Comprehensive Inorganic Chemistry III. Volume 6: Heterogeneous Inorganic Catalysis [6, 3 ed.] 9780128231449

Citation preview

COMPREHENSIVE INORGANIC CHEMISTRY III

COMPREHENSIVE INORGANIC CHEMISTRY III EDITORS IN CHIEF

Jan Reedijk Leiden Institute of Chemistry, Leiden University, Leiden, the Netherlands

Kenneth R. Poeppelmeier Department of Chemistry, Northwestern University, Evanston, IL, United States

VOLUME 6

Heterogeneous Inorganic Catalysis VOLUME EDITORS

Rutger A. van Santen Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, the Netherlands

Emiel Hensen Laboratory of Inorganic Materials and Catalysis, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, the Netherlands

Amsterdam • Boston • Heidelberg • London • New York • Oxford Paris • San Diego • San Francisco • Singapore • Sydney • Tokyo

Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom 50 Hampshire Street, 5th Floor, Cambridge MA 02139, United States Copyright Ó 2023 Elsevier Ltd. All rights reserved No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers may always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN 978-0-12-823144-9

For information on all publications visit our website at http://store.elsevier.com

Publisher: Oliver Walter Acquisitions Editors: Clodagh Holland-Borosh and Blerina Osmanaj Content Project Manager: Pamela Sadhukhan Associate Content Project Manager: Abraham Lincoln Samuel Designer: Victoria Pearson Esser

CONTENTS OF VOLUME 6 Editor Biographies

vii

Volume Editors

ix

Contributors to Volume 6

xv

Preface

xix

6.01

Introduction: A short history of single site catalysis Rutger A van Santen and Emiel JM Hensen

6.02

Synthesis and application of (nano) zeolites Ana Palcic and Valentin Valtchev

18

6.03

Mesostructured materials Feng Yu and Feng-Shou Xiao

41

6.04

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts Christophe Copéret and Maciej Damian Korzynski

67

Challenges with atomically dispersed supported metal catalysts: Controlling performance, improving stability, and enhancing metal loading Samira Fatma Kurtoglu-Öztulum and Alper Uzun

86

6.05

1

6.06

Metal containing nanoclusters in zeolites Guanna Li and Evgeny A Pidko

112

6.07

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites Dieter Plessers, Max L Bols, Hannah M Rhoda, Alexander J Heyer, Edward I Solomon, Bert F Sels, and Robert A Schoonheydt

148

6.08

Dynamic evolution of catalytic active sites within zeolite catalysis Massimo Bocus, Samuel E Neale, Pieter Cnudde, and Véronique Van Speybroeck

165

6.09

Nanocluster heterogeneous catalysts: Insights from theory Geng Sun and Philippe Sautet

201

6.10

Imaging of single atom catalysts Stephen Porter and Abhaya K Datye

222

6.11

Metal-support interfaces in ceria-based catalysts Valery Muravev, Nikolay Kosinov, and Emiel JM Hensen

244

v

vi

Contents of Volume 6

6.12

Solid acid catalysis; Part I, the zeolite protonic site Rutger A van Santen

271

6.13

Solid acid catalysis; Part II, catalytic chemistry of proton activation Rutger A van Santen

293

6.14

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins Hao Zhang, Emiel JM Hensen, and Nikolay Kosinov

311

6.15

Inorganic catalysis for methane conversion to chemicals Guangzong Fang, Dunfeng Gao, Xiulian Pan, Guoxiong Wang, and Xinhe Bao

327

6.16

Promoted Fischer-Tropsch catalysts Paul B Webb and Ivo AW Filot

354

6.17

Selective oxidation by mixed metal nanoparticles Hannah Rogers and Simon J Freakley

381

6.18

Doped semiconductor photocatalysts Takashi Hisatomi and Kazunari Domen

401

6.19

Structure-reactivity relations in electrocatalysis Sebastian A Watzele, Batyr Garlyyev, Elena Gubanova, and Aliaksandr S Bandarenka

419

EDITOR BIOGRAPHIES Editors in Chief Jan Reedijk Jan Reedijk (1943) studied chemistry at Leiden University where he completed his Ph.D. (1968). After a few years in a junior lecturer position at Leiden University, he accepted a readership at Delft University of Technology in 1972. In 1979 he accepted a call for Professor of Chemistry at Leiden University. After 30 years of service, he retired from teaching in 2009 and remained as an emeritus research professor at Leiden University. In Leiden he has acted as Chair of the Department of Chemistry, and in 1993 he became the Founding Director of the Leiden Institute of Chemistry. His major research activities have been in Coordination Chemistry and Bioinorganic Chemistry, focusing on biomimetic catalysis, molecular materials, and medicinal inorganic chemistry. Jan Reedijk was elected member of the Royal Netherlands Academy of Sciences in 1996 and he was knighted by the Queen of the Netherlands to the order of the Dutch Lion (2008). He is also lifetime member of the Finnish Academy of Sciences and Letters and of Academia Europaea. He has held visiting professorships in Cambridge (UK), Strasbourg (France), Münster (Germany), Riyadh (Saudi Arabia), Louvain-la-Neuve (Belgium), Dunedin (New Zealand), and Torun (Poland). In 1990 he served as President of the Royal Netherlands Chemical Society. He has acted as the Executive Secretary of the International Conferences of Coordination Chemistry (1988–2012) and served IUPAC in the Division of Inorganic Chemistry, first as a member and later as (vice/past) president between 2005 and 2018. After his university retirement he remained active as research consultant and in IUPAC activities, as well as in several editorial jobs. For Elsevier, he acted as Editor-in-Chief of the Reference Collection in Chemistry (2013–2019), and together with Kenneth R. Poeppelmeier for Comprehensive Inorganic Chemistry II (2008–2013) and Comprehensive Inorganic Chemistry III (2019-present). From 2018 to 2020, he co-chaired the worldwide celebrations of the International Year of the Periodic Table 2019. Jan Reedijk has published over 1200 papers (1965–2022; cited over 58000 times; h ¼ 96). He has supervised 90 Ph.D. students, over 100 postdocs, and over 250 MSc research students. Kenneth R. Poeppelmeier Kenneth R. Poeppelmeier (1949) completed his undergraduate studies in chemistry at the University of Missouri (1971) and then volunteered as an instructor at Samoa CollegedUnited States Peace Corps in Western Samoa (1971–1974). He completed his Ph.D. (1978) in Inorganic Chemistry with John Corbett at Iowa State University (1978). He joined the catalysis research group headed by John Longo at Exxon Research and Engineering Company, Corporate Research–Science Laboratories (1978–1984), where he collaborated with the reforming science group and Exxon Chemicals to develop the first zeolite-based light naphtha reforming catalyst. In 1984 he joined the Chemistry Department at Northwestern University and the recently formed Center for Catalysis and Surface Science (CCSS). He is the Charles E. and Emma H. Morrison Professor of Chemistry at Northwestern University and a NAISE Fellow joint with Northwestern University and the Chemical Sciences and Engineering Division, Argonne National Laboratory. Leadership positions held include Director, CCSS, Northwestern University; Associate Division Director for Science, Chemical Sciences and Engineering Division, Argonne National Laboratory; President of the Chicago Area Catalysis Club; Associate Director, NSF Science and Technology Center for Superconductivity; and Chairman of the ACS Solid State Subdivision of the Division of Inorganic Chemistry. His major research activities have been in Solid State and Inorganic Materials Chemistry focusing on heterogeneous catalysis, solid state chemistry, and materials chemistry. His awards include National Science Council of Taiwan Lecturer (1991); Dow Professor of Chemistry (1992–1994); AAAS Fellow, the American Association for the Advancement of Science (1993); JSPS Fellow, Japan Society for the Promotion of Science (1997); Natural Science Foundation of China Lecturer (1999); National Science Foundation Creativity Extension Award (2000

vii

viii

Editor Biographies

and 2022); Institut Universitaire de France Professor (2003); Chemistry Week in China Lecturer (2004); Lecturer in Solid State Chemistry, China (2005); Visitantes Distinguidos, Universid Complutenses Madrid (2008); Visiting Professor, Chinese Academy of Sciences (2011); 20 years of Service and Dedication Award to Inorganic Chemistry (2013); Elected foreign member of Spanish National Academy: Real Academia de Ciencia, Exactas, Fisicas y Naturales (2017); Elected Honorary Member of the Royal Society of Chemistry of Spain (RSEQ) (2018); and the TianShan Award, Xinjiang Uygur Autonomous Region of China (2021). He has organized and was Chairman of the Chicago Great Lakes Regional ACS Symposium on Synthesis and Processing of Advanced Solid State Materials (1987), the New Orleans National ACS Symposium on Solid State Chemistry of Heterogeneous Oxide Catalysis, Including New Microporous Solids (1987), the Gordon Conference on Solid State Chemistry (1994) and the First European Gordon Conference on Solid State Chemistry (1995), the Spring Materials Research Society Symposium on Environmental Chemistry (1995), the Advisory Committee of Intense Pulsed Neutron Source (IPNS) Program (1996–1998), the Spring Materials Research Society Symposium on Perovskite Materials (2003), the 4th International Conference on Inorganic Materials, University of Antwerp (2004), and the Philadelphia National ACS Symposium on Homogeneous and Heterogeneous Oxidation Catalysis (2004). He has served on numerous Editorial Boards, including Chemistry of Materials, Journal of Alloys and Compounds, Solid State Sciences, Solid State Chemistry, and Science China Materials, and has been a co-Editor for Structure and Bonding, an Associate Editor for Inorganic Chemistry, and co-Editor-in-Chief with Jan Reedijk for Comprehensive Inorganic Chemistry II (published 2013) and Comprehensive Inorganic Chemistry III (to be published in 2023). In addition, he has served on various Scientific Advisory Boards including for the World Premier International Research Center Initiative and Institute for Integrated Cell-Material Sciences Kyoto University, the European Center SOPRANO on Functional Electronic Metal Oxides, the Kyoto University Mixed-Anion Project, and the Dresden Max Planck Institute for Chemistry and Physics. Kenneth Poeppelmeier has published over 500 papers (1971–2022) and cited over 28000 times (h-index ¼ 84). He has supervised over 200 undergraduates, Ph.D. students, postdocs, and visiting scholars in research.

VOLUME EDITORS Risto S. Laitinen Risto S. Laitinen is Professor Emeritus of Chemistry at the University of Oulu, Finland. He received the M.Sc and Ph.D. degrees from Helsinki University of Technology (currently Aalto University). His research interests are directed to synthetic, structural, and computational chemistry of chalcogen compounds focusing on selenium and tellurium. He has published 250 peer-reviewed articles and 15 book chapters and has edited 2 books: Selenium and Tellurium Reagents: In Chemistry and Materials Science with Raija Oilunkaniemi (Walther de Gruyter, 2019) and Selenium and Tellurium Chemistry: From Small Molecules to Biomolecules and Materials with Derek Woollins (Springer, 2011). He has also written 30 professional and popular articles in chemistry. He is the Secretary of the Division of Chemical Nomenclature and Structure Representation, International Union of Pure and Applied Chemistry, for the term 2016–2023. He served as Chair of the Board of Union of Finnish University Professors in 2007–2010. In 2017, Finnish Cultural Foundation (North Ostrobothnia regional fund) gave him an award for excellence in his activities for science and music. He has been a member of Finnish Academy of Science and Letters since 2003.

Vincent L. Pecoraro Professor Vincent L. Pecoraro is a major contributor in the fields of inorganic, bioinorganic, and supramolecular chemistries. He has risen to the upper echelons of these disciplines with over 300 publications (an h-index of 92), 4 book editorships, and 5 patents. He has served the community in many ways including as an Associate Editor of Inorganic Chemistry for 20 years and now is President of the Society of Biological Inorganic Chemistry. Internationally, he has received a Le Studium Professorship, Blaise Pascal International Chair for Research, the Alexander von Humboldt Stiftung, and an Honorary PhD from Aix-Maseille University. His many US distinctions include the 2016 ACS Award for Distinguished Service in the Advancement of Inorganic Chemistry, the 2021 ACS/SCF FrancoAmerican Lectureship Prize, and being elected a Fellow of the ACS and AAAS. He also recently cofounded a Biomedical Imaging company, VIEWaves. In 2022, he was ranked as one of the world’s top 1000 most influential chemists.

ix

x

Volume Editors

Zijian Guo Professor Zijian Guo received his Ph.D. from the University of Padova and worked as a postdoctoral research fellow at Birkbeck College, the University of London. He also worked as a research associate at the University of Edinburgh. His research focuses on the chemical biology of metals and metallodrugs and has authored or co-authored more than 400 peer-reviewed articles on basic and applied research topics. He was awarded the First Prize in Natural Sciences from Ministry of Education of China in 2015, the Luigi Sacconi Medal from the Italian Chemical Society in 2016, and the Outstanding Achievement Award from the Society of the Asian Biological Inorganic Chemistry in 2020. He founded Chemistry and Biomedicine Innovation Center (ChemBIC) in Nanjing University in 2019, and is serving as the Director of ChemBIC since then. He was elected to the Fellow of the Chinese Academy of Sciences in 2017. He served as Associated Editor of Coord. Chem. Rev and an editorial board member of several other journals.

Daniel C. Fredrickson Daniel C. Fredrickson is a Professor in the Department of Chemistry at the University of WisconsinMadison. He completed his BS in Biochemistry at the University of Washington in 2000, where he gained his first experiences with research and crystals in the laboratory of Prof. Bart Kahr. At Cornell University, he then pursued theoretical investigations of bonding in intermetallic compounds, the vast family of solid state compounds based on metallic elements, under the mentorship of Profs. Stephen Lee and Roald Hoffmann, earning his Ph.D. in 2005. Interested in the experimental and crystallographic aspects of complex intermetallics, he then carried out postdoctoral research from 2005 to 2008 with Prof. Sven Lidin at Stockholm University. Since starting at UW-Madison in 2009, his research group has created theory-driven approaches to the synthesis and discovery of new intermetallic phases and understanding the origins of their structural features. Some of his key research contributions are the development of the DFT-Chemical Pressure Method, the discovery of isolobal bonds for the generalization of the 18 electron rule to intermetallic phases, models for the emergence of incommensurate modulations in these compounds, and various design strategies for guiding complexity in solid state structures.

Patrick M. Woodward Professor Patrick M. Woodward received BS degrees in Chemistry and General Engineering from Idaho State University in 1991, an MS in Materials Science, and a Ph.D. in Chemistry from Oregon State University (1996) under the supervision of Art Sleight. After a postdoctoral appointment in the Physics Department at Brookhaven National Laboratory (1996–1998), he joined the faculty at Ohio State University in 1998, where he holds the rank of Professor in the Department of Chemistry and Biochemistry. He is a Fellow of the American Chemical Society (2020) and a recipient of an Alfred P. Sloan Fellowship (2004) and an NSF Career Award (2001). He has co-authored two textbooks: Solid State Materials Chemistry and the popular general chemistry textbook, Chemistry: The Central Science. His research interests revolve around the discovery of new materials and understanding links between the composition, structure, and properties of extended inorganic and hybrid solids.

Volume Editors

xi

P. Shiv Halasyamani Professor P. Shiv Halasyamani earned his BS in Chemistry from the University of Chicago (1992) and his Ph.D. in Chemistry under the supervision of Prof. Kenneth R. Poeppelmeier at Northwestern University (1996). He was a Postdoctoral Fellow and Junior Research Fellow at Christ Church College, Oxford University, from 1997 to 1999. He began his independent academic career in the Department of Chemistry at the University of Houston in 1999 and has been a Full Professor since 2010. He was elected as a Fellow of the American Association for the Advancement of Science (AAAS) in 2019 and is currently an Associate Editor of the ACS journals Inorganic Chemistry and ACS Organic & Inorganic Au. His research interests involve the design, synthesis, crystal growth, and characterization of new functional inorganic materials.

Ram Seshadri Ram Seshadri received his Ph.D. in Solid State Chemistry from the Indian Institute of Science (IISc), Bangalore, working under the guidance of Professor C. N. R. Rao FRS. After some years as a Postdoctoral Fellow in Europe, he returned to IISc as an Assistant Professor in 1999. He moved to the Materials Department (College of Engineering) at UC Santa Barbara in 2002. He was recently promoted to the rank of Distinguished Professor in the Materials Department and the Department of Chemistry and Biochemistry in 2020. He is also the Fred and Linda R. Wudl Professor of Materials Science and Director of the Materials Research Laboratory: A National Science Foundation Materials Research Science and Engineering Center (NSF-MRSEC). His work broadly addresses the topic of structure–composition– property relations in crystalline inorganic and hybrid materials, with a focus on magnetic materials and materials for energy conversion and storage. He is Fellow of the Royal Society of Chemistry, the American Physical Society, and the American Association for the Advancement of Science. He serves as Associate Editor of the journals, Annual Reviews of Materials Research and Chemistry of Materials.

Serena Cussen Serena Cussen née Corr studied chemistry at Trinity College Dublin, completing her doctoral work under Yurii Gun’ko. She then joined the University of California, Santa Barbara, working with Ram Seshadri as a postdoctoral researcher. She joined the University of Kent as a lecturer in 2009. She moved to the University of Glasgow in 2013 and was made Professor in 2018. She moved to the University of Sheffield as a Chair in Functional Materials and Professor in Chemical and Biological Engineering in 2018, where she now serves as Department Head. She works on next-generation battery materials and advanced characterization techniques for the structure and properties of nanomaterials. Serena is the recipient of several awards including the Journal of Materials Chemistry Lectureship of the Royal Society of Chemistry. She previously served as Associate Editor of Royal Society of Chemistry journal Nanoscale and currently serves as Associate Editor for the Institute of Physics journal Progress in Energy.

xii

Volume Editors

Rutger A. van Santen Rutger A. van Santen received his Ph.D. in 1971 in Theoretical Chemistry from the University of Leiden, The Netherlands. In the period 1972–1988, he became involved with catalysis research when employed by Shell Research in Amsterdam and Shell Development Company in Houston. In 1988, he became Full Professor of Catalysis at the Technical University Eindhoven. From 2010 till now he is Emeritus Professor and Honorary Institute Professor at Technical University Eindhoven. He is a member of Royal Dutch Academy of Sciences and Arts and Foreign Associate of the United States National Academy of Engineering (NAE). He has received several prestigious awards: the 1981 golden medal of the Royal Dutch Chemical Society; in 1992, the F.G. Chiappetta award of the North American Catalysis Society; in 1997, the Spinoza Award from the Dutch Foundation for Pure and Applied Research; and in 2001, the Alwin Mittasch Medal Dechema, Germany, among others. His main research interests are computational heterogeneous catalysis and complex chemical systems theory. He has published over 700 papers, 16 books, and 22 patents.

Emiel J. M. Hensen Emiel J. M. Hensen received his Ph.D. in Catalysis in 2000 from Eindhoven University of Technology, The Netherlands. Between 2000 and 2008, he worked at the University of Amsterdam, Shell Research in Amsterdam, and Eindhoven University of Technology on several topics in the field of heterogeneous catalysis. Since July 2009, he is Full Professor of Inorganic Materials and Catalysis at TU/e. He was a visiting professor at the Katholieke Universiteit Leuven (Belgium, 2001–2016) and at Hokkaido University (Japan, 2016). He is principal investigator and management team member of the gravitation program Multiscale Catalytic Energy Conversion, elected member of the Advanced Research Center Chemical Building Blocks Consortium, and chairman of the Netherlands Institute for Catalysis Research (NIOK). Hensen was Head of the Department of Chemical Engineering and Chemistry of Eindhoven University of Technology from 2016 to 2020. Hensen received Veni, Vidi, Vici, and Casmir grant awards from the Netherlands Organisation for Scientific Research. His main interests are in mechanism of heterogeneous catalysis combining experimental and computation studies. He has published over 600 papers, 20 book chapters, and 7 patents.

Artem M. Abakumov Artem M. Abakumov graduated from the Department of Chemistry at Moscow State University in 1993, received his Ph.D. in Chemistry from the same University in 1997, and then continued working as a Researcher and Vice-Chair of Inorganic Chemistry Department. He spent about 3 years as a postdoctoral fellow and invited professor in the Electron Microscopy for Materials Research (EMAT) laboratory at the University of Antwerp and joined EMAT as a research leader in 2008. Since 2015 he holds a Full Professor position at Skolkovo Institute of Science and Technology (Skoltech) in Moscow, leading Skoltech Center for Energy Science and Technology as a Director. His research interests span over a wide range of subjects, from inorganic chemistry, solid state chemistry, and crystallography to battery materials and transmission electron microscopy. He has extended experience in characterization of metal-ion battery electrodes and electrocatalysts with advanced TEM techniques that has led to a better understanding of charge–discharge mechanisms, redox reactions, and associated structural transformations in various classes of cathode materials on different spatial scales. He has published over 350 papers, 5 book chapters, and 12 patents.

Volume Editors

xiii

Keith J. Stevenson Keith J. Stevenson received his Ph.D. in 1997 from the University of Utah under the supervision of Prof. Henry White. Subsequently, he held a postdoctoral appointment at Northwestern University (1997– 2000) and a tenured faculty appointment (2000–2015) at the University of Texas at Austin. At present, he is leading the development of a new graduate level university (Skolkovo Institute for Science and Technology) in Moscow, Russia, where he is Provost and the former Director of the Center for Energy Science and Technology (CEST). To date he has published over 325 peer-reviewed publications, 14 patents, and 6 book chapters in this field. He is a recipient of Society of Electroanalytical Chemistry Charles N. Reilley Award (2021).

Evgeny V. Antipov Evgeny V. Antipov graduated from the Department of Chemistry at Moscow State University in 1981, received his Ph.D. in Chemistry in 1986, DSc degree in Chemistry in 1998, and then continued working at the same University as a Researcher, Head of the Laboratory of Inorganic Crystal Chemistry, Professor, Head of Laboratory of fundamental research on aluminum production, and Head of the Department of Electrochemistry. Since 2018 he also holds a professor position at Skolkovo Institute of Science and Technology (Skoltech) in Moscow. Currently his research interests are mainly focused on inorganic materials for application in batteries and fuel cells. He has published more than 400 scientific articles and 14 patents.

Vivian W.W. Yam Professor Vivian W.W. Yam is the Chair Professor of Chemistry and Philip Wong Wilson Wong Professor in Chemistry and Energy at The University of Hong Kong. She received both her B.Sc (Hons) and Ph.D. from The University of Hong Kong. She was elected to Member of Chinese Academy of Sciences, International Member (Foreign Associate) of US National Academy of Sciences, Foreign Member of Academia Europaea, Fellow of TWAS, and Founding Member of Hong Kong Academy of Sciences. She was Laureate of 2011 L’Oréal-UNESCO For Women in Science Award. Her research interests include inorganic and organometallic chemistry, supramolecular chemistry, photophysics and photochemistry, and metal-based molecular functional materials for sensing, organic optoelectronics, and energy research. Also see: https://chemistry.hku.hk/wwyam.

xiv

Volume Editors

David L. Bryce David L. Bryce (B.Sc (Hons), 1998, Queen’s University; Ph.D., 2002, Dalhousie University; postdoctoral fellow, 2003–04, NIDDK/NIH) is Full Professor and University Research Chair in Nuclear Magnetic Resonance at the University of Ottawa in Canada. He is the past Chair of the Department of Chemistry and Biomolecular Sciences, a Fellow of the Royal Society of Chemistry, and an elected Fellow of the Royal Society of Canada. His research interests include solid-state NMR of low-frequency quadrupolar nuclei, NMR studies of materials, NMR crystallography, halogen bonding, mechanochemistry, and quantum chemical interpretation of NMR interaction tensors. He is the author of approximately 200 scientific publications and co-author of 1 book. He is the Editor-in-Chief of Solid State Nuclear Magnetic Resonance and Section Editor (Magnetic Resonance and Molecular Spectroscopy) for the Canadian Journal of Chemistry. He has served as the Chair of Canada’s National Ultrahigh-Field NMR Facility for Solids and is a past co-chair of the International Society for Magnetic Resonance conference and of the Rocky Mountain Conference on Magnetic Resonance Solid-State NMR Symposium. His work has been recognized with the Canadian Society for Chemistry Keith Laidler Award and with the Gerhard Herzberg Award of the Canadian Association of Analytical Sciences and Spectroscopy.

Paul R. Raithby Paul R. Raithby obtained his B.Sc (1973) and Ph.D. (1976) from Queen Mary College, University of London, working for his Ph.D. in structural inorganic chemistry. He moved to the University of Cambridge in 1976, initially as a postdoctoral researcher and then as a faculty member. In 2000, he moved to the University of Bath to take up the Chair of Inorganic Chemistry when he remains to the present day, having been awarded an Emeritus Professorship in 2022. His research interests have spanned the chemistry of transition metal cluster compounds, platinum acetylide complexes and oligomers, and lanthanide complexes, and he uses laboratory and synchrotron-based X-ray crystallographic techniques to determine the structures of the complexes and to study their dynamics using time-resolved photocrystallographic methods.

Angus P. Wilkinson

Angus P. Wilkinson completed his bachelors (1988) and doctoral (1992) degrees in chemistry at Oxford University in the United Kingdom. He spent a postdoctoral period in the Materials Research Laboratory, University of California, Santa Barbara, prior to joining the faculty at the Georgia Institute of Technology as an assistant professor in 1993. He is currently a Professor in both the Schools of Chemistry and Biochemistry, and Materials Science and Engineering, at the Georgia Institute of Technology. His research in the general area of inorganic materials chemistry makes use of synchrotron X-ray and neutron scattering to better understand materials synthesis and properties.

CONTRIBUTORS TO VOLUME 6 Aliaksandr S Bandarenka Physics of Energy Conversion and Storage, PhysikDepartment, Technische Universität München, Garching bei München, Germany Xinhe Bao The State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, The Chinese Academy of Science, Beijing, China; and University of Science and Technology of China, Hebei, China Massimo Bocus Center for Molecular Modeling, Ghent University, Tech Lane Ghent Science Park Campus A, Zwijnaarde, Belgium Max L Bols Department of Microbial and Molecular Systems, Centre for Sustainable Catalysis and Engineering, KU Leuven, Leuven, Belgium Pieter Cnudde Center for Molecular Modeling, Ghent University, Tech Lane Ghent Science Park Campus A, Zwijnaarde, Belgium Christophe Copéret Department of Chemistry and Applied Biosciences, ETH Zürich, Zürich, Switzerland Abhaya K Datye Department of Chemical and Biological Engineering, University of New Mexico, Albuquerque, NM, United States Kazunari Domen Research Initiative for Supra-Materials, Shinshu University, Nagano, Japan Guangzong Fang The State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, The Chinese Academy of Science, Beijing, China

Ivo AW Filot Eindhoven University of Technology, Inorganic Materials and Catalysis, Eindhoven, The Netherlands Simon J Freakley Department of Chemistry, University of Bath, Bath, United Kingdom Dunfeng Gao The State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, The Chinese Academy of Science, Beijing, China Batyr Garlyyev Physics of Energy Conversion and Storage, PhysikDepartment, Technische Universität München, Garching bei München, Germany Elena Gubanova Physics of Energy Conversion and Storage, Physik-Department, Technische Universität München, Garching bei München, Germany Emiel JM Hensen Laboratory of Inorganic Materials and Catalysis, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, MB Eindhoven, The Netherlands Alexander J Heyer Department of Chemistry, Stanford University, Stanford, CA, United States Takashi Hisatomi Research Initiative for Supra-Materials, Shinshu University, Nagano, Japan; and PRESTO, Japan Science and Technology Agency (JST), Nagano, Japan Maciej Damian Korzy nski Department of Chemistry and Applied Biosciences, ETH Zürich, Zürich, Switzerland

xv

xvi

Contributors to Volume 6

Nikolay Kosinov Laboratory of Inorganic Materials and Catalysis, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, MB Eindhoven, The Netherlands Samira Fatma Kurto glu-Öztulum Department of Chemical and Biological Engineering, Koç University, Istanbul, Turkey; and Koç University TÜPRAŞ Energy Center (KUTEM), Koç University, Istanbul, Turkey Guanna Li Biobased Chemistry and Technology, Wageningen University & Research, Wageningen, The Netherlands; and Laboratory of Organic Chemistry, Wageningen University & Research, Wageningen, The Netherlands Valery Muravev Laboratory of Inorganic Materials and Catalysis, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, MB Eindhoven, The Netherlands

Hannah Rogers Department of Chemistry, University of Bath, Bath, United Kingdom Philippe Sautet Department of Chemical and Biomolecular Engineering, University of California, Los Angeles, Los Angeles, CA, United States; Department of Chemistry and Biochemistry, University of California, Los Angeles, Los Angeles, CA, United States; and California Nano Systems Institute, Los Angeles, CA, United States Robert A Schoonheydt Department of Microbial and Molecular Systems, Centre for Sustainable Catalysis and Engineering, KU Leuven, Leuven, Belgium Bert F Sels Department of Microbial and Molecular Systems, Centre for Sustainable Catalysis and Engineering, KU Leuven, Leuven, Belgium

Samuel E Neale Center for Molecular Modeling, Ghent University, Tech Lane Ghent Science Park Campus A, Zwijnaarde, Belgium

Edward I Solomon Department of Chemistry, Stanford University, Stanford, CA, United States; and Photon Science, SLAC National Accelerator Laboratory, Menlo Park, CA, United States

Ana Palcic Division of Materials Chemistry, Laboratory for Synthesis of New Materials, RuCer Boskovic Institute, Zagreb, Croatia

Geng Sun Department of Chemical and Biomolecular Engineering, University of California, Los Angeles, Los Angeles, CA, United States

Xiulian Pan The State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, The Chinese Academy of Science, Beijing, China

Alper Uzun Department of Chemical and Biological Engineering, Koç University, Istanbul, Turkey; Koç University TÜPRAŞ Energy Center (KUTEM), Koç University, Istanbul, Turkey; and Koç University Surface Science and Technology Center (KUYTAM), Koç University, Istanbul, Turkey

Evgeny A Pidko Inorganic Systems Engineering group, Department of Chemical Engineering, Faculty of Applied Sciences, Delft University of Technology, Delft, The Netherlands Dieter Plessers Department of Microbial and Molecular Systems, Centre for Sustainable Catalysis and Engineering, KU Leuven, Leuven, Belgium Stephen Porter Department of Chemical and Biological Engineering, University of New Mexico, Albuquerque, NM, United States Hannah M Rhoda Department of Chemistry, Stanford University, Stanford, CA, United States

Valentin Valtchev Laboratoire Catalyse et Spectrochimie, Normandie Université, ENSICAEN, UNICAEN, CNRS, Caen Cedex, France Rutger A van Santen Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, MB Eindhoven, The Netherlands Véronique Van Speybroeck Center for Molecular Modeling, Ghent University, Tech Lane Ghent Science Park Campus A, Zwijnaarde, Belgium

Contributors to Volume 6

Guoxiong Wang The State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, The Chinese Academy of Science, Beijing, China Sebastian A Watzele Physics of Energy Conversion and Storage, Physik-Department, Technische Universität München, Garching bei München, Germany Paul B Webb University of St Andrews, School of Chemistry, St Andrews, Fife, United Kingdom

Feng-Shou Xiao Key Lab of Applied Chemistry of Zhejiang Province, Department of Chemistry, Zhejiang University, Hangzhou, China; and Key Lab of Biomass Chemical Engineering of Ministry of Education, College of Chemical and Biological Engineering, Zhejiang University, Hangzhou, China Feng Yu Key Lab of Applied Chemistry of Zhejiang Province, Department of Chemistry, Zhejiang University, Hangzhou, China Hao Zhang Laboratory of Inorganic Materials and Catalysis, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, Eindhoven, The Netherlands

xvii

PREFACE Comprehensive Inorganic Chemistry III is a new multi-reference work covering the broad area of Inorganic Chemistry. The work is available both in print and in electronic format. The 10 Volumes review significant advances and examines topics of relevance to today’s inorganic chemists with a focus on topics and results after 2012. The work is focusing on new developments, including interdisciplinary and high-impact areas. Comprehensive Inorganic Chemistry III, specifically focuses on main group chemistry, biological inorganic chemistry, solid state and materials chemistry, catalysis and new developments in electrochemistry and photochemistry, as well as on NMR methods and diffractions methods to study inorganic compounds. The work continues our 2013 work Comprehensive Inorganic Chemistry II, but at the same time adds new volumes on emerging research areas and techniques used to study inorganic compounds. The new work is also highly complementary to other recent Elsevier works in Coordination Chemistry and Organometallic Chemistry thereby forming a trio of works covering the whole of modern inorganic chemistry, most recently COMC-4 and CCC-3. The rapid pace of developments in recent years in all areas of chemistry, particularly inorganic chemistry, has again created many challenges to provide a contemporary up-to-date series. As is typically the challenge for Multireference Works (MRWs), the chapters are designed to provide a valuable long-standing scientific resource for both advanced students new to an area as well as researchers who need further background or answers to a particular problem on the elements, their compounds, or applications. Chapters are written by teams of leading experts, under the guidance of the Volume Editors and the Editors-inChief. The articles are written at a level that allows undergraduate students to understand the material, while providing active researchers with a ready reference resource for information in the field. The chapters are not intended to provide basic data on the elements, which are available from many sources including the original CIC-I, over 50-years-old by now, but instead concentrate on applications of the elements and their compounds and on high-level techniques to study inorganic compounds. Vol. 1: Synthesis, Structure, and Bonding in Inorganic Molecular Systems; Risto S. Laitinen In this Volume the editor presents an historic overview of Inorganic Chemistry starting with the birth of inorganic chemistry after Berzelius, and a focus on the 20th century including an overview of “inorganic” Nobel Prizes and major discoveries, like inert gas compounds. The most important trends in the field are discussed in an historic context. The bulk of the Volume consists of 3 parts, i.e., (1) Structure, bonding, and reactivity in inorganic molecular systems; (2) Intermolecular interactions, and (3) Inorganic Chains, rings, and cages. The volume contains 23 chapters. Part 1 contains chapters dealing with compounds in which the heavy p-block atom acts as a central atom. Some chapters deal with the rich synthetic and structural chemistry of noble gas compounds, low-coordinate p-block elements, biradicals, iron-only hydrogenase mimics, and macrocyclic selenoethers. Finally, the chemistry and application of weakly coordinating anions, the synthesis, structures, and reactivity of carbenes containing non-innocent ligands, frustrated Lewis pairs in metal-free catalysis are discussed. Part 2 discusses secondary bonding interactions that play an important role in the properties of bulk materials. It includes a chapter on the general theoretical considerations of secondary bonding interactions, including halogen and chalcogen bonding. This section is concluded by the update of the host-guest chemistry of the molecules of p-block elements and by a comprehensive review of closed-shell metallophilic interactions. The third part of the Volume is dedicated to chain, ring and cage (or cluster) compounds in molecular inorganic chemistry. Separate

xix

xx

Preface

chapters describe the recent chemistry of boron clusters, as well as the chain, ring, and cage compounds of Group13 and 15, and 16 elements. Also, aromatic compounds bearing heavy Group 14 atoms, polyhalogenide anions and Zintl-clusters are presented. Vol. 2: Bioinorganic Chemistry and Homogeneous Biomimetic Inorganic Catalysis; Vincent L. Pecoraro and Zijian Guo In this Volume, the editors have brought together 26 chapters providing a broad coverage of many of the important areas involving metal compounds in biology and medicine. Readers interested in fundamental biochemistry that is assisted by metal ion catalysis, or in uncovering the latest developments in diagnostics or therapeutics using metal-based probes or agents, will find high-level contributions from top scientists. In the first part of the Volume topics dealing with metals interacting with proteins and nucleic acids are presented (e.g., siderophores, metallophores, homeostasis, biomineralization, metal-DNA and metal-RNA interactions, but also with zinc and cobalt enzymes). Topics dealing with iron-sulfur clusters and heme-containing proteins, enzymes dealing with dinitrogen fixation, dihydrogen and dioxygen production by photosynthesis will also be discussed, including bioinspired model systems. In the second part of the Volume the focus is on applications of inorganic chemistry in the field of medicine: e.g., clinical diagnosis, curing diseases and drug targeting. Platinum, gold and other metal compounds and their mechanism of action will be discussed in several chapters. Supramolecular coordination compounds, metal organic frameworks and targeted modifications of higher molecular weight will also be shown to be important for current and future therapy and diagnosis. Vol. 3: Theory and Bonding of Inorganic Non-molecular Systems; Daniel C. Fredrickson This volume consists of 15 chapters that build on symmetry-based expressions for the wavefunctions of extended structures toward models for bonding in solid state materials and their surfaces, algorithms for the prediction of crystal structures, tools for the analysis of bonding, and theories for the unique properties and phenomena that arise in these systems. The volume is divided into four parts along these lines, based on major themes in each of the chapters. These are: Part 1: Models for extended inorganic structures, Part 2: Tools for electronic structure analysis, Part 3: Predictive exploration of new structures, and Part 4: Properties and phenomena. Vol. 4: Solid State Inorganic Chemistry; P. Shiv Halasyamani and Patrick M. Woodward In a broad sense the field of inorganic chemistry can be broken down into substances that are based on molecules and those that are based on extended arrays linked by metallic, covalent, polar covalent, or ionic bonds (i.e., extended solids). The field of solid-state inorganic chemistry is largely concerned with elements and compounds that fall into the latter group. This volume contains nineteen chapters covering a wide variety of solid-state inorganic materials. These chapters largely focus on materials with properties that underpin modern technology. Smart phones, solid state lighting, batteries, computers, and many other devices that we take for granted would not be possible without these materials. Improvements in the performance of these and many other technologies are closely tied to the discovery of new materials or advances in our ability to synthesize high quality samples. The organization of most chapters is purposefully designed to emphasize how the exceptional physical properties of modern materials arise from the interplay of composition, structure, and bonding. Not surprisingly this volume has considerable overlap with both Volume 3 (Theory and Bonding of Inorganic NonMolecular Systems) and Volume 5 (Inorganic Materials Chemistry). We anticipate that readers who are interested in this volume will find much of interest in those volumes and vice versa Vol. 5: Inorganic Materials Chemistry; Ram Seshadri and Serena Cussen This volume has adopted the broad title of Inorganic Materials Chemistry, but as readers would note, the title could readily befit articles in other volumes as well. In order to distinguish contributions in this volume from

Preface

xxi

those in other volumes, the editors have chosen to use as the organizing principle, the role of synthesis in developing materials, reflected by several of the contributions carrying the terms “synthesis” or “preparation” in the title. It should also be noted that the subset of inorganic materials that are the focus of this volume are what are generally referred to as functional materials, i.e., materials that carry out a function usually through the way they respond to an external stimulus such as light, or thermal gradients, or a magnetic field.

Vol. 6: Heterogeneous Inorganic Catalysis; Rutger A. van Santen and Emiel J. M. Hensen This Volume starts with an introductory chapter providing an excellent discussion of single sites in metal catalysis. This chapter is followed by 18 chapters covering a large part of the field. These chapters have been written with a focus on the synthesis and characterization of catalytic complexity and its relationship with the molecular chemistry of the catalytic reaction. In the 1950s with the growth of molecular inorganic chemistry, coordination chemistry and organometallic chemistry started to influence the development of heterogeneous catalysis. A host of new reactions and processes originate from that time. In this Volume chapters on major topics, like promoted Fischer-Tropsch catalysts, structure sensitivity of well-defined alloy surfaces in the context of oxidation catalysis and electrocatalytic reactions, illustrate the broadness of the field. Molecular heterogeneous catalysts rapidly grew after high-surface synthetic of zeolites were introduced; so, synthesis, structure and nanopore chemistry in zeolites is presented in a number of chapters. Also, topics like nanocluster activation of zeolites and supported zeolites are discussed. Mechanistically important chapters deal with imaging of single atom catalysts. An important development is the use of reducible supports, such as CeO2 or Fe2O3 where the interaction between the metal and support is playing a crucial role.

Vol. 7: Inorganic Electrochemistry; Keith J. Stevenson, Evgeny V. Antipov and Artem M. Abakumov This volume bridges several fields across chemistry, physics and material science. Perhaps this topic is best associated with the book “Inorganic Electrochemistry: Theory, Practice and Applications” by Piero Zanello that was intended to introduce inorganic chemists to electrochemical methods for study of primarily molecular systems, including metallocenes, organometallic and coordination complexes, metal complexes of redox active ligands, metal-carbonyl clusters, and proteins. The emphasis in this Volume of CIC III is on the impact of inorganic chemistry on the field of material science, which has opened the gateway for inorganic chemists to use more applied methods to the broad areas of electrochemical energy storage and conversion, electrocatalysis, electroanalysis, and electrosynthesis. In recognition of this decisive impact, the Nobel Prize in Chemistry of 2019 was awarded to John B. Goodenough, M. Stanley Whittingham, and Akira Yoshino for the development of the lithium-ion battery.

Vol. 8: Inorganic Photochemistry; Vivian W. W. Yam In this Volume the editor has compiled 19 chapters discussing recent developments in a variety of developments in the field. The introductory chapter overviews the several topics, including photoactivation and imaging reagents. The first chapters include a discussion of using luminescent coordination and organometallic compounds for organic light-emitting diodes (OLEDs) and applications to highlight the importance of developing future highly efficient luminescent transition metal compounds. The use of metal compounds in photo-induced bond activation and catalysis is highlighted by non-sacrificial photocatalysis and redox photocatalysis, which is another fundamental area of immense research interest and development. This work facilitates applications like biological probes, drug delivery and imaging reagents. Photochemical CO2 reduction and water oxidation catalysis has been addressed in several chapters. Use of such inorganic compounds in solar fuels and photocatalysis remains crucial for a sustainable environment. Finally, the photophysics and photochemistry of lanthanoid compounds is discussed, with their potential use of doped lanthanoids in luminescence imaging reagents.

xxii

Preface

Vol. 9: NMR of Inorganic Nuclei; David L. Bryce Nuclear magnetic resonance (NMR) spectroscopy has long been established as one of the most important analytical tools at the disposal of the experimental chemist. The isotope-specific nature of the technique can provide unparalleled insights into local structure and dynamics. As seen in the various contributions to this Volume, applications of NMR spectroscopy to inorganic systems span the gas phase, liquid phase, and solid state. The nature of the systems discussed covers a very wide range, including glasses, single-molecule magnets, energy storage materials, bioinorganic systems, nanoparticles, catalysts, and more. The focus is largely on isotopes other than 1H and 13C, although there are clearly many applications of NMR of these nuclides to the study of inorganic compounds and materials. The value of solid-state NMR in studying the large percentage of nuclides which are quadrupolar (spin I > ½) is apparent in the various contributions. This is perhaps to be expected given that rapid quadrupolar relaxation can often obfuscate the observation of these resonances in solution. Vol. 10: X-ray, Neutron and Electron Scattering Methods in Inorganic Chemistry; Angus P. Wilkinson and Paul R. Raithby In this Volume the editors start with an introduction on the recent history and improvements of the instrumentation, source technology and user accessibility of synchrotron and neutron facilities worldwide, and they explain how these techniques work. The modern facilities now allow inorganic chemists to carry out a wide variety of complex experiments, almost on a day-to-day basis, that were not possible in the recent past. Past editions of Comprehensive Inorganic Chemistry have included many examples of successful synchrotron or neutron studies, but the increased importance of such experiments to inorganic chemists motivated us to produce a separate volume in CIC III dedicated to the methodology developed and the results obtained. The introduction chapter is followed by 15 chapters describing the developments in the field. Several chapters are presented covering recent examples of state-of-the-art experiments and refer to some of the pioneering work leading to the current state of the science in this exciting area. The editors have recognized the importance of complementary techniques by including chapters on electron crystallography and synchrotron radiation sources. Chapters are present on applications of the techniques in e.g., spin-crossover materials and catalytic materials, and in the use of time-resolved studies on molecular materials. A chapter on the worldwide frequently used structure visualization of crystal structures, using PLATON/PLUTON, is also included. Finally, some more specialized studies, like Panoramic (in beam) studies of materials synthesis and high-pressure synthesis are present. Direct observation of transient species and chemical reactions in a pore observed by synchrotron radiation and X-ray transient absorption spectroscopies in the study of excited state structures, and ab initio structure solution using synchrotron powder diffraction, as well as local structure determination using total scattering data, are impossible and unthinkable without these modern diffraction techniques. Jan Reedijk, Leiden, The Netherlands Kenneth R. Poeppelmeier, Illinois, United States March 2023

6.01

Introduction: A short history of single site catalysis

Rutger A. van Santen and Emiel J.M. Hensen, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, MB Eindhoven, The Netherlands © 2023 Elsevier Ltd. All rights reserved.

6.01.1 6.01.2 6.01.2.1 6.01.2.2 6.01.2.3 6.01.2.3.1 6.01.2.3.2 6.01.2.4 6.01.2.4.1 6.01.2.4.2 6.01.2.5 6.01.2.5.1 6.01.2.5.2 6.01.2.6 References

Introduction Single site catalystsdThe modern age of heterogeneous catalysis Introduction The disproportionation or metathesis reaction, polymerization catalysis Selective oxidation Propylene epoxidation Other oxidation reactions catalyzed by zeolites Bifunctional catalysts; hydrocarbon activation Activation of short alkanes by Ga or Zn Methane to aromatics catalysis; the methane dehydro-aromatization reaction Single atom catalysis Single atoms; reducible supports Solid solution catalysts Summary; from solid state to molecular nano-clusters catalysts

1 2 2 2 4 4 7 8 9 10 11 11 12 13 14

Abstract This chapter provides an outline of the contents of the book. A sketch of the historic development of molecular heterogeneous catalysis is given as a context to the selection of collected chapters.

6.01.1

Introduction

This introductory chapter provides an introduction to the chapters of this book. It also contains a short summary of the historic development of molecular heterogeneous catalysis, which found its start in the 50s and 60s of the previous century. We have organized the respective chapters in sections: catalyst synthesis and functionalization, characterization and theory and catalyst/structure and catalyst reactivity relations. In contrast to the molecular heterogeneous catalysts, which consist of well-defined nanoclusters or cations immobilized on a support, the catalysts as invented in the early 20th century were complex reactive inorganic materials used in their bulk form. In time they became replaced by catalytically reactive particles distributed over an also high surface area support. Whereas initially apart from composition, pore size was the main physical property of concern, supported catalysts generate the need to characterize the reactive particles on the relevant nanometer level. This became possible in the 40s of the previous century. Chapter 6.16 by Webb and Filot, which deals with promoted Fischer-Tropsch catalysts, provides a review of these pre-1950 developments for synthesis gas conversion catalysis. It became soon discovered that differences in size and shape of the supported catalytically active particles can give large variation in catalytic reactivity for some reactions. This led to the important concept of structure sensitivity or insensitivity of catalytic reactivity.1Chapter 6.17 by Rogers and Freakley discusses structure sensitivity of well-defined alloy surfaces in the context of oxidation catalysis, while Chapter 6.19 by Watzele, Garlyyev, Gubanova and Bandarenka does the same for electrocatalytic reactions. Currently, differences in reactivity are understood on the basis of the atomic structure of the catalyst surface. It is experimentally based on the modern instrumental tools of nanomaterial as well as surface science.2 This went in parallel with paradigmatic changes of catalyst materials. In the middle of the previous century molecular heterogeneous catalysis began with the exploration of inorganic synthesis tools from coordination and organometallic chemistry, which came to fruition around that time. It led to the design of well-defined molecular clusters anchored to a high surface area support. True molecular definition became possible when also the support material and its porous structure could be chemically defined on a molecular level. Molecular heterogeneous catalysts with high surface area became available with the discovery of synthesis of zeolites, which have such a atomically defined nanostructure. The synthesis of zeolites is dealt with in contributions by Palcic and Valtchev (see chapter 6.02) and Yu and Xiao (see chapter 6.03). Zeolites by themselves have unique reactivity characteristics. The Chapter 6.04 written by Copéret and Korzy nski introduces the surface organometallic approach to catalyst synthesis and related catalytic chemistry. The incorporation of molecular complexes into the zeolite nanopore and a comparison of such systems

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00152-7

1

2

Introduction: A short history of single site catalysis

with non-zeolite systems is discussed in Chapter 6.05 by Kurto glu-Öztulum and Uzun and in Chapter 6.06 by Li and Pidko. The characterization and exploration of zeolites activated by nanoclusters is a subject in several other chapters in this book, which we will highlight in the following sections. In the course of time, it became realized that the nature of the support materials is not only important to provide anchoring sites of the molecule complexes, but could also have a synergetic interaction. A rich portfolio of atomically defined catalysts became discovered with reducible supports as CeO2 of Fe2O3. Chapter 6.10 by Porter and Datye on Imaging of single atom catalysts and Chapter 6.11 on Metal-support interfaces in ceria-based catalysts by Muravev, Kosinov and Hensen provide an introduction to such systems and their characterization. Molecular heterogeneous catalysis is not possible without the sophisticated characterization arsenal and computational tools that are currently available. Chapter 6.07 on Single Site Spectroscopy of Transition Metal Ions and Reactive Oxygen Complexes in Zeolites written by Plessers, Bols, Rhoda, Heyer, Solomon, Sels and Schoonheydt and Chapter 6.06 on Metal containing nanoclusters in zeolites by Li and Pidko discuss various types of spectroscopy for characterization of nanoclusters embedded in the micropores of zeolites. Computational advances that deal with the dynamical changes of the catalytic site affected by reactions are presented in Chapter 6.08 on Dynamic Evolution of catalytic active sites within zeolite catalysis by Bocus, Neale, Cnudde and Van Speybroeck and Chapter 6.09 on Nano Cluster Heterogeneous Catalysts: Insights from Theory by Sun and Sautet. In the following sections we will sketch the history of successfully applied catalyst materials and corresponding catalysis, which leads up to the current state of the art of molecular heterogeneous catalysis. A rich variety of single atom or single site catalysts will be met and, with it, also new reactivity concepts that often refer to the state of the reacting site. The chapter will be concluded with a short summarizing section that also will refer to the relation between molecular properties of the solid-state material and supported nanoclusters.

6.01.2

Single site catalystsdThe modern age of heterogeneous catalysis

6.01.2.1

Introduction

Molecular inorganic chemistry became to fruition in the 60s of the previous century. Coordination chemistry and metal organic chemistry became well established and homogenous catalysts were part of important catalytic processes. This chemistry became also widely explored in heterogeneous catalysis and has led to a wealth of new processes and new catalytic systems. Well-defined molecular complexes are anchored to a solid material of high surface area. These reaction centers are physically separated and react independently. The reactive complex can consist of a single cation that is part of a molecular complex or a nanocluster built from several cations. It has led to a host of new discoveries of catalysts that we will discuss in the next sections. We will focus on the relation between the chemistry of the catalytic reactive site and mechanism of the reactions catalyzed. In the early 50s, major discoveries were made in polymerization and metathesis catalysis, that we will discuss in the next section. The other important class of reactions with major advances are selective oxidation reactions as propylene epoxidation, which are presented in Section 6.01.2.3. Also in the early 50s the discovery of bifunctional catalysts for hydrocarbon activation had a major impact. This will be discussed in Section 6.01.2.4, where not only hydrocarbon activation of solid acid zeolite catalysts by nanometal clusters will be met, but also hydrocarbon activation by oxidic nanoclusters occluded in zeolite pores. Section 6.01.2.5 deals with single atom catalysts dissolved in reducible oxide supports.

6.01.2.2

The disproportionation or metathesis reaction, polymerization catalysis

One of the first heterogeneous single site catalysts is the silica-supported chromate catalyst discovered by Banks and Hogan in 1953 at the Phillips Petroleum Company in Oklahoma.3 This catalyst was successfully applied to the low-pressure polymerization of ethylene. Around the same time the serendipitous discovery of the TiCl4/(Al(C2H5)2Cl) or a-TiCl3/(Al(C2H5)2Cl) catalytic systems was made by Ziegler and coworkers4,5 at the Institute of Kohlenforschung in Mülheim, Germany.6–8 Ziegler and Natta were awarded the Nobel Prize in Chemistry for this work in 1963. Besides low-pressure polymerization of ethylene, it led to the development of the important polymerization of propylene to stereoregular isotactic polypropylene. These polymerization catalysts are single site catalysts where the reaction center is a reducible cation, for both Cr and Ti.9–11 Fig. 1 gives the structure and proposed mechanism of the Cr-catalyzed ethylene polymerization reaction. The Cr3þ ion is connected through two SieOeCr bonds to the silica surface and its free ligand positions are occupied by the adsorbed ethylene molecule which inserts into the growing hydrocarbon alkyl chain. Interestingly, 10 years after the discovery of this polymerization catalyst again at the Phillips Petroleum Company another single site catalyst was discovered again by Banks jointly with Bailey.12 It catalyzed a remarkable new reaction: the disproportionation reaction of propylene to ethylene and butene. This catalyst was prepared by treating silica with Mo-carbonyl followed by calcination The deprotonation reaction became later also named the metathesis reaction, when it became an important reaction of new organometallic chemistry. New homogeneous catalysts based on the metathesis reaction became discovered for ring-opening catalysis of cyclic olefins (ROMP). W, Ta or Re are metals useful as part of metathesis catalysts. Their contribution to metathesis catalysis led to the Nobel Prize in Chemistry in 2005 of Chauvin, Grubbs and Schrock.13 The metathesis reaction and related reactions catalyzed by

Introduction: A short history of single site catalysis

3

Fig. 1 The catalytic reaction cycle of the chromate Phillips ethylene polymerization catalyst. Reproduced with permission from Brown, C.; Lita, A.; Tao, Y.; Peek, N.; Crosswhite, M.; Mileham, M.; Krzystek, J.; Achey, R.; Fu, R.; Bindra, J. K.; Polinski, M.; Wang, Y.; van de Burgt, L. J.; Jeffcoat, D.; Salvatore Profeta, J.; Stiegman, A. E.; Scott, S. L. Mechanism of Initiation in the Phillips Ethylene Polymerization Catalyst: Ethylene Activation by Cr(II) and the Structure of the Resulting Active Site. ACS Catal. 2017, 7, 7442–7455. doi: 10.1021/ACSCATAL.7B02677.

heterogeneous catalysts are discussed in Chapter 6.4 that deals with surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts. The Ziegler-Natta TiCl3/MgCl2 catalyst as well as the Banks-Bailey disproportionation catalyst are also the first catalytic systems that became probed by early quantum-chemical models and calculations. Both are coordination complexes containing single metal centers. Ligand field theory had then recently developed,14,15 which could describe the relation between the electronic structure of the transition metal cations and the relative stability of the coordination complex. Cossee proposed in 196416 the insertion mechanism of ethylene and propylene polymerization with an empirical ligand field quantum-chemical model. When ab-initio first principle quantum-chemical methods became developed, the first calculations applied to catalysis by Clementi et al.,17 then at IBM, were undertaken to probe the Cossee semi-empirical quantum-chemical model. The Cossee reactions mechanism can be extended to also explain stereoregular propylene polymerization18 The key intermediate of the Cossee-Arlman olefin polymerization mechanism is shown in Fig. 2. The chain-growth reaction consists of the insertion of the ethylene or propylene molecule coordinated to the Ti4þ ion into the growing alkyl polymer chain (which had formerly a negative charge). Similar to Fig. 1, the reaction is initiated by addition of ethylene or propylene to a hydrogen atom to give adsorbed alkyl intermediate. The hydrogen atom is regenerated in a chain-growth termination reaction. Density function theory (DFT) quantumchemical calculations that support the Cossee-Arlman mechanism are from Ziegler et al. in 2003.19 In 1971, Herisson and Chauvin from the French Institute de Petrole proposed the carbene mechanism of the metathesis reaction based on detailed investigations of product patterns of a variety of alkene mixtures. A statistical redistribution of carbon atoms was

Fig. 2 The surface complex of titanium chloride which illustrates the insertion of the propylene molecule inserting into the alkyl polymerization chain according to the Cossee–Arlman mechanism of propylene polymerization. After Arlman, E. J.; Cossee, P. Ziegler-Natta Catalysis III. Stereospecific Polymerization of Propene With the Catalyst System TiCl3AlEt3. J. Catal. 1964, 3, 99–104. doi: 10.1016/0021-9517(64)90097-1.

4

Introduction: A short history of single site catalysis

observed that allowed them to deduce the now generally accepted Chauvin metathesis (Chauvin, Nobel Prize in Chemistry 2005) reaction mechanism.20 According to the Chauvin mechanism, the metathesis reaction proceeds through formation of an intermediate metal-carbene species. Upon contact with an alkene molecule, a metallocycle intermediate forms. Cleavage of the weakened C]C bond of reacting alkene gives a new carbene intermediate. The product molecule forms through a new C]C bond with the carbene ligand originally present. The reaction then resumes. Fig. 3b shows a calculated structure of Mo]CH2 attached to a SiO2 surface. DFT calculations by Handzlik and Sautet21 predict such and related Mo]CH2 site species as reaction intermediates. The reactivity for disproportionation catalysis is found to be a strong function of the structure of the silica surface. Also the Chauvin mechanism provides for a route to polymerization. When cyclic alkenes react, their C]C bonds will open and the atoms become available for new C]C bond formation with other molecules. This the ring-opening monomer polymerization (ROMP) reaction. It is useful to summarize the difference between the Cossee mechanism of alkene polymerization and the Chauvin mechanism of metathesis catalysis. The Cossee mechanism is based on an alkyl substituent (ligand), which grows through insertion of the olefin. The alkene oligomerization reaction is catalyzed by a transition metal of the first row in the periodic system of the transition metals, which are cationic, nearly or completely ionized and in a low valence state. In the Chauvin mechanism, the reactive substituent is attached to the transition metal is a carbene species (CH2) or its analog when the hydrogen atoms are substituted by alkyl or other groups. This reaction is catalyzed by a transition metal cation of the second and third row of the transition metals in a high valence state with near redox states.

6.01.2.3

Selective oxidation

Here we will discuss discoveries of selective oxidation reactions with catalysts that have been designed respectively as single atom catalytic sites attached to a support surface, as solid solutions in an inorganic matrix or as oxidic nanoclusters in the nanopores of a zeolitic material. In the next section we will discuss three prototypes. The Ti/SiO2 catalyst used to epoxidize propylene with an hydroperoxide reagent is an example of a single atom site attached to the support surface. Ti atoms dissolved in the framework of a siliceous zeolite for epoxidation of propylene with hydrogen peroxide is a prototype example of a solid solution. The solid solution approach with reactive cations dissolved in zeolitic frameworks has generated a host of new catalytic systems which will be discussed as well. An early catalyst with its reaction center occluded in zeolite nanopores is Fe promoted zeolite for the hydroxylation of benzene. The chemistry of nanoclusters occluded in zeolites has become widely explored and is discussed in several chapters of this book (see chapters 6.5, 6.6, 6.10, 6.16).

6.01.2.3.1

Propylene epoxidation

The quest to replace the commonly used epichlorohydrine process for epoxidation of propylene, which is a stoichiometric reaction propylene with Ca(OH)2 and HOCl co-producing environmentally harmful chloride, by a catalytic process without the use of chlorides has resulted into important successful homogeneous and heterogeneous catalytic processes. Catalysis of the heterogenous type, both based on the unique reactivity Ti cationic sites will be discussed here.

(A)

(B)

Fig. 3 (a) The Chauvin reaction mechanism of the metathesis reaction. (b) The four center metallocycle intermediates in ethene metathesis. Panel (a): After Chauvin, Y. Olefin Metathesis: The Early Days (Nobel Lecture). Angew. Chem. Int. Ed. Engl. 2006, 45, 3740–3747. doi: 10.1002/ anie.200601234. Panel (b): Reproduced with permission from Handzlik, J.; Sautet, P. Active Sites of Olefin Metathesis on Molybdena-Alumina System: A Periodic DFT Study. J. Catal. 2008, 256, 1–14. doi: 10.1016/J.JCAT.2008.02.016.

Introduction: A short history of single site catalysis

5

Whereas epoxidation of ethylene oxygen is an established process catalyzed by Ag,22 the epoxidation reaction of propylene when catalyzed by the Ag catalyst of ethylene epoxidation only occurs with low yield.23 This lower selectivity of propylene epoxidation is because, in contrast to ethylene, that readily gives intermediate allyl prone to total oxidation. It had already been discovered in the 60s that through a consecutive reaction with intermediate allyl propylene oxidation catalyzed by Cu gives acrolein instead of propylene oxide.23,24 Selective epoxidation of propylene is possible when instead of molecular oxygen a peroxide is used as reactant. Important to what follows is that in water the epoxide readily hydrolyses to the corresponding glycol. Inspired by this principle, two heterogeneous catalytic processes became discovered for the selective oxidation of propylene-to-propylene oxide. One process uses a hydroperoxide organic molecule in aprotic medium, the other uses hydrogen peroxide but selects a catalyst with its reactive site localized in an apolar environment. Both processes are based on a Ti cation as the active site. When the hydroperoxide is used the insertion of the oxygen atom in a desired product, the epoxide co-produces an alcohol, while with hydrogen peroxide water is the co-product. This involves homogeneous catalytic epoxidation by complexes containing Mo6þ, W6þ, V5þ or Ti4þ25 that became applied industrially around 1970. The Shell SMPO (Styrene Monomer Propylene Oxide) process26,27 became operational in 1979. It employs a single site catalyst consisting of a Ti complex immobilized on silica. This catalyst is used to epoxidize propylene with ethylbenzene hydroperoxide and produces the epoxide and ethylbenzene alcohol. In the overall process, the latter is dehydrated and converted into styrene monomer. Instead of ethylbenzene also isobutane can be used to generate, by autocatalytic reaction, the corresponding hydroperoxide. Then in the overall process isobutene is co-produced, which is also a useful product.28 The Ti catalyst is prepared by reaction of a Ti(alkoxy)4 complex with a silica surface. Upon calcination the Ti complex becomes anchored through two or three SieOeTi bonds to the silica surface. This reaction of an organometallic complex with silica is an early example of the by now well-developed surface organometallic chemistry approach for the preparation well defined molecule complexes attached to oxidic surfaces (see chapters 6.4, 6.5).29–31 Around the same time as the heterogenous epoxidation processes, enantiomeric epoxidation of alkenes catalyzed by an organometallic Ti complex using chiral tartrate ligands were discovered by the American chemist Sharpless (Nobel Prize in Chemistry 2001).32–34 The mechanism of this homogenous catalytic reaction is very similar to that deduced from EXAFS and DFT calculations on a Ti cation grafted to the surface of a porous silica catalyst.35

Fig. 4 Schematic representation of propylene epoxidation by hydroperoxide with Ti/SiO2 catalyst in which the proton changes position between peroxide atoms. After Catlow, C. R.; French, S.; Sokol, A.; Thomas, J. Computational Approaches to the Determination of Active Site Structures and Reaction Mechanisms in Heterogeneous Catalysts. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 2005 363, 913–936. doi: 10.1098/RSTA.2004.1529.

6

Introduction: A short history of single site catalysis

This is schematically illustrated in Fig. 4. Ti4þ accommodates the ROO peroxy species. The oxygen atom attached to Ti ion is transferred to the epoxide. After reaction, the alcoholate remains adsorbed to Ti, which reacts with another ROOH molecules to initiate the next reaction cycle. Its proton reacts with the alcoholate and hydrogen-bridges with the Ti attached ROOH. Because of its bond to Ti and the hydrogen promotion, the oxygen atom that inserts into the alkene p bond is highly electrophilic. A few years after the Shell SMPO process, the epoxidation reaction of propylene with H2O2 became discovered in 1983 at the Enichem company in Italy.36 It is used in the Dow-BASF epoxidation process since 2008. The catalyst of this process, named TS1 (titanosilicalite-1) by the inventors, was originally thought to be a single atom site of Ti. The Ti4þ cation substitutes for Si in the framework of a synthetic nanoporous material called silicalite-1.37,38 It is a low concentration solid solution of Ti ions in the silica framework of a polymorph of quartz with the topology of MFI zeolite. The siliceous environment of the Ti site makes the reactive site is aprotic. It repulses water and hence prevents hydrolysis propylene oxide to glycol. The invention of the TS1 propylene epoxidation catalyst illustrates the reinforcement of scientific progress by co-discoveries in seemingly unrelated fields. For the TS1 catalyst to be discovered, previous scientific advances in zeolite synthesis had to occur that made silicalite-1 available. Silicalite-1 is one of the compounds that became synthesized after the 1967–1968 discoveries at the Mobil Oil company of the synthesis of high-silica framework zeolites.39 Zeolites are nanoporous alumino-silicate materials with important applications in the chemical and petrochemical industry. The framework of the zeolite is a network of tetrahedra connected through their apices. The nanopores are part of the crystalline structure and created by the network topology of the zeolite framework, that has high variability. An introduction to zeolite materials, their synthesis and use can be found in Chapter 6.02 by Palcic and Valtchev (see Fig. 1 in that chapter 6.02). When a Si4þ ion is replaced by an Al3þ ion the zeolite framework is no longer electrostatically neutral. In the material, this can be neutralized by an additional proton which will be located on the oxygen atom that bridges the Si4þ and Al3þ cation containing tetrahedra. Protonated zeolites react as solid acids that we will shortly discuss in Section 6.01.2.4 (see chapters 6.13 and 6.14). Substitution of Si4þ by Ti4þ maintains charge neutrality of the framework. It changes the siliceous zeolite material to a catalysts useful to selective oxidation as discussed above. It has recently been discovered that the catalytic site is not a single Ti atom but is actually is a site where two Ti atoms have a synergetic interaction. This has been shown by Gordon et al. using labeling studies with 18 O2 in combination with advanced spectroscopic studies and DFT simulations.40 As is schematically illustrated in Fig. 5, the reactive site consists of two Ti atoms that are located in two adjacent tetrahedra. When reacted with hydrogen peroxide, the single oxygen atom bridge becomes replaced by two OOH hydroperoxide substituents bridge the two Ti cations. The coordination of the Ti atoms becomes five-coordinated. Reaction with propylene is analogous to the epoxidation mechanism of grafted TI shown in Fig. 4. The hydrogen atom of the peroxide forms a hydrogen bond with an oxygen atom next to Ti of the zeolite framework. Similar as for the single atom model site, there is also an organometallic dimer complex, the di-m-oxotitanium(salene) molecule that reacts analogously to the dual site Ti epoxidation mechanism.41 The concentration of Ti in the material is very low. It cannot be excluded that in parallel also single atom Ti sites are reactive. As we will see also later, for nanoclusters in the zeolite nanopore several differently structured sites can be active in parallel. The

Fig. 5 Reaction scheme of propylene epoxidation by dual Ti silicalite site. After Gordon, C. P.; Engler, H.; Tragl, A. S.; Plodinec, M.; Lunkenbein, T.; Berkessel, A.; Teles, J. H.; Parvulescu, A.-N.; Copéret, C. Efficient Epoxidation over Dinuclear Sites in Titanium Silicalite-1. Nature 2020, 586, 708– 713. doi: 10.1038/s41586-020-2826-3.

Introduction: A short history of single site catalysis

7

question whether a single atom site or nanoclusters with two or more metal centers is the selective site is a general key question to be addressed in many related systems. The epoxidation reaction of propylene with hydrogen peroxide or hydroperoxides can be described as a reaction catalyzed by Ti4þ that acts as a Lewis acidic site. The valency of the Ti cation does not change. It provides an anchor site to the intermediate hydroperoxide substituents. It promotes the electrophilic insertion of the oxygen atom into the alkene p bond. Cavity size and composition of the zeolite affects selectivity reaction importantly. In the particular case of the epoxidation reaction the hydrophobicity of the reaction site is essential.

6.01.2.3.2

Other oxidation reactions catalyzed by zeolites

Inspired by the siliceous TS1 catalyst, substitution of the zeolite lattice Si by Lewis acid cations other than Ti has also been explored. Siliceous zeolitic frameworks with Si4þ substituted by Sn4þ or Zr4þ in the framework of zeolite b, which has larger nanopore dimensions than TS1, have been investigated for the conversion of biomass derived molecules. In the first decade of this century, catalysts were discovered for the selective oxidation of biomass derived molecules as lactones or monosaccharides.42–44 Important to the further development of zeolitic oxidation catalysts is the discovery of nanoporous zeolitic structures not based on the SiO2 tetrahedral network but on networks of the AlPO4 composition. In 1982, Wilson45 succeeded in the synthesis of such nanoporous materials with nanopores useful for catalysis. By substitution of P5þ by Si4þ the AlPO4-based materials can be converted into solid acids. Al3þ can also be substituted by cations of the same charge or cations with lower valency. An extensive portfolio of zeolitic systems with a variety of compositions became available through the inventions by Wilson and Flanigan. The Si-AlPO4, SAPO-34, catalyst discovered in 1986 by Union Carbide researchers is currently the preferred catalyst for the methanol-to-olefin (MTO) process currently used at large scale in China to convert coal to olefins.46 The rich variety in composition and structure of the AlPO4 systems makes them attractive to exploration in selective oxidation and related reactions. Around the beginning of this century a host of such single site heterogeneous catalysts of use to fine chemical production and biomass molecule conversion became developed and investigated by Thomas.31 An example are the Co/Mn AlPO18 or FeAlPO-31 materials designed in the 90s for the selective oxidation of alkanes by molecular oxygen. The reaction with oxygen usually is initiated by radical autocatalysis. Reaction of the intermediate peroxides is influenced by redox and Lewis acid catalysis with lattice cations47 Important to selectivity is the quenching of the radical reaction chain by the nanopores of the material narrow pores. Radical chain quenching by selection of cavity size constraint is a general mechanism to suppress total oxidation as has been recently also demonstrated for the selective oxidation of methane.48 Also since the 90s of the previous century, nanoporous metal-organic-framework (MOF) materials are explored for catalysis, mainly for fine chemicals synthesis. The organic framework is held together by cations. This building principle has made available a large portfolio of different structures and compositions with possibilities of detailed molecular design. They are however much less robust than zeolites. This so far has inhibited catalytic breakthrough applications.49 On the other hand, MOF materials have promising properties as adsorbents for CO2 capture,50 which is a desirable technology within the context of climate change. So far, we discussed zeolitic materials with the reactive cation substituted for one of the lattice cations. A very important additional class of zeolite catalysts has reactive cations or nanocluster not located in the zeolite framework but located in their nanopores (i.e., outside the framework itself). One of the first zeolite catalysts with extra frame work catalytic reaction centers that has been explored for selective oxidation is the Panov reaction. This reaction, which we discuss in the next section, catalyzes hydroxylation of benzene to phenol with N2O. The Fe-promoted zeolite catalyst of this reaction was found to consist of single or dual Fe atom molecular complexes located in the nanopore of the zeolite. After discussing this catalyst, we will highlight in Section 6.01.2.3.2.2 other important reactions catalyzed by nanoclusters in zeolite nanopore. 6.01.2.3.2.1 The Panov oxidation reaction The Panov reaction, which oxidizes benzene with N2O to phenol, is catalyzed by a low Fe content ZSM-5 zeolite catalyst. This catalyst has the same MFI zeolite structure as silicalite-1 discussed in the previous section.51 This reaction is attractive, since it is a possible outlet for N2O, which is an environmentally undesirable greenhouse gas. N2O is a stoichiometric co-product of nylon manufacture.52 So far there is no alternative to the current nylon producing process. The N2O reaction with benzene to produce phenol has been discovered originally by Japanese chemists in 1983.53 In 1988 the zeolite ZSM-5 was found to have high activity. Subsequent research led in 2000 to the announcement by Solutia (formerly Monsanto) of commercial implementation. The catalysts used by Solutia had been developed jointly with the Panov group from the Boreskov Institute of Catalysis in Novosibirsk, Russia.54 This happened after the discovery that the low impurity level of Fe in the ZSM-5 catalysts is responsible for catalyst activity and not the presence of zeolitic protons as had initially be proposed. The Fe3þ ion is present as a substitute for Al3þ in the framework of the zeolite. During reaction, it leaches from the framework. Extensive characterization of the system by many research groups demonstrated that a distribution of singe atom or multiple Fe cation containing clusters is in the nanopores of the zeolite catalyst. It became soon clear that there is only one uniquely reactive Fe site, which activates oxygen to selectively insert into the benzene CH bond.55,56 The selective catalytic site of the Panov catalyst is most likely a single extra-framework Fe2þ ion stabilized by the negative charge around the zeolite framework Al3þ containing tetrahedra. The Fe2þ ion converts into FeO2þ when reacted with N2O. The selective elementary reaction step that produces phenol happens by insertion of O into the CH bond of benzene adsorbed to the cationic complex. The current generally accepted mechanism of the reaction is illustrated in Fig. 6.

8

Introduction: A short history of single site catalysis

Fig. 6 The catalytic cycle for benzene to phenol oxidation with N2O over extraframework iron (FeEF) sites. With permission from Li, G.; Pidko, E. A.; Van Santen, R. A.; Feng, Z.; Li, C.; Hensen, E. J. M. Stability and Reactivity of Active Sites for Direct Benzene Oxidation to Phenol in Fe/ZSM-5: A Comprehensive Periodic DFT Study. J. Catal. 2011, 284, 194–206. doi: 10.1016/j.jcat.2011.07.008.

Catalysis by a single cation is selective. Dual or higher oxycationic clusters are non-selective. In these cationic clusters, the oxygen atom readily accepts the phenol proton. This elementary reaction step promotes deactivating phenol oligomerization reactions. 6.01.2.3.2.2 Oxidation catalysis with nanoclusters The discovery also in other catalyst systems that single or multinuclear metal cationic clusters occluded into the micropores of zeolites have unique reactivity has led to exploration of a variety of synthesis approaches of such systems. Coordination complex synthesis, activation of metal organic compounds as well as classical impregnation methods are used57,58 (see chapters 6.5, 6.6, and 6.10). Chapter 6.07 and Chapter 6.06 by Plessers et al. and Li and Pidko provide reviews of the characterization and theoretical study of metal cations and cationic clusters in zeolites. Zeolites activated with cationic reaction centers in their nanopores have been widely explored, which has resulted in important applications. Here we will mention oxidation-related reactions. In Section 6.01.2.4, we will discuss dehydrogenation reactions. An industrial application of metal/zeolite is in automotive catalysts, where Cu ions in chabazite catalyze NOx from the aerobic diesel exhaust by reaction with added ammonia (typically obtained by decomposition of added urea). Zeolite catalysis for this application was discovered König and co-workers around 1985 at Volkswagen, who then used a Cu-exchanged mordenite catalyst.59,60 Methane oxidation to methanol is a highly desirable reaction. It would replace current methanol synthesis processes from synthesis gas. Analogous to the benzene to phenol reaction discussed in previous section, the methane to methanol reaction is an insertion reaction of an oxygen atom. Direct oxidation with O2 to methanol is a low selectivity reaction, because activation of methanol or other products as ethylene is significantly more facile than methane.61,62 In Chapters 6.6 and 6.10, Fe-containing zeolites are discussed for selective oxidation of methane to methanol. Not only Fecontaining catalysts, but also Cu-containing zeolite catalysts can form methanol. Such reaction has to be done in successive cycles of catalyst oxidation and reduction by methane with a low reaction yield. Different from the benzene hydroxylation reaction, the methane to methanol reaction is a radical reaction. Recently, Snyder et al.48 found that the difference in yield when catalysis by single Fe cations is compared in different zeolite frameworks relates to the mobility of intermediate CH3 radicals. In the narrow-pore chabazite framework, a substantially higher yield is observed than in the wide pore zeolite Beta framework. The narrow pore windows of the chabazite cavity prevent escape of the methyl radical from the reaction cavity. Deactivation happens by escaping methyl radicals that form deactivating methoxy species with non-reacted oxidations. Hutchings and co-workers63 found that the reaction can be executed in a single step in water with hydrogen peroxide. Again the reaction proceeds through a radical reaction. Using isotope labeling, they demonstrated that the oxygen atom of methanol can also derive from water. This implies that intermediate OH radicals are part of the reaction mechanism. For more details on this reaction, we refer to Chapters 6.6 and 6.10.

6.01.2.4

Bifunctional catalysts; hydrocarbon activation

The discovery in the 40s of the previous century by Haensel at UOP of the bifunctional gasoline reforming catalyst that converts alkanes into aromatics not only gave catalysis an important new catalyst concept, but it also became of immense importance to

Introduction: A short history of single site catalysis

9

the petroleum understory.64 In 1949 the first platforming process that converts naphtha into high octane gasoline became operational. The breakthrough discovery of Haensel was the discovery that a noble metal as Pt distributed with extremely low concentration distributed over an acidified support (in his case chlorinated alumina) can be used for his reaction. The catalyst is an early example of a nanomaterial because the Pt is dispersed as nanosized particles. The reforming process is operated at high hydrogen pressure. The Pt nanoparticle serves to activate hydrogen and to cleave and form CH bonds of alkanes. Because of the alkane–alkene equilibrium at reaction conditions, the steady-state concentration of alkene is extremely low. These alkenes are converted to aromatics by consecutive reactions catalyzed by the acidic protons of the support. The combination of transition metal activation and consecutive reaction by protons makes the catalytic system bifunctional. A low steady-state concentration of alkene is important to suppress acid-catalyzed deactivating carbonaceous molecule formation (see Chapter 6.14). Also the Pt nanoparticle itself can become deactivated. Alloying of Pt, for instance with Re,65 is used to prevent such deactivation. The large interest, industrially as well as academically in alloy catalysis, finds its origin in this deactivation question.66,67 Alloy catalysis is discussed in Chapter 6.17 by Rogers and Freakley in relation with oxidation catalysis and Chapter 6.19 by Watzele et al. in relations with electrocatalysis. Since then, also the structure, distribution and Pt particle size has been extensively investigated. Chapter 6.09 on Nanocluster Heterogeneous Catalysts: Insights from theory by Sun and Sautet provides a state of the art summary of the modeling of these particles. A salient feature is the sensitivity of their structure to catalyst support and their ease of reconstruction when exposed to reactants. Another important discovery with major impact on industrial catalysis is the discovery of strong acidity of zeolites68 by Rabo and colleagues at Union Carbide. In the early days of zeolite catalysis discovery, they were working with high Al/Si systems. In 1957, the discovery was made that ion exchange with Ca2þ leads to stable and highly reactive solid acid catalysts. The catalytic cracking process became invented at Mobil Oil in 1964,69 which at high temperature cracks heavy oil into gasoline grade automotive fuel is based on zeolite Y catalysts from Union Carbide. It is currently a major catalytic process in the petrochemical industry. The chapters by Rutger van Santen give an introduction to zeolite Brønsted acidity and solid acid catalysis. A relevant question is whether the metal particle is located in the zeolite micropore or externally at the surface of the zeolite binder material, used to provide strength and shape to the catalyst particle. Kinetic studies show that usually there is no equilibrium between external composition and reactant and product concentration in the nanopore70 Pt particles should reside near the protons residing in the nanopores. Selective deposition of the metal particle exploits cation exchange with cations located in the zeolite nanopore. Hence to deposit Pt particles inside the zeolite nanopore Pt complexes that have a positive charge should be used. This in contrast to the negatively charges Pt chloride complexes conventionally used to impregnate the surfaces of alumina catalysts that when impregnated carry a positive charge.71 It as has recently been demonstrated that, by using Pt(NH3)42þ to deposit Pt72,73 the nanoparticle can be selectively positioned in the zeolite nanopore. In the Chapter 6.06 by Li and Pidko, alternative ways to occlude metal particles in the zeolite nanocavity by encapsulation are discussed. The interplay of transition metal clusters with proton acidity is reviewed in Ref. 74 by Gates, now at University of Davis, who started in the early 90s an important research program on the chemistry and catalysis of nanoclusters in zeolites that continues up to today.75,76 Early research that reminds of the successful disproportionation catalyst preparation of Banks and Bailey (see Section 6.01.2.2), designed supported clusters represented as Ir4, Ir6, Rh4, and Rh6 in the supercages of faujasite zeolites by ship-in-a-bottle synthesis of the metal carbonyl cluster precursors [Ir4(CO)12], [Ir6(CO)16], [Rh4(CO)12], and [Rh6(CO)16], respectively, which were subsequently decarbonylated. These studies indicate also significant differences in reactivity of such very small nanoclusters. Chapter 6.6 highlights such and more recent nanocluster chemistry. Not only transition metals as Pt can be used in bifunctional catalytic reactions that activate alkanes. Also zeolites activated by Lewis acid cations or Zn or Ga are useful hydrocarbon activation catalysts. As discussed in the next section this property is exploited in reactions where hydrogen cannot be used. Such processes are the oligomerization of short alkanes as ethane or propane, available form natural gas to more heavy molecules as aromatics. In Section 6.01.2.4.2 related non-oxidative activation of methane activation is introduced, which requires activation by more reactive cations similar as disproportionation catalysis.

6.01.2.4.1

Activation of short alkanes by Ga or Zn

HZSM-5 zeolite promoted with Ga is a catalyst that selectively converts short alkanes into aromatics and hydrogen. This reaction is endothermic and requires a temperature of 500  C, which has been discovered in the 70s of the previous century by researchers at the British Petroleum company.77 The catalytic reaction is fundamental to the BP-UOP Cyclar process, announced in 1991 to be commercialized. Light alkanes are an attractive feedstock available from ample resources of natural gas that in the 70s became available. The process would make the world less dependable on crude oil. Ga-oxide clusters were thought to catalyze selectively dehydrogenation of alkanes. Subsequent reaction catalyzed by the zeolitic protons converts intermediate olefins into desirable aromatic molecules. However, because hydrogen cannot be used, the catalyst deactivates readily with coke formation. Catalyst recycle through high-temperature coke removal is part of the overall process.

10

Introduction: A short history of single site catalysis

Whereas noble metal as Pt can also be used for this reaction, in the absence of hydrogen the Pt nanoparticle rapidly deactivates due to non-selective CeC bond cleavage reactions.78 Since its original invention, substantial additional research has been done and related processes have been proposed. Other elements than Ga can also be used. Only Zn has a comparable reactivity as Ga.79 Chapter 6.10 reviews some of this chemistry. This catalytic system is of general interest because as previously the key mechanistic questions is whether the dehydrogenation site is the single Ga or Zn cation or is this site a dual od higher metal oxide cation site. Early research of the 90s resolved the question whether the Ga3þ ion that can substitute for the Al3þ cation in the aluminosilicate matrix of the zeolite framework is active when part of the zeolite lattice or is only active when located in the zeolite microchannel as a cationic complex.80 Ga or Zn are only active when present as isolated ions or oxycationic clusters in the zeolite nanopore.81 In the course of time the relation between the structure of the cationic site and catalytic reactivity has been elucidated through computational and spectroscopic studies in combination with model catalytic experiments.82–87 Dimetal oxycationic complexes have superior activity with respect to the activation of the CeH bonds of the alkanes. To regenerate the complex hydrogen has to desorb. Due to the strong interaction with the oxycation, its rate of desorption can be slow. The preferred catalytic reaction site is a single cation stabilized by the negative charge of the zeolite lattice. In the dimetal oxycation, CH activation is a heterolytic dissociation reaction. The H atom of the CH bond binds with the extraframework O atom and the alkyl fragment to the Ga3þ or Zn2þ ion part of the oxycationic complex. On a single atom Gaþ 1 or Zn2þ ion site, the alkyl intermediate attaches to the Zn2þ or Gaþ 1 ion and the hydrogen atom adsorbs as a proton on the zeolite framework O atom. The alkene molecule desorbs through a b-CH cleavage reaction of the alkyl intermediate. The alkene molecule desorbs and hydride anion and proton recombine to give hydrogen. Because of the high barrier of hydrogen desorption from the Zn dimetal oxycationic complex, the Zn2þ ion is the preferred catalytic site. In the case of Ga, the dimetal oxycation reaction site is the more reactive one. Alkane activation and hydrogen desorption have comparable activation energies.88 However, the catalyst deactivates readily because the rate of water desorption competes with that of hydrogen molecule desorption.85 Therefore, also in the case of the Ga catalysts at steady state the catalyst is a single site catalyst and Ga is reactive as a Gaþ 1 ion. Once the alkene molecules are formed, they is converted by the zeolite protons by subsequent oligomerization and hydride transfer reactions to aromatics. Ga is the preferred H ZSM-5 promotor, since Zn is too volatile to be sustained through the catalyst recycle process. The catalyst belongs to the category of single atom reaction site catalysts.

6.01.2.4.2

Methane to aromatics catalysis; the methane dehydro-aromatization reaction

As for oxidative conversion of methane the total oxidation of product molecules has to be abated, the challenge in non-oxidative methane conversion is to suppress deactivation by coke formation. The reaction than converts methane into aromatics is called methane dehydroaromatization (MDH). There are two catalytic systems that show promise. One system originates from the 80s of the previous century, when Russian scientists.89 demonstrated the possibility of the reaction of methane to aromatics catalyzed by metal-promoted ZSM-5 catalysts. The reaction is highly endothermic and requires a temperature of 970 K. A major advance was the discovery by Wang et al.90 that Mo promoted H-ZSM-5 shows attractive performance. Chapter 6.11 by Muravev, Kosinov and Hensen provides an introduction to this system and current state of the art. Additional reviews can be found in Refs. 91,92. After the following short introduction of the second system, that is non-zeolitic, we will continue by highlighting molecular mechanistic aspects of the zeolitic system. A second and later development is the discovery by Bao’s group.93 They did not use a zeolitic system, but a uniquely prepared single site Fe/SiO2 system. The temperature of reaction used is 1400 K. At this temperature, the aromatization reaction is mainly a gas-phase radical reaction. The main role of Fe is to initiate the radical reaction. Due to its single site nature, the Fe site will not deactivate by coking and catalyst deactivation relates to oil and coke formation intrinsic to the radical reaction.94,95 This reaction and catalytic system is highlighted in Chapter 6.17. A prime question on the zeolitic systems is the question of the nuclearity of the cationic nanocluster that activates methane. For HZSM-5 promoted by Fe, which is also active for the methane aromatization96,97 as well as Mo-HZSM-5, this has been investigated. The single site cation is part of catalysts that are the most stable. Contact with methane will reduce the single metal containing oxocationic cluster. Then a carbiding process sets in, which can remain a single metal atom site,97 but according to some other authors leads to a multinuclear carbided active centers.98 Selective carbided cationic sites need to remain located in the zeolite nanopore. Interestingly, also the Fe/SiO2 site of the high-temperature methane conversion reaction referred to earlier is transformed into a species coordinated to two carbon atoms and the Si atom.93 The original interpretation of the catalyst mechanism is that it is a bifunctional catalyst. The metal cation activates methane and produces an alkene intermediate that in consecutive reactions is converted by the zeolite proton to aromatics, that thermodynamically next to carbon can be formed as product. There is consensus that methane is activated by the cation. Fig. 7 schematically illustrates a mechanistic reaction path supported by DFT calculations toward formation of the ethylene intermediate from methane. The key reaction intermediate is a CH2 carbene species attached to the metal cation. Methane can react with CH2 to give a 2 methyl intermediates. Alternatively it has also been suggested that analogous to heterolytic dissociation methane reacts with the metal cation and nanocluster carbon98 to give a CH and CH3 intermediate. Subsequent hydrogen transfer would give two carbene intermediates. The group of Hensen99,100 demonstrated that silicalite-1, which does not contain protons, promoted by Mo also catalyzes the methane aromatization reaction. This is in conflict with the bifunctional mechanistic model of the methane to aromatics reaction. They suggest that a radical reaction leads to intermediate occluded polyaromatics. By using 13C-labeled methane, they show that the

Introduction: A short history of single site catalysis

11

Fig. 7 Reaction pathway for the entire catalytic cycle of methane dehydrogenation and coupling to ethylene over a single Mo cation. With permission from Zhou, D.; Zuo, S.; Xing, S. Methane Dehydrogenation and Coupling to Ethylene Over a Mo/HZSM-5 Catalyst: A Density Functional Theory Study. J. Phys. Chem. C 2012, 116, 4060–4070. doi: 10.1021/JP209037C.

reaction proceeds through an intermediate hydrocarbon pool, analogous to the catalytic cyclopentadienyl or benzyl cations suggested for the acid catalyzed methanol to aromatics reactions.101,102

6.01.2.5 6.01.2.5.1

Single atom catalysis Single atoms; reducible supports

Catalysts with reactive cations on reducible supports as CeO2 have unique reactivity. Their reactive sites are often single atom cations. The unique reactivity derives from reactive synergy with the reducible support atoms. The chapters by Datye et al. and Maravev et al. of this book describe frontier research of such reducible support catalysts. Interest in these catalysts started with their use in automotive catalysis. We will shortly outline these developments and then highlight mechanistic aspects of the catalytic reactions. Gasoline en diesel engine exhaust gas treatment require different technologies since their respective exhaust pipe conditions are very different. The Cu zeolite systems shortly discussed in Section 6.01.2.3.2.2 earlier refer to the treatment technology used in diesel engines, where the NOx removal is done by reaction with NH3 in oxidative environment. The reaction conditions for exhaust treatment of gasoline engines is regulated with sensors and operates near oxidation-reduction stoichiometry neutral condition. The catalyst consists of precious metals such as Pt, Pd and Rh distributed on a CeO2/ZrO2 reducible oxide and alumina dispersed on a ceramic cordierite honeycomb structure. Pd mainly functions as oxidation catalyst for CO and hydrocarbons, while Rh is an active NOx reduction catalyst with hydrogen, CO or hydrocarbon. Pt can serve both functions. The reducible mixed CeO2/ZrO2 support oxide is used to balance the oxygen stoichiometry. It stores oxygen when there is excess oxygen and provides oxygen at reducing conditions. It took some time before it became recognized that there is also a unique catalytic role of CeO2. Gasoline exhaust catalysts became widely implemented in automotive vehicles after 1981. They could be used due to the introduction of unleaded fuel, mandatory since 1975. The other important gasoline grade condition is low sulfur level, which also became available due to tightening governmental legislations since the 1980s.103 One aspect of the superior performance of the ceria-supported catalysts is the dissolution of noble metal atoms in the CeO2 matrix.104 Recent, high-resolution transmission electron microscopy has visualized these mono-atomically dispersed atoms.105,106 The metal atoms substitute as cations for the Ce cations. Charge imbalance introduced by the difference in charge of the cations is accommodated through oxygen vacancy formation in the oxide. The uniqueness of the CeO2 support relates to its low energy of surface oxygen atom vacancy formation.107 The surface vacancy oxygen vacancy formation energy is of the order of 200 kJ/mol for CeO2.108 This is to be compared with surface oxygen vacancy formation energies of materials used as oxidation selective oxidation catalysts as MoO3 or V2O5 that are of the order of 600 kJ/mol.109 CeO2 and also other reducible supports as Fe2O3 or TiO2 dissolve readily catalytically reactive transition metal cations. One of the most spectacular reactivity aspects of these systems is their extreme reactivity. At room temperature, they are able to oxidize CO or catalyze the water-gas shift reaction of CO with H2O that gives CO2 and H2. Because of this property, the catalyst is of use in preferential oxidation (PROX) where CO is selectively is oxidized in the presence of hydrogen.110,111 This reaction is useful to fuel cell operation, where the PROX reaction prevents CO poisoning of the Pt electrode. The reasons for this superior reactivity are twofold. Whereas CO adsorbs strongly to reduced transition metal clusters or atoms, this interaction energy is substantially weaker for CO adsorption to a cation (representative numbers are 180 kJ/mol for CO adsorption to Rh surface vs 80 kJ/mol for CO adsorbed to the Rh3þ cation of Rh2O3). This reduces the temperature of CO poisoning from 220  C on a reduced transition metal to only 50  C for the cation. It is for instance the reason for the high reactivity at low temperature of the CO oxidation reaction catalyzed by small Rh2O3 particles dispersed on CeO2 support.112 Adsorbed CO will readily react with O from the CeO2 support. Fig. 8a illustrates the mechanism of CO oxidation as modeled quantum-chemically on a Pd/CeO2

12

Introduction: A short history of single site catalysis

Fig. 8 (a) Helmholtz free energy analysis of CO oxidation on a Pd1O/CeO298 surface model (energies in kJ/mol) are shown for T ¼ 50  C. The corresponding reaction intermediate structures are shown (color scheme: red, surface O; white, surface Ce; pink, O of adsorbed species; blue, Pd). (b) Catalytic cycle derived for the 2 NO þ 2 CO / N2 þ 2CO2 reaction on Pd1/CeO2.108 Panel (a): With permission from Spezzati, G.; Benavidez, A. D.; DeLaRiva, A. T.; Su, Y.; Hofmann, J. P.; Asahina, S.; Olivier, E. J.; Neethling, J. H.; Miller, J. T.; Datye, A. K.; Hensen, E. J. M. CO Oxidation by Pd Supported on CeO2(100) and CeO2(111) Facets. Appl. Catal. Environ. 2019, 243, 36–46. doi: 10.1016/J.APCATB.2018.10.015. Panel (b): With permission from Ding, W.-C.; Gu, X.-K.; Su, H.-Y.; Li, W.-X. Single Pd Atom Embedded in CeO2(111) for NO Reduction With CO: A First-Principles Study. J. Phys. Chem. C 2014, 118, 12216–12223. doi: 10.1021/JP503745C.

catalyst. CO adsorbs to the Pd ion to which an oxygen is attached. It then reacts with O of the CeO2 surface to CO2. The surface vacancy is regenerated by dissociation of O2.113 The second important feature is the presence of oxygen atom vacancies of the CeO2 support. The presence of an oxygen vacancy cannot only assist O2 dissociation, but also for instance the dissociation of NO necessary in the reaction of CO with NO that gives N2 and CO2.114–116 The mechanism of the NO reduction reaction with CO to give N2 is shown in Fig. 8b. The latter illustrates the mechanism where two NO molecules initially recombine on Pd to give the N2O2 intermediate. Subsequently the N2O2 decomposes to N2 by cleavage of the respective NeO bonds and adsorption of its two oxygen atoms on surface vacant sites of CeO2. The surface vacancies are regenerated by successive reaction with CO that gives CO2. An additional important feature of the single atom catalysts supported by reducible supports as CeO2 or Fe2O3 that it increases catalyst stability. The early discovery by Japanese scientists in 1987 of low temperature CO oxidation by catalysts with Au ions dispersed on Fe2O3 or other reducible supports cannot only be considered a milestone because of the discovery that Au can be part of an active heterogeneous catalysis, but also since it is fundamentally the discovery of the activity of single atom/reducible support systems.117 A few years later Bond et al.118 discovered that Au present at extreme low concentration on a non-reducible alumina support is active as alkene hydrogenation catalyst. Au catalysis has since then widely been explored and has shown promising activity for several important other reactions.119–122 It is also an essential component of catalysts that produce hydrogen peroxide from hydrogen and oxygen. Au catalysis has also been explored for selective conversion of biomass derived molecules. This catalysis is discussed in Chapter 6.17 on Selective Oxidation by Mixed Metal Nanoparticles by Rogers and Freakley.

6.01.2.5.2

Solid solution catalysts

Single atom cations dissolved in non-reducible, rigid supports as MgO and CaO have also been explored since the 60s of the previous century by Stone and colleagues from the University of Bath.123,124 These materials are well-defined solid solutions. Deviations from stoichiometry are extremely small. These oxide solid solutions strongly resist deviations of charge or coordination for solute ions. For these reasons at the time research on these materials started were also considered as interesting model systems to study the relation between catalyst reactivity and reactive cation electronic structure. A major conclusion is that local coordination, cation electron count and electron spin state (see also Ref. 125) dominate differences in reactivity of different cations. Recent quantum-chemical calculations and electrochemical studies have confirmed this view.126 Solution occurs extensively when ions isovalent to the mother matrix cations are used (i.e., TM2 þ in MgO, TM3 þ in La2O3, etc.) provided that certain requirements are met, notably size and preferred coordination. The reactivity of these systems compared to those with reducible oxides is quite low. For instance, different from solid solution in reducible oxides, a solid solutions of Co2þ in MgO will only oxidize CO at 170  C.127 Solid solutions of complex oxides as the versatile perovskites or scheelites have important applications. Chapter 6.18 Doped Semiconductor Photocatalysts by Hisatomi and Domen discusses applications of

Introduction: A short history of single site catalysis

13

perovskite oxides substituted with cations of different charge in photocatalysis. Perovskites are also important as high temperature catalysts of the oxygen evolution reaction.126

6.01.2.6

Summary; from solid state to molecular nano-clusters catalysts

The past decades the science of heterogeneous catalysis has integrated the molecular sciences, which has become the major lens through which the chemistry of the catalyst surface is interpreted. This is a position far from the origin of heterogeneous catalysis in the first part of previous century. The relation between catalyst performance and material properties of the catalyst remained long ill understood. This, too a large extent, was due to the lack of knowledge of the structure of the inorganic solids and their exposed surfaces. Instrumental techniques to study catalysis at the nanolevel still had to be discovered. The chapters of this book in heterogeneous catalysis science built on the chemistry that became introduced in the 50s and 60s of the previous century. Then, molecular inorganic chemistry made its impact and heterogeneous catalysts became introduced designed using the then new coordination chemistry and organometallic chemistry. A host of new reactions and processes originate from that time. The novelty and unexpected nature of these reactions result often from serendipitous discoveries. The TiCl3/MgCl2 catalyst is an example of this. The discovery of the Fe-promoted Panov reaction is another example. Also, the discovery that zeolites can be converted into a highly active solid acid catalysts came as a surprise.68 The new inorganic molecular synthesis capabilities made a large impact. This went together with the advanced state of materials science with characterization tools that allow their characterization not only at nanoscale, but even at the atomic level. Such spectroscopic techniques became complemented with powerful computational simulation tools. The latter became only available as a reliable computational tool to simulate catalysts at the beginning of this century. Surface science, developed also in the second part of the previous century, contributes largely to current theories of heterogeneous catalysis, due to the introduction of model catalyst systems that are well defined at atomistic level. Catalytic science is about the relationship of catalyst functionality and its structure and composition, defined at the atomistic and nanoscale level. This relationship is described by reaction kinetics. In order to model kinetics the mechanism of the reaction has to be formulated. This is the network of elementary reaction steps that composes the catalytic reaction cycle. The major advance of today is that elementary reaction rate constants can be deduced from the chemistry that occurs at the catalyst surface. Macroscopic catalyst performance depends in a complex way on surface dynamic events. Catalytic kinetics is multiscale. Longer time and spatial events are deduced from phenomena of shorter molecular time and length scales. Microkinetics predicts the state of the catalyst surface, which in turn determines macroscopic global kinetics.128 The surface state ultimately defines the relation between catalyst performance and catalyst composition. The surface state of the catalyst strongly depends on reaction conditions. Also the inorganic chemistry of the catalyst in its active catalytic reactive state is intrinsically complex. Not only, as mentioned, will the state of the catalytic site depend on the reaction and the reaction conditions, also the structure and even composition of the site may change during the ongoing reaction. For the surfaces, this phenomenon is recognized as surface reconstruction, while for nanoclusters, even at steady state, it may cause the reactive catalytic site to have only an indirect relation with the state of the catalyst before reaction. The synthetic challenge to incorporate this self-organization into the design of the catalyst reflects into the increasing attention for in operando study of the catalyst and computational interest to capture this dynamics of the catalytic system. A major challenge to current catalysis science is to predict how this state varies with reaction conditions. The chapters that are part of this book have been written with a focus on the synthesis and characterization of catalytic complexity and its relation with the molecular chemistry of the catalytic reaction. Catalysis also develops as a response to the needs for new reactions as time evolves. At the start of molecular heterogeneous catalysis in the middle of the previous century after the Second World War the petrochemical industry was rapidly expanding. New processes became invented for the oil industry now used at large scale in refineries. New materials as polymers or detergents became manufactured from oil derived products through the new catalytic inventions. With the oil crises of the 70s of the previous century as described in some of the chapters a renewed interest started for processes based in synthesis gas from coal or based on natural gas as raw material. With the huge parallel expansion of chemical industry in the 90s of the previous century increasing environmental concerns made a large impact on the direction of catalytic exploration. Catalysts, as the zeolitic oxidation catalysts, became investigated to replace harmful stoichiometric organic reactions by catalytic environmentally friendly processes and automotive exhaust catalysis became an important new branch of catalysis science. At present due to the perceived need to reduce CO2 emissions, the search for alternative processes based on renewable resources as biomass or electricity from solar or windmill electricity directs significantly catalytic innovation. It implies a renewed interest in electrocatalysis and solar energy conversion. These developments define the catalytic reactions and respective catalytic systems that are addressed in in this book. The chapters in this book have been written with a focus on the synthesis, characterization and modeling of catalytic complexity and its relation with the molecular chemistry of the catalytic reaction.

14

Introduction: A short history of single site catalysis

References 1. Boudart, M. Catalysis by Supported Metals. Adv. Catal. 1969, 20 (C), 153–166. https://doi.org/10.1016/S0360-0564(08)60271-0. 2. Li, Y.; Somorjai, G. A. Introduction to Surface Chemistry and Catalysis, 2nd ed.| Wiley, Wiley, 2010. 3. Hogan, J. P. Ethylene Polymerization Catalysis over Chromium Oxide. J. Polym. Sci., Part A-1: Polym. Chem. 1970, 8 (9), 2637–2652. https://doi.org/10.1002/ POL.1970.150080929. 4. Ziegler, K.; Holzkamp, E.; Breil, H.; Martin, H. Polymerisation von Äthylen und anderen Olefinen. Angew. Chem. 1955, 67 (16), 426. https://doi.org/10.1002/ ANGE.19550671610. 5. Ziegler, K.; Holzkamp, E.; Breil, H.; Martin, H. Das M{ü}lheimer Normaldruck-Poly{ä}thylen-Verfahren. Angew. Chem. 1955, 67 (19–20), 541–547. https://doi.org/10.1002/ ange.19550671902. 6. Natta, G. Une nouvelle classe de polymeres d’a-olefines ayant une régularité de structure exceptionnelle. J. Polym. Sci. A Polym. Chem. 1955, 16 (82), 143–154. https:// doi.org/10.1002/POL.1955.120168205. 7. Natta, G.; Pino, P.; Corradini, P.; Danusso, F.; Mantica, E.; Mazzanti, G.; Moraglio, G. Crystalline High Polymers of a-Olefins. J. Am. Chem. Soc. 2002, 77 (6), 1708–1710. https://doi.org/10.1021/JA01611A109. 8. Natta, G.; Pino, P.; Mazzanti, G. Sintesi e struttura di alcuni poli-idrocarburi cristallini contenenti atomi di carbonio asimmetrici nella catena principale. Chim. Ind. 1955, 37, 927. 9. Brown, C.; Lita, A.; Tao, Y.; Peek, N.; Crosswhite, M.; Mileham, M.; Krzystek, J.; Achey, R.; Fu, R.; Bindra, J. K.; Polinski, M.; Wang, Y.; van de Burgt, L. J.; Jeffcoat, D.; Salvatore Profeta, J.; Stiegman, A. E.; Scott, S. L. Mechanism of Initiation in the Phillips Ethylene Polymerization Catalyst: Ethylene Activation by Cr(II) and the Structure of the Resulting Active Site. ACS Catal. 2017, 7 (11), 7442–7455. https://doi.org/10.1021/ACSCATAL.7B02677. 10. Eley, D. D.; Rochester, C. H.; Scurrell, M. S. The Polymerization of Ethylene on Chromium Oxide Catalysts. I. Kinetic Studies. Proc. R. Soc. Lond. A Math. Phys. Sci. 1972, 329 (1579), 361–373. Accessed: Sep. 02, 2021. [Online]. Available: https://www.jstor.org/stable/78238?refreqid¼excelsior3A546b1974ea19b71285016bafd7e51c18 &seq¼1#metadata_info_tab_contents. 11. Cossee, P. Ziegler-Natta Catalysis I. Mechanism of Polymerization of {$}Alpha;-Olefins With Ziegler-Natta Catalysts*1. J. Catal. 1964, 3 (1), 80–88. https://doi.org/10.1016/ 0021-9517(64)90095-8. 12. Banks, R. L.; Bailey, G. C. Olefin Disproportionation. A New Catalytic Process. Ind. Eng. Chem. Prod. Res. Dev. 2002, 3 (3), 170–173. https://doi.org/10.1021/I360011A002. 13. Casey, C. P. 2005 Nobel Prize in Chemistry. Development of the Olefin Metathesis Method in Organic Synthesis. J. Chem. Educ. 2006, 83 (2), 192–195. https://doi.org/ 10.1021/ED083P192. 14. Chatt, J.; Duncanson, L. A. 586. Olefin Co-Ordination Compounds. Part III. Infra-red Spectra and Structure: Attempted Preparation of Acetylene Complexes. J. Chem. Soc. 1953, (0), 2939–2947. https://doi.org/10.1039/JR9530002939. 15. Ballhausen, B. J. Introduction to Ligand Field Theory, McGraw-Hill, 1962. 16. Cossee, P. Ziegler-Natta Catalysis I. Mechanism of Polymerization of a-Olefins With Ziegler-Natta Catalysts. J. Catal. 1964, 3 (1), 80–88. https://doi.org/10.1016/00219517(64)90095-8. 17. Novaro, O.; Blaisten-Barojas, E.; Clementi, E.; Giunchi, G.; Ruiz-Vizcaya, M. E. Theoretical Study on a Reaction Pathway of Ziegler–Natta-Type Catalysis. J. Chem. Phys. 1978, 68 (5), 2337–2351. https://doi.org/10.1063/1.436004. 18. Arlman, E. J.; Cossee, P. Ziegler-Natta Catalysis III. Stereospecific Polymerization of Propene With the Catalyst System TiCl3AlEt3. J. Catal. 1964, 3 (1), 99–104. https:// doi.org/10.1016/0021-9517(64)90097-1. 19. Seth, M.; Ziegler, T. Polymerization Properties of a Heterogeneous Ziegler Natta Catalyst Modified by a Base: A Theoretical Study. Macromolecules 2003, 36 (17), 6613– 6623. https://doi.org/10.1021/MA0301247. 20. Jean-Louis Hérisson, P.; Chauvin, Y. Catalyse de transformation des ol{é}fines par les complexes du tungst{è}ne. II. T{é}lom{é}risation des ol{é}fines cycliques en pr{é}sence d’ol{é}fines acycliques. Makromol. Chem. 1971, 141 (1), 161–176. https://doi.org/10.1002/macp.1971.021410112. 21. Handzlik, J.; Sautet, P. Active Sites of Olefin Metathesis on Molybdena-Alumina System: A Periodic DFT Study. J. Catal. 2008, 256 (1), 1–14. https://doi.org/10.1016/ J.JCAT.2008.02.016. 22. Voge, H. H.; Adams, C. R. Catalytic Oxidation of Olefins. Adv. Catal. 1967, 17 (C), 151–221. https://doi.org/10.1016/S0360-0564(08)60687-2. 23. Akimoto, M.; Ichikawa, K.; Echigoya, E. Kinetic and Adsorption Studies on Vapor-Phase Catalytic Oxidation of Olefins over Silver. J. Catal. 1982, 76 (2), 333–344. https:// doi.org/10.1016/0021-9517(82)90264-0. 24. Sachtler, W. M. H. Mechanism of the Catalytic Oxidations of Propene to Acrolein and of Isomeric Butenes to Butadiene. Recl. Trav. Chim. Pays-Bas 2010, 82 (3), 243–245. https://doi.org/10.1002/recl.19630820305. 25. Landau, R.; Sullivan, G. A.; Brown, D. Propylene Oxide by the Co-Product Processes. Chem. Tech. 1979, 9, 602. 26. Buijink, J. K. F.; Lange, J. P.; Bos, A. N. R.; Horton, A. D.; Niele, F. G. M. Propylene Epoxidation via Shell’s SMPO Process: 30 Years of Research and Operation. In Mechanisms in Homogeneous and Heterogeneous Epoxidation Catalysis, Elsevier, 2008; pp 355–371. 27. Buijink, J. K. F.; Van Vlaanderen, J. J. M.; Crocker, M.; Niele, F. G. M. Propylene Epoxidation over Titanium-on-Silica CatalystdThe Heart of the SMPO Process. Catal. Today 2004, 93–95, 199–204. https://doi.org/10.1016/J.CATTOD.2004.06.041. 28. Oxidation products ethylene. In Industrial Organic Chemistry; Ted Oyama, S., Ed., Elsevier, 2008; pp 145–192. 29. Copéret, C.; Chabanas, M.; Romain, P. S.-A.; Basset, J.-M. Homogeneous and Heterogeneous Catalysis: Bridging the Gap through Surface Organometallic Chemistry. Angew. Chem. Int. Ed. Engl. 2003, 42 (2), 156–181. https://doi.org/10.1002/ANIE.200390072. 30. Guzman, J.; Gates, B. C. Supported Molecular Catalysts: Metal Complexes and Clusters on Oxides and Zeolites. Dalton Trans. 2003, 3 (17), 3303–3318. https://doi.org/ 10.1039/B303285J. 31. Thomas, J. M. Design and Applications of Single-Site Heterogeneous Catalysts: Contributions to Green Chemistry, Clean Technology and Sustainability, Imperial College Press, 2012. 32. Finn, M. G.; Sharpless, K. B. Mechanism of Asymmetric Epoxidation. 2. Catalyst Structure. J. Am. Chem. Soc. 2002, 113 (1), 113–126. https://doi.org/10.1021/ JA00001A019. 33. Katsuki, T.; Sharpless, K. B. The First Practical Method for Asymmetric Epoxidation. J. Am. Chem. Soc. 2002, 102 (18), 5974–5976. https://doi.org/10.1021/JA00538A077. 34. Jacobsen, E. N. Transition Metal-Catalyzed Oxidations: Asymmetric Epoxidation. In Comprehensive Organometallic Chemistry II; 1995; pp 1097–1135. https://doi.org/ 10.1016/B978-008046519-7.00137-4. 35. Catlow, C. R.; French, S.; Sokol, A.; Thomas, J. Computational Approaches to the Determination of Active Site Structures and Reaction Mechanisms in Heterogeneous Catalysts. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 2005, 363 (1829), 913–936. https://doi.org/10.1098/RSTA.2004.1529. 36. Clerici, M. G.; Bellussi, G.; Romano, U. Synthesis of Propylene Oxide From Propylene and Hydrogen Peroxide Catalyzed by Titanium Silicalite. J. Catal. 1991, 129 (1), 159– 167. https://doi.org/10.1016/0021-9517(91)90019-Z. 37. Bordiga, S.; Coluccia, S.; Lamberti, C.; Marchese, L.; Zecchina, A.; Boscherini, F.; Buffa, F.; Genoni, F.; Leofanti, G. XAFS Study of Ti-Silicalite: Structure of Framework Ti(IV) in the Presence and Absence of Reactive Molecules (H2O, NH3) and Comparison With Ultraviolet-Visible and IR Results. J. Phys. Chem. 2002, 98 (15), 4125–4132. https:// doi.org/10.1021/J100066A036.

Introduction: A short history of single site catalysis

15

38. Bordiga, S.; Bonino, F.; Damin, A.; Lamberti, C. Reactivity of Ti(IV) Species Hosted in TS-1 Towards H2O2–H2O Solutions Investigated by ab initio Cluster and Periodic Approaches Combined With Experimental XANES and EXAFS Data: A Review and New Highlights. Phys. Chem. Chem. Phys. 2007, 9 (35), 4854–4878. https://doi.org/ 10.1039/B706637F. 39. Flanigen, E. M. Chapter 2 Zeolites and Molecular Sieves an Historical Perspective. Stud. Surf. Sci. Catal. 1991, 58 (C), 13–34. https://doi.org/10.1016/S0167-2991(08) 63599-5. 40. Gordon, C. P.; Engler, H.; Tragl, A. S.; Plodinec, M.; Lunkenbein, T.; Berkessel, A.; Teles, J. H.; Parvulescu, A.-N.; Copéret, C. Efficient Epoxidation over Dinuclear Sites in Titanium Silicalite-1. Nature 2020, 586 (7831), 708–713. https://doi.org/10.1038/s41586-020-2826-3. 41. Sawada, Y.; Matsumoto, K.; Katsuki, T. Titanium-Catalyzed Asymmetric Epoxidation of Non-Activated Olefins with Hydrogen Peroxide. Angew. Chem. Int. Ed. 2007, 46 (24), 4559–4561. https://doi.org/10.1002/ANIE.200700949. 42. Corma, A. Attempts to Fill the Gap Between Enzymatic, Homogeneous, and Heterogeneous Catalysis. Catal. Rev. Sci. Eng. 2004, 46 (3 {&} 4), 369–417. https://doi.org/ 10.1081/CR-200036732. 43. Holm, M. S.; Saravanamurugan, S.; Taarning, E. Conversion of Sugars to Lactic Acid Derivatives Using Heterogeneous Zeotype Catalysts. Science 2010, 328 (5978), 602– 605. https://doi.org/10.1126/science.1183990. 44. Moliner, M.; Román-Leshkov, Y.; Davis, M. E. Tin-Containing Zeolites Are Highly Active Catalysts for the Isomerization of Glucose in Water. Proc. Natl. Acad. Sci. U. S. A. 2010, 107 (14), 6164–6168. https://doi.org/10.1073/pnas.1002358107. 45. Wilson, S. T.; Lok, B. M.; Messina, C. A.; Cannan, T. R.; Flanigen, E. M. Aluminophosphate Molecular Sieves: A New Class of Microporous Crystalline Inorganic Solids. J. Am. Chem. Soc. 2002, 104 (4), 1146–1147. https://doi.org/10.1021/JA00368A062. 46. Tian, P.; Wei, Y.; Ye, M.; Liu, Z. Methanol to Olefins (MTO): From Fundamentals to Commercialization. ACS Catal. 2015, 5 (3), 1922–1938. https://doi.org/10.1021/ ACSCATAL.5B00007. 47. Gómez-Hortigüela, L.; Corà, F.; Sankar, G.; Zicovich-Wilson, C. M.; Catlow, C. R. A. Catalytic Reaction Mechanism of Mn-Doped Nanoporous Aluminophosphates for the Aerobic Oxidation of Hydrocarbons. Chem. A Eur. J. 2010, 16 (46), 13638–13645. https://doi.org/10.1002/CHEM.201001876. 48. Snyder, B. E. R.; Bols, M. L.; Rhoda, H. M.; Plessers, D.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. Cage Effects Control the Mechanism of Methane Hydroxylation in Zeolites. Science 2021, 373 (6552), 327–331. https://doi.org/10.1126/SCIENCE.ABD5803. 49. Lee, J.; Farha, O. K.; Roberts, J.; Scheidt, K. A.; Nguyen, S. T.; Hupp, J. T. Metal–Organic Framework Materials as Catalysts. Chem. Soc. Rev. 2009, 38 (5), 1450–1459. https://doi.org/10.1039/B807080F. 50. Sumida, K.; Rogow, D. L.; Mason, J. A.; McDonald, T. M.; Bloch, E. D.; Herm, Z. R.; Bae, T.-H.; Long, J. R. Carbon Dioxide Capture in Metal–Organic Frameworks. Chem. Rev. 2011, 112 (2), 724–781. https://doi.org/10.1021/CR2003272. 51. Panov, G. I. Advances in Oxidation Catalysis; Oxidation of Benzene to Phenol by Nutrous Oxide. CATTECH 2000, 4 (1), 18–31. https://doi.org/10.1023/A:1011991110517. 52. Thiemens, M. H.; Trogler, W. C. Nylon Production: An Unknown Source of Atmospheric Nitrous Oxide. Science 1991, 4996 (251), 932–934. https://doi.org/10.1126/ SCIENCE.251.4996.932. 53. Iwamoto, M.; Hirata, J.; Matsukami, K.; Kagawa, S. Catalytic Oxidation by Oxide Radical Ions. 1. One-Step Hydroxylation of Benzene to Phenol Over Group 5 and 6 Oxides Supported on Silica Gel. J. Phys. Chem. 2002, 87 (6), 903–905. https://doi.org/10.1021/J100229A001. 54. Uriarte, A. K.; Rodkin, M. A.; Gross, M. J.; Kharitonov, A. S.; Panov, G. I. Direct Hydroxylation of Benzene to Phenol by Nitrous Oxide. Stud. Surf. Sci. Catal. 1997, 110, 857– 864. https://doi.org/10.1016/S0167-2991(97)81048-8. 55. Jia, J.; Pillai, K. S.; Sachtler, W. M. H. One-Step Oxidation of Benzene to Phenol With Nitrous Oxide over Fe/MFI Catalysts. J. Catal. 2004, 221 (1), 119–126. https://doi.org/ 10.1016/J.JCAT.2003.08.010. 56. Hensen, E. J. M.; Zhu, Q.; Janssen, R. A. J.; Magusin, P. C. M. M.; Kooyman, P. J.; Van Santen, R. A. Selective Oxidation of Benzene to Phenol with Nitrous Oxide over MFI Zeolites: 1. On the Role of iron and Aluminum. J. Catal. 2005, 233 (1), 123–135. https://doi.org/10.1016/J.JCAT.2005.04.009. 57. Gates, B. C. Metal Clusters in Zeolites: Nearly Molecular Catalysts for Hydrocarbon Conversion. Stud. Surf. Sci. Catal. 1996, 100 (C), 49–63. https://doi.org/10.1016/S01672991(96)80007-3. 58. De Vos, D. E.; Dams, M.; Sels, B. F.; Jacobs, P. A. Ordered Mesoporous and Microporous Molecular Sieves Functionalized with Transition Metal Complexes as Catalysts for Selective Organic Transformations. Chem. Rev. 2002, 102 (10), 3615–3640. https://doi.org/10.1021/CR010368U. 59. Deka, U.; Lezcano-Gonzalez, I.; Weckhuysen, B. M.; Beale, A. M. Local Environment and Nature of Cu Active Sites in Zeolite-Based Catalysts for the Selective Catalytic Reduction of NOx. ACS Catal. 2013, 3 (3), 413–427. https://doi.org/10.1021/CS300794S. 60. Han, L.; Cai, S.; Gao, M.; Hasegawa, J.; Wang, P.; Zhang, J.; Shi, L.; Zhang, D. Selective Catalytic Reduction of NOx with NH3 by Using Novel Catalysts: State of the Art and Future Prospects. Chem. Rev. 2019, 119 (19), 10916–10976. https://doi.org/10.1021/ACS.CHEMREV.9B00202. 61. Lunsford, J. H. Catalytic Conversion of Methane to more Useful Chemicals and Fuels: A Challenge for the 21st Century. Catal. Today 2000, 63 (2–4), 165–174. https:// doi.org/10.1016/S0920-5861(00)00456-9. 62. Farrell, B. L.; Igenegbai, V. O.; Linic, S. A Viewpoint on Direct Methane Conversion to Ethane and Ethylene Using Oxidative Coupling on Solid Catalysts. ACS Catal. 2016, 6 (7), 4340–4346. https://doi.org/10.1021/ACSCATAL.6B01087. 63. Freakley, S. J.; Dimitratos, N.; Willock, D. J.; Taylor, S. H.; Kiely, C. J.; Hutchings, G. J. Methane Oxidation to Methanol in Water. Acc. Chem. Res. 2021, 54 (11), 2614– 2623. https://doi.org/10.1021/ACS.ACCOUNTS.1C00129. 64. Haensel, V.; Bloch, H. B. Duofunctional Platinum Catalysts in the Petroleum Industry. Platin. Met. Rev. 1964, 8 (1), 2–8. Accessed: Sep. 02, 2021. [Online]. Available: https:// www.technology.matthey.com/article/8/1/2-8/. 65. Xiao, J.; Puddephatt, R. J. Pt-re Clusters and Bimetallic Catalysts. Coord. Chem. Rev. 1995, 143 (C), 457–500. https://doi.org/10.1016/0010-8545(94)07008-8. 66. Sinfelt, J. H. Catalysis by Alloys and Bimetallic Clusters. Acc. Chem. Res. 1977, 10 (1), 15–20. https://doi.org/10.1021/ar50109a003. 67. Sachtler, W. M. H. Chemisorption Complexes on Alloy Surfaces. Catal. Rev. Sci. Eng. 2006, 14 (1), 193–210. https://doi.org/10.1080/03602457608073411. 68. Rabo, J. A.; Schoonover, M. W. Early Discoveries in Zeolite Chemistry and Catalysis at Union Carbide, and Follow-Up in Industrial Catalysis. Appl. Catal. A Gen. 2001, 222 (1– 2), 261–275. https://doi.org/10.1016/S0926-860X(01)00840-7. 69. Plank, C. J.; Rosinski, E. J.; Hawthorne, W. P. Acidic Crystalline Aluminosilicates. New Superactive, Superselective Cracking Catalysts. Ind. Eng. Chem. Prod. Res. Dev. 2002, 9 (3), 325–334. https://doi.org/10.1021/I360011A001. 70. Steijns, M.; Froment, G. F. Hydroisomerization and Hydrocracking. 3. Kinetic Analysis of Rate Data for N-Decane and n-Dodecane. Ind. Eng. Chem. Prod. Res. Dev. 2002, 20 (4), 660–668. https://doi.org/10.1021/I300004A014. 71. Sachtler, W. M. H.; Zhang, Z. Zeolite-Supported Transition Metal Catalysts. Adv. Catal. 1993, 39 (C), 129–220. https://doi.org/10.1016/S0360-0564(08)60578-7. 72. Kim, J.; Kim, W.; Seo, Y.; Kim, J. C.; Ryoo, R. N-Heptane Hydroisomerization Over Pt/MFI Zeolite Nanosheets: Effects of Zeolite Crystal Thickness and Platinum Location. J. Catal. 2013, 301, 187–197. https://doi.org/10.1016/J.JCAT.2013.02.015. 73. Oenema, J.; Hofmann, J. P.; Hensen, E. J. M.; Zecevic, J.; de Jong, K. P. Assessment of the Location of Pt Nanoparticles in Pt/Zeolite Y/g-Al2O3 Composite Catalysts. ChemCatChem 2020, 12 (2), 615–622. https://doi.org/10.1002/CCTC.201901617. 74. Sachtler, W. M. H. Metal Clusters in Zeolites: An Intriguing Class of Catalysts. Acc. Chem. Res. 2002, 26 (7), 383–387. https://doi.org/10.1021/AR00031A005. 75. Kawi, S.; Chang, J. R.; Gates, B. C. Characterization of NaY Zeolite-Encaged Tetrairidium Clusters by Infrared and X-ray Absorption Spectroscopies. J. Phys. Chem. 2002, 97 (41), 10599–10606. https://doi.org/10.1021/J100143A014. 76. Hegedus, L. Catalyst Design: Progress and Perspectives, Wiley: New York, 1987.

16

Introduction: A short history of single site catalysis

77. Biscardi, J. A.; Iglesia, E. Structure and Function of Metal Cations in Light Alkane Reactions Catalyzed by Modified H-ZSM5. Catal. Today 1996, 31 (3–4), 207–231. https:// doi.org/10.1016/S0920-5861(96)00028-4. 78. Doolan, P. C.; Pujado, P. R. No Title. Hydrocarb. Process. 1989, 72, 68–90. 79. Xiang, Y.; Wang, H.; Cheng, J.; Matsubu, J. Progress and Prospects in Catalytic Ethane Aromatization. Cat. Sci. Technol. 2018, 8 (6), 1500–1516. https://doi.org/10.1039/ C7CY01878A. 80. Bayense, C. R.; van der Pol, A. J. H. P.; van Hooff, J. H. C. Aromatization of Propane Over MFI-Gallosilicates. Appl. Catal. 1991, 72 (1), 81–98. https://doi.org/10.1016/ 0166-9834(91)85030-Y. 81. Getsoian, A.“. B.”.; Das, U.; Camacho-Bunquin, J.; Zhang, G.; Gallagher, J. R.; Hu, B.; Cheah, S.; Schaidle, J. A.; Ruddy, D. A.; Hensley, J. E.; Krause, T. R.; Curtiss, L. A.; Miller, J. T.; Hock, A. S. Organometallic Model Complexes Elucidate the Active Gallium Species in Alkane Dehydrogenation Catalysts Based on Ligand Effects in Ga K-edge XANES. Cat. Sci. Technol. 2016, 6 (16), 6339–6353. https://doi.org/10.1039/C6CY00698A. 82. Kazansky, V. B.; Pidko, E. A. Intensities of IR Stretching Bands as a Criterion of Polarization and Initial Chemical Activation of Adsorbed Molecules in Acid Catalysis. Ethane Adsorption and Dehydrogenation by Zinc Ions in ZnZSM-5 Zeolitey. J. Phys. Chem. B 2004, 109 (6), 2103–2108 https://doi.org/10.1021/JP049224K. 83. Kazansky, V. B.; Serykh, A. I. Unusual Localization of Zinc Cations in MFI Zeolites Modified by Different Ways of Preparation. Phys. Chem. Chem. Phys. 2004, 6 (13), 3760– 3764. https://doi.org/10.1039/B401853B. 84. Pidko, E. A.; van Santen, R. A. Activation of Light Alkanes Over Zinc Species Stabilized in ZSM-5 Zeolite: A Comprehensive DFT Study. J. Phys. Chem. C 2007, 111 (6), 2643–2655. https://doi.org/10.1021/JP065911V. 85. Hensen, E. J. M.; Pidko, E. A.; Rane, N.; Van Santen, R. A. Water-Promoted Hydrocarbon Activation Catalyzed by Binuclear Gallium Sites in ZSM-5 Zeolite. Angew. Chem. Int. Ed. 2007, 46 (38), 7273–7276. https://doi.org/10.1002/anie.200702463. 86. Gonzales, N. O.; Chakraborty, A. K.; Bell, A. T. A Density Functional Theory Study of Hydrogen Recombination and Hydrogen-Deuterium Exchange on Ga/H-ZSM-5. Top. Catal. 1999, 9 (3), 207–213. https://doi.org/10.1023/A:1019118823908. 87. Joshi, Y. V.; Thomson, K. T. High Ethane Dehydrogenation Activity of [GaH]2 þ Al Pair Sites in Ga/H-[Al]ZSM-5: A DFT Thermochemical Analysis of the Catalytic Sites Under Reaction Conditions. J. Catal. 2007, 246 (2), 249–265. https://doi.org/10.1016/J.JCAT.2006.11.032. 88. Pidko, E. A.; van Santen, R. A.; Hensen, E. J. M. Multinuclear Gallium-Oxide Cations in High-Silica Zeolites. Phys. Chem. Chem. Phys. 2009, 11 (16), 2893–2902. https:// doi.org/10.1039/B815943B. 89. Yakerson, V. I.; Vasina, T. V.; Lafer, L. I.; Sytnyk, V. P.; Dykh, G. L.; Mokhov, A. V.; Bragin, O. V.; Minachev, K. M. The Properties of Zinc and Gallium Containing PentasilsdThe Catalysts for the Aromatization of Lower Alkanes. Catal. Lett. 1989, 3 (4), 339–345. https://doi.org/10.1007/BF00766072. 90. Wang, L.; Tao, L.; Xie, M.; Xu, G.; Huang, J.; Xu, Y. Dehydrogenation and Aromatization of Methane under Non-oxidizing Conditions. Catal. Lett. 1993, 21 (1–2), 35–41. https://doi.org/10.1007/BF00767368. 91. Kosinov, N.; Hensen, E. J. M. Reactivity, Selectivity, and Stability of Zeolite-Based Catalysts for Methane Dehydroaromatization. Adv. Mater. 2020, 32 (44), 2002565. https:// doi.org/10.1002/ADMA.202002565. 92. Schwach, P.; Pan, X.; Bao, X. Direct Conversion of Methane to Value-Added Chemicals Over Heterogeneous Catalysts: Challenges and Prospects. Chem. Rev. 2017, 117 (13), 8497–8520. https://doi.org/10.1021/ACS.CHEMREV.6B00715. 93. Guo, X.; Fang, G.; Li, G.; Ma, H.; Fan, H.; Yu, L.; Ma, C.; Wu, X.; Deng, D.; Wei, M.; Tan, D.; Si, R.; Zhang, S.; Li, J.; Sun, L.; Tang, Z.; Pan, X.; Bao, X. Direct, Nonoxidative Conversion of Methane to Ethylene, Aromatics, and Hydrogen. Science 2014, 344 (6184), 616–619. https://doi.org/10.1126/SCIENCE.1253150. 94. Hao, J.; Schwach, P.; Fang, G.; Guo, X.; Zhang, H.; Shen, H.; Huang, X.; Eggart, D.; Pan, X.; Bao, X. Enhanced Methane Conversion to Olefins and Aromatics by H-Donor Molecules Under Nonoxidative Condition. ACS Catal. 2019, 9 (10), 9045–9050. https://doi.org/10.1021/ACSCATAL.9B01771. 95. Van Der Zwet, G. P.; Hendriks, P. A. J. M.; Van Santen, R. A. Pyrolysis of Methane and the Role of Surface Area. Catal. Today 1989, 4 (3–4), 365–369. https://doi.org/ 10.1016/0920-5861(89)85032-1. 96. Lai, Y.; Veser, G. The Nature of the Selective Species in Fe-HZSM-5 for Non-oxidative Methane Dehydroaromatization. Cat. Sci. Technol. 2016, 6 (14), 5440–5452. https:// doi.org/10.1039/C5CY02258D. 97. Kosinov, N.; Wijpkema, A. S. G.; Uslamin, E.; Rohling, R.; Coumans, F. J. A. G.; Mezari, B.; Parastaev, A.; Poryvaev, A. S.; Fedin, M. V.; Pidko, E. A.; Hensen, E. J. M. Confined Carbon Mediating Dehydroaromatization of Methane over Mo/ZSM-5. Angew. Chem. Int. Ed. 2018, 57 (4), 1016–1020. https://doi.org/10.1002/ANIE.201711098. 98. Gao, J.; Zheng, Y.; Jehng, J.-M.; Tang, Y.; Wachs, I. E.; Podkolzin, S. G. Identification of Molybdenum Oxide Nanostructures on Zeolites for Natural Gas Conversion. Science 2015, 348 (6235), 686–690. https://doi.org/10.1126/science.aaa7048. 99. Kosinov, N.; Coumans, F. J. A. G.; Uslamin, E. A.; Wijpkema, A. S. G.; Mezari, B.; Hensen, E. J. M. Methane Dehydroaromatization by Mo/HZSM-5: Mono- or Bifunctional Catalysis? ACS Catal. 2016, 7 (1), 520–529 https://doi.org/10.1021/ACSCATAL.6B02497. 100. Agote-Arán, M.; Fletcher, R. E.; Briceno, M.; Kroner, A. B.; Sazanovich, I. V.; Slater, B.; Rivas, M. E.; Smith, A. W. J.; Collier, P.; Lezcano-González, I.; Beale, A. M. Implications of the Molybdenum Coordination Environment in MFI Zeolites on Methane Dehydroaromatisation Performance. ChemCatChem 2020, 12 (1), 294–304. https:// doi.org/10.1002/CCTC.201901166. 101. Dahl, I. M.; Kolboe, S. On the Reaction Mechanism for Hydrocarbon Formation from Methanol over SAPO-34: 2. Isotopic Labeling Studies of the Co-Reaction of Propene and Methanol. J. Catal. 1996, 161 (1), 304–309. https://doi.org/10.1006/JCAT.1996.0188. 102. Arstad, B.; Nicholas, J. B.; Haw, J. F. Theoretical Study of the Methylbenzene Side-Chain Hydrocarbon Pool Mechanism in Methanol to Olefin Catalysis. J. Am. Chem. Soc. 2004, 126 (9), 2991–3001. https://doi.org/10.1021/JA035923J. 103. Tamaru, K.; Mills, G. A. Chapter 8 Catalysts for Control of Exhaust Emissions. Catal. Today 1994, 22 (2), 349–360. https://doi.org/10.1016/0920-5861(94)80109-6. 104. Flytzani-Stephanopoulos, M.; Gates, B. C. Atomically dispersed supported metal catalysts. Annu. Rev. Chem. Biomol. Eng. 2012, 3, 545–574. https://doi.org/10.1146/ ANNUREV-CHEMBIOENG-062011-080939. 105. Jones, J.; Xiong, H.; DeLaRiva, A. T.; Peterson, E. J.; Pham, H.; Challa, S. R.; Qi, G.; Oh, S.; Wiebenga, M. H.; Hernández, X. I. P.; Wang, Y.; Datye, A. K. Thermally Stable Single-Atom Platinum-on-Ceria Catalysts Via Atom Trapping. Science 2016, 353 (6295), 150–154. https://doi.org/10.1126/SCIENCE.AAF8800. 106. Nie, L.; Mei, D.; Xiong, H.; Peng, B.; Ren, Z.; Hernandez, X. I. P.; DeLaRiva, A.; Wang, M.; Engelhard, M. H.; Kovarik, L.; Datye, A. K.; Wang, Y. Activation of Surface Lattice Oxygen in Single-Atom Pt/CeO2 for Low-Temperature CO Oxidation. Science 2017, 358 (6369), 1419–1423. https://doi.org/10.1126/SCIENCE.AAO2109. 107. Trovarelli, A. Catalysis by Ceria and Related Materials; vol. 2; Published By Imperial College Press And Distributed By World Scientific Publishing Co, 2002. 108. Nolan, M.; Grigoleit, S.; Sayle, D. C.; Parker, S. C.; Watson, G. W. Density Functional Theory Studies of the Structure and Electronic Structure of Pure and Defective Low Index Surfaces of Ceria. Surf. Sci. 2005, 576 (1–3), 217–229. https://doi.org/10.1016/J.SUSC.2004.12.016. 109. Hermann, K. Physical and Chemical Properties of Oxygen at Vanadium and Molybdenum Oxide Surfaces: Theoretical Case Studies. In Computational Methods in Catalysis and Materials Science: An Introduction for Scientists and Engineers; van Santen, R. A., Sautet, P., Eds., Wiley-VCH, 2009; pp 375–415. 110. Korotkikh, O.; Farrauto, R. Selective Catalytic Oxidation of CO in H2: Fuel Cell Applications. Catal. Today 2000, 62 (2–3), 249–254. https://doi.org/10.1016/S0920-5861(00) 00426-0. 111. Liu, K.; Wang, A.; Zhang, T. Recent Advances in Preferential Oxidation of CO Reaction Over Platinum Group Metal Catalysts. ACS Catal. 2012, 2 (6), 1165–1178. https:// doi.org/10.1021/CS200418W. 112. Ligthart, D. A. J. M.; van Santen, R. A.; Hensen, E. J. M. Supported Rhodium Oxide Nanoparticles As Highly Active CO Oxidation Catalysts. Angew. Chem. 2011, 123 (23), 5418–5422. https://doi.org/10.1002/ANGE.201100190. 113. Spezzati, G.; Benavidez, A. D.; DeLaRiva, A. T.; Su, Y.; Hofmann, J. P.; Asahina, S.; Olivier, E. J.; Neethling, J. H.; Miller, J. T.; Datye, A. K.; Hensen, E. J. M. CO Oxidation by Pd Supported on CeO2(100) and CeO2(111) Facets. Appl. Catal. Environ. 2019, 243, 36–46. https://doi.org/10.1016/J.APCATB.2018.10.015.

Introduction: A short history of single site catalysis

17

114. Hegde, M. S.; Madras, G.; Patil, K. C. Noble Metal Ionic Catalysts. Acc. Chem. Res. 2009, 42 (6), 704–712. https://doi.org/10.1021/AR800209S. 115. Weiyu, S.; Jansen, A. P. J.; Hensen, E. J. M. A Computational Study of the Influence of the Ceria Surface Termination on the Mechanism of CO Oxidation of Isolated Rh Atoms. Faraday Discuss. 2013, 162, 281–292. https://doi.org/10.1039/C3FD20129E. 116. Kibis, L. S.; Svintsitskiy, D. A.; Derevyannikova, E. A.; Kardash, T. Y.; Slavinskaya, E. M.; Stonkus, O. A.; Svetlichnyi, V. A.; Boronin, A. I. From Highly Dispersed Rh3 þ to Nanoclusters and Nanoparticles: Probing the Low-Temperature NO þ CO Activity of Rh-Doped CeO2 Catalysts. Appl. Surf. Sci. 2019, 493, 1055–1066. https://doi.org/ 10.1016/J.APSUSC.2019.07.043. 117. Masatake, H.; Tetsuhiko, K.; Hiroshi, S.; Nobumasa, Y. Novel Gold Catalysts for the Oxidation of Carbon Monoxide at a Temperature Far Below 0  C. Chem. Lett. 2006, 16 (2), 405–408. https://doi.org/10.1246/CL.1987.405. 118. Bond, G. C.; Sermon, P. A. Gold Catalysts for Olefin Hydrogenation. Gold Bull. 1973, 6 (4), 102–105. https://doi.org/10.1007/BF03215018. 119. Thompson, D. New Advances in Gold Catalysis Part I. Gold Bull. 1998, 31 (4), 111–118. https://doi.org/10.1007/BF03214775. 120. Thompson, D. New Advances in Gold Catalysis Part II. Gold Bull. 1999, 32 (1), 12–19. https://doi.org/10.1007/BF03214784. 121. Hutchings, G. J. Heterogeneous Gold Catalysis. ACS Cent. Sci. 2018, 4 (9), 1095–1101. https://doi.org/10.1021/ACSCENTSCI.8B00306. 122. Cattaneo, S.; Stucchi, M.; Villa, A.; Prati, L. Gold Catalysts for the Selective Oxidation of Biomass-Derived Products. ChemCatChem 2019, 11 (1), 309–323. https://doi.org/ 10.1002/CCTC.201801243. 123. Cimino, A.; Stone, F. S. Oxide Solid Solutions. In Handbook of Heterogenous Catalysis; vol. 3; Wiley-VCH, 2008; pp 1362–1373. 124. Stone, F. S. The Significance for Oxide Catalysis of Electronic Properties and Structure. J. Solid State Chem. 1975, 12 (3–4), 271–281. https://doi.org/10.1016/00224596(75)90319-9. 125. Albright, T. A.; Burdett, J. K.; Whangbo, M. H. Orbital Interactions in Chemistry: Second Edition, Wiley Interscience, 2013; pp 1–819. 126. Suntivich, J.; Gasteiger, H. A.; Yabuuchi, N.; Nakanishi, H.; Goodenough, J. B.; Shao-Horn, Y. Design Principles for Oxygen-Reduction Activity on Perovskite Oxide Catalysts for Fuel Cells and Metal–Air Batteries. Nat. Chem. 2011, 3 (8), 647. https://doi.org/10.1038/nchem.1093. 127. Marchetti, L.; Forni, L. Catalytic Combustion of Methane over Perovskites. Appl. Catal. Environ. 1998, 15 (3–4), 179–187. https://doi.org/10.1016/S0926-3373(97) 00045-3. 128. Sengar, A.; van Santen, R. A.; Kuipers, J. A. M. Deactivation Kinetics of the Catalytic Alkylation Reaction. ACS Catal. 2020, 6988–7006. https://doi.org/10.1021/ acscatal.0c00932.

6.02

Synthesis and application of (nano) zeolites

Ana Palcica and Valentin Valtchevb, a Division of Materials Chemistry, Laboratory for Synthesis of New Materials, Ruđer Boskovic Institute, Zagreb, Croatia; and b Laboratoire Catalyse et Spectrochimie, Normandie Université, ENSICAEN, UNICAEN, CNRS, Caen Cedex, France © 2023 Elsevier Ltd. All rights reserved.

6.02.1 6.02.2 6.02.3 6.02.3.1 6.02.4 6.02.4.1 6.02.4.2 6.02.4.3 6.02.4.4 6.02.4.5 6.02.4.5.1 6.02.4.5.2 6.02.4.5.3 6.02.4.5.4 6.02.4.5.5 6.02.4.5.6 6.02.5 6.02.5.1 6.02.5.1.1 6.02.5.1.2 6.02.5.1.3 6.02.5.1.4 6.02.5.1.5 6.02.5.2 6.02.5.3 6.02.5.4 6.02.6 References

Introduction The structure of zeolites The properties of zeolites The properties of nanosized zeolites The synthesis of zeolites Components of the synthesis mixture Nucleation and crystal growth Interzeolite conversion Zeolite synthesis by the assembly of pre-formed layers The synthesis of nanosized zeolites The conventional synthesis of nanosized zeolites Synthesis of nanosized zeolites via interzeolite conversion Seed-assisted synthesis of nanosized zeolites Synthesis of nanosized zeolites by modifying the initial precursor Special cases of nanosized zeolites Alternative reaction conditions for the synthesis of nanosized zeolites The application of (nano) zeolites Catalysis FCC process MTO process Abatement of nitrogen oxides (deNOx) Biomass valorization Other reactions Adsorption and gas separation Ion-exchange Other application fields Summary and outlook

18 19 19 20 20 21 22 23 24 25 25 26 26 27 27 28 29 29 29 30 31 31 32 33 34 34 34 35

Abstract The article is devoted to the synthesis of zeolite materials in an accent on the nano zeolites. After a general introduction of zeolite materials, the factors controlling zeolite nucleation and growth are addressed. This information is further used to discuss the key issues related to the synthesis of nano zeolites. A comprehensive analysis of the different nano zeolites synthesis methods is performed, and the synthesisdproperties relationship is addressed. The physicochemical properties of (nano) zeolites are discussed in light of their practical uses. The article also includes an overview of the current and potential applications of zeolites. The review is completed with an outlook of the field and the prospects for (nano) zeolites.

6.02.1

Introduction

The term “zeolites” refers to three-dimensional (3D) crystalline alumosilicate porous materials. According to IUPAC classification, the zeolites are microporous materials with pores smaller than 2 nm.1 Their chemical composition is represented by the general empirical formula M2/nO : Al2O3 : y SiO2 : w H2O, where y > 2, n is the cation charge, and w represents the water confined in the voids of the zeolite. Besides the aluminosilicate zeolites, there are zeolite-type (zeotype) materials (silicoaluminophosphates, aluminophosphates, germanosilicates, zincosilicates, etc.). The common denominator of all zeolitic materials is their structure built of tetrahedra having central T atom (T ¼ Si, Al, Ge, B, Ga, P, .) and vertex oxygen atoms. Each vertice is shared with the neighboring tetrahedron.2 When possible variations of the zeolite framework TeOeT angles and distances together with framework densities as well as energies of postulated configurations are accounted for, it has been calculated there could be over two million prospective zeolite structures.3 However, there is a marked discrepancy between the theory and the experimental feasibility as currently there are

18

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00005-4

Synthesis and application of (nano) zeolites

19

253 known zeolite framework types approved by the Structure Commission of the International Zeolite Association.4 Novel framework types are continuously discovered, yet there are many challenges to be circumvented by the synthetic chemists to achieve a substantially larger number of zeolite structures. Furthermore, it should be stressed that many zeolite materials occur as natural minerals, yet the majority of synthetic materials do not have their natural counterparts. Namely, nowadays, approximately 70 known minerals have zeolite-type structure.5 The major large-scale zeolite usage fields are heterogenous catalysis, molecular adsorption, gas separation, and ion-exchange. Zeolites are employed in water treatment, air purification, petrochemical, fine chemicals industry, and agriculture. The specialty applications involve paints and coatings, plastics, the personal care industry, and the automotive industry.6 In 2018 the global synthetic zeolites market reached a value of US$ 5.2 billion, and it is projected to grow steadily.7 Hence, it is necessary to research this area. Present studies of zeolites are in accordance with the societal demands and follow the general established trends in materials science: the preparation of new materials (zeolite framework types) with novel properties as well as the improvement of the properties of the already known zeolites for further enhancement of their performance in standard industrial processes and possible new applications. Also, nanosized zeolites are becoming a more and more important class of zeolitic materials due to their versatility in finding a solution to surmount the issue of diffusion limitations and active site accessibility within micron-sized zeolites. Therefore, a detailed understanding of the basic mechanisms of the processes occurring during zeolite synthesis (nucleation and crystal growth) to control their crystal size remains a challenge. This article will address the basic properties of zeolites, with an emphasis on the nano zeolites, and outline various synthesis avenues that lead to the preparation of zeolite materials with desired properties. The already established applications and the emerging uses of nano zeolites will be addressed in the final part of the article.

6.02.2

The structure of zeolites

The zeolite framework is built of connected tetrahedra that form a defined 3D network of voids and channels within the structure, as displayed in Fig. 1. In this respect, there are one-, two- and three-dimensional channel-system zeolites, as in the case of two- and three-dimensional channel systems, the channels are separated or interconnected. Normally, zeolite frameworks are depicted as an assembly of simpler subunits of various combinations of central tetrahedral (T) atoms, neglecting the oxygen atoms. The so-called secondary building units (SBUs) are postulating by the condition that the whole network will be constructed by the multiplication of a single building unit.8 Very often, one structure can be represented by more SBUs. Composite building units (CBUs) are structural fragments that appear in several zeolite frameworks, making them convenient for comparing frameworks. From the application point of view is important the zeolite structure representation in terms of the size of rings that are the limiting pore openings. There are zeolite materials with pore openings restricted by rings having 8 T-atoms (8-membered ringsd8MR, small pore zeolites), 10-membered rings (10MR, medium pore zeolites) and 12-membered rings (12MR, large pore zeolites). Further, there are zeolites of pore openings larger than 12MRs, known as extra-large pore zeolites.

6.02.3

The properties of zeolites

All silica zeolite materials have an electrically neutral framework and are thus hydrophobic. Once other T atoms such as trivalent aluminum get incorporated in the framework, its charge becomes negative, thus generating the material’s hydrophilic nature. In the latter case, the charge-compensating cations in the cavities are necessary to achieve structural neutrality. These cations are exchangeable, and therefore zeolites act as ion-exchangers. The molecular sieving ability (the separation of molecules on the grounds of size

Fig. 1

Representations of CHA- and MFI-type zeolite framework structures.

20

Synthesis and application of (nano) zeolites

exclusion) is a direct consequence of their open framework with pores and the pore openings of molecular dimensions as well as these voids’ shape. Furthermore, the materials with different chemical compositions yet of the same framework type exhibit different acid-base properties. Namely, the zeolites possess both Brønsted and Lewis type of acid sites. In zeolites, the Brønsted sites originate at the bridging hydroxylsdOH groups that ensue at oxygen atoms that bridge two different T atoms, e.g., Si and Al.9 The OH group’s acid strength is a function of the zeolite framework, more precisely the TeOeT bond angles and the chemical composition of a particular material.10 On the other hand, certain computational studies indicate that the acid strength is independent of the zeolite framework, yet it results from the confinement effect.11,12 Lewis acid sites are coordinatively unsaturated and are able to accept electrons such as extra-framework Al species that may get formed upon the calcination of the zeolite.9 The high internal surface area of zeolites is a consequence of their open structure and enables the adsorption of various molecules. Generally, the (hydro)thermal stability of zeolites depends on their chemical composition, with high silica zeolites being rather hydrothermally stable. In contrast, the stability of materials with an increased fraction of framework Al is reduced. In addition, zeolites may exhibit redox properties when in the framework are embedded T atoms that may undergo oxidation state change (Ti, Fe, Cr). In the end, it is important to emphasize that various post-synthesis treatments can be used to alter the properties of a zeolite.13,14 However, this pathway of obtaining zeolites with pre-specified features is out of this article’s scope.

6.02.3.1

The properties of nanosized zeolites

Reducing the size of the crystals gives rise to new opportunities in tuning the performance of a material in an established application, yet it may open possibilities for some new fields of usage. Nanocrystals have a greater external surface area than micron-sized crystals. Thus, a larger fraction of framework atoms are exposed to the crystal surface. Further, nanosized crystals of porous materials such as zeolites have shorter intra-crystal pathway; thus, the guest molecules’ diffusion path length is shorter. Besides, the active sites are more accessible and available for interactions. The crystal size does not affect the ion-exchange capacity nor the adsorption capacity. However, it might affect the efficiency of these processes and namely, the reaction kinetics. Accordingly, the catalytic efficiency of nanosized zeolites is altered with respect to micron-sized crystals. The nano zeolites are often compared with the hierarchical zeolite in term of catalytic performance. The key advantage of nanosized zeolites over hierarchical zeolite materials, which are also designed to overcome diffusion issues in zeolites, lies in the fact that nanosized zeolites are mostly prepared by in situ approach. The direct synthesis is an integrated process with a greater reproducibility degree due to fewer preparation steps. The preparation of hierarchical materials requires either post-synthesis treatment or complex sacrificial templates with all related negative consequences. On the other side, the major obstacle in large-scale production and use of nanosized zeolites is the post-synthesis processing, namely their purification. Furthermore, to prevent the aggregation of the crystals, the freeze-drying procedure is required.

6.02.4

The synthesis of zeolites

In 1756, Swedish mineralogist Axel von Crönsted discovered the first natural zeolite stilbite. He named the mineral zeoliteda “boiling stone” (Greek words “zeo, z3 u”dI boil and “lithos, lιq o2”dstone) since upon heating in blowpipe, it became swollen.15 Natural deposits of zeolites occur at geological locations of different genesis (saline lakes, alkaline soils, marine sediments, cooled volcanic flows, hydrothermal alterations, burial metamorphism, etc.). The common factors are that they are formed at elevated pH in a water-rich environment at temperatures up to 350  C from volcanic glass, biogenic silica, poorly crystalline clays, plagioclase, nepheline, and quartz.16 H. de St Claire Deville, in 1862, reported the first hydrothermal synthesis of a zeolite (levynite).17 Until the mid-1930s, main zeolite properties have been identified, yet the synthesis work cannot be corroborated due to insufficient characterization and the inability to reproduce the reactions. The synthesis of zeolites as we know it today began with the work of Richard M. Barrer in the mid-1930s and Robert M. Milton in the 1940s.18 The work of Barrer was inspired by the processes taking place during the genesis of the natural zeolites. He treated known mineral phases in strong salt solutions at temperatures 170–270  C. In 1948, the group of Barrer reported that among the products were the first zeolites without natural counterpartdzeolites P and Q (KFI structure). Milton was the first to employ freshly precipitated alumosilicate gels, more reactive starting mixtures for the zeolite synthesis using lower temperatures and pressure conditions. By 1955, Milton’s group at Linde Division of the Union Carbide Corporation had synthesized 20 zeolites (A, X, .), and 14 of them were unknown as natural minerals. In 1961 Barrer and Denny reported on introducing quaternary ammonium cations into alumosilicate gels for the zeolite synthesis. Latter this approach enabled the preparation of many high-silica zeolites (zeolite beta, ZSM-5, .). Since 1982, when the first aluminophosphate was discovered, many “zeotypes” (zeolite-like materials) with different T atoms were synthesized: silicoalumophosphates, germanosilicates, gallosilicates, titanosilicates, zincosilicates, etc.19,20 Zeotypes often have a new type of framework. Some of them possess very large poresdaluminophosphate VPI-5 (VFItype) has a 14-membered ring pore; cloverite, a gallophosphate with an interrupted framework, 20-membered ring pore; ITQ-37 is a germanosilicate with 30-membered ring pore.4 A new class of ordered mesoporous materials was discovered in 1992 by Mobil researchers using surfactant molecules as templates.21,22 They have periodic structures with pore sizes much larger than zeolites, but the pore walls are amorphous which impacts strongly their properties. Likewise, since the late 1980s and in the 1990s, new research directions focusing on particular

Synthesis and application of (nano) zeolites

21

properties have been initiated, and other new types of microporous materials have been obtained, including extra-large pore, chiral, and 2D zeolites.23–28 Target properties of zeolites involve their size, as the quest for nanomaterials is high;29,30 morphology;31,32 hydrophobicity/ hydrophilicity;33,34 acidity in terms of the nature, quantity, and location of acid sites;35 additional pore systems (hierarchical materials);36,37 hydrothermal stability38,39 and chemical composition40,41. The new developments of zeolite materials have been facilitated by utilizing novel organic structure-directing agents, incorporation of T atoms other than Si and Al which have different structure-directing action as well as hydrolysis tendency (B, Ge) but also by introducing and/or developing innovative synthesis procedures such as high-throughput and ultrafast synthesis approach, along with numerous post-synthesis treatments of zeolite crystals.42–49 It is important to highlight that the preparation of zeolite materials both in industry and academia is an application-driven process. The zeolites’ properties are engineered on the grounds of the requirements that directly arise from their practical usage.

6.02.4.1

Components of the synthesis mixture

Zeolite yielding systems (Fig. 2) comprise four essential elements: (1) T atom source; (2) solvent; (3) mineralizer (mobilizer); and (4) structure-directing agent (SDA). The crystal growth process is a function of the physicochemical properties, especially the T atom source solubility. Further, certain T atom source chemical may preferably produce particular zeolite material or crystal morphology.50,51 Water is the prevalent solvent for zeolite synthesis due to its ability to dissolve the precursor species, stability, thermal conductivity, availability, bio-compatibility, etc. Solvents other than water (ionic liquids, ethylene glycol, triethylamine, glycerol)52–55 have been efficient when employed in zeolite synthesis enabling new reaction conditions and offering new opportunities in designing zeolite materials. In zeolite reaction mixtures, hydroxide and fluoride anions act as mineralizers. The mineralizer’s role is to solubilize the reactants by forming complexes of various polycondensation degrees depending on the reaction conditions and mineralizer concentration. The chemistry of silicate species differs in the presence of a specific mineralizer, which will be discussed in the following section. The preparation of high-silica zeolites in hydroxide media leads to materials of a rather high number of framework silanol defects that balance SDA’s positive charge. Since there is a deficiency of trivalent T-atoms, the framework generates a negative charge by the siloxy –Si-O sites. In fluoride systems, the fluoride anion bonds to Si to form penta-coordinated SiO4F sites that neutralize the SDA charge. Fluorides are removed upon calcination resulting in defectless material.56 Besides, the presence of fluoride anions in the synthesis mixture enabled obtaining all-silica counterparts of some zeolite materials, and several new zeolite frameworks have been discovered.57–60 In addition, owing to the stabilizing interactions with the SDAs, fluorides facilitate the formation of extralarge pore zeolites.61 Furthermore, fluoride systems usually yield larger crystals than hydroxide ones.61 In a hydroxide medium, a high supersaturation of (alumino)silicate species is achieved. Thus, plenty of nuclei are formed, resulting in a crystalline product of smaller crystal size. The arranged network of zeolite voids and channels is a consequence of the electrostatic interactions between structure-directing agents and (alumino)silicate precursors.62 Alkali and alkali-earth cations with their hydration spheres act as structure-directing agents for natural and low-silica synthetic zeolites.51 The cation prefers to reside at a specific site within a particular framework.63 The inorganic cations present within the voids are exchangeable, and zeolite ion-exchange ability is a function of their quantity and location.64 A range of organic molecules may act as organic structure agents (OSDAs): quaternary, di-quaternary and tri-quaternary alkylammonium cations, surfactants, imidazolium derivatives, amines, alcohols, quaternary and di-quaternary alkylphosphonium cations, quaternary and di-quaternary alkyl-aminophosphonium cations, phosphazene bases, sulphonium cations, aza-, and oxocrown macrocycles, metal complexes, proton sponges, self-assembled compounds, etc.42 Furthermore, using OSDAs results in an increase in the zeolite Si/Al ratio since fewer large organic molecules can get confined within zeolite cavities than small inorganic cations. The removal of entrapped organic species in the final zeolite usually takes place during calcination by thermal decomposition. Besides, successful zeolite synthesis with the degradable (and sometimes recyclable OSDAs) has been reported in 2003 by Zones and Davis.65 The idea was to eliminate the combustion step, but this approach was not largely pursued. The OSDAs sometimes serve as space-fillers in the zeolite void space, providing stability from van der Waals interactions with the zeolite network. This is exemplified by the fact that different OSDAs can make one zeolite type.66 On the other hand, the size and the

Fig. 2

Schematic representation of the zeolite synthesis process.

22

Synthesis and application of (nano) zeolites

shape of the molecule can fit within void space and consequently direct zeolite structures more specifically.67–69 Recently, a new concept to zeolite synthesis was introduced. It is based on using the organic species that mimic the transition state of the preestablished reaction to be catalyzed.70 Indeed, materials prepared in this way present higher catalytic activity in the target reaction. Clearly, each of the properties of the OSDA (size, shape, hydrophobicity/hydrophilicity, polarity, electronic properties, rigidity, stability) affects its structure direction action, and consequently, the zeolite properties.71–73 In addition, combining alkali cations and OSDAs is a common practice in zeolite synthesis that also affects the zeolite formation processes. Thus, changing the ratio of alkali cations and OSDAs can produce different products in terms of crystalline phase, chemical composition, morphology, and acidic properties.74–77 Accordingly, materials of different properties are produced by combining two different OSDAs.78,79

6.02.4.2

Nucleation and crystal growth

Upon mixing the ingredients, the initial zeolite preparation mixture is formed. According to their appearance to the naked eye, there are two principal types of zeolite preparation systems: (i) homogeneous systemsdoptically clear solutions without any macroscopic evidence of the presence of the solid phase, (ii) heterogeneous systemsdhydrogels having discrete solid (gel) and liquid (mother liquor) phases.80,81 Ultradense gels without bulk liquid phase are also used in particular zeolite synthesis.80 Nucleation and (crystal) growth in the synthesis system are the principal processes that lead to the final zeolite product. A schematic representation of classical nucleation pathways is depicted in Fig. 3. Nuclei are the smallest entities within a system that possess structural features and can evolve to the final material. The number of nuclei determines the size of the ultimate crystals; hence the control of nucleation is essential for obtaining nanosized zeolites. Besides, the initial system’s chemical composition and the reaction conditions impact the reaction kinetics and finally its completion. Zeolite yielding systems are considered highly complex, and thus it is difficult to predict the outcome of the synthesis process.82 The synthesis mixture may comprise a plethora of the (reactive) species whose distribution is a factor of the source materials and the ratio between the reactants (OSDA/Si, Si/Al, Si/H2O, pH, etc.), crystallization temperature, heating period, agitation, etc.83–86 In such a system, numerous interactions between these species occur as polymerization/depolymerization, dissolution, gel formation, gel reorganization, etc. The enthalpies of these reactions are very similar, meaning that the competing reactions, equilibria, and interactions are kinetically controlled, yielding a metastable end product, i.e., the zeolites.87–89 Since the early periods of zeolite research, various approaches towards the understanding of zeolite synthesis mechanism have been developed.90–98 The majority of zeolite scientists accept the concept of heterogeneous nucleation in zeolite formation.99–101 It is proposed that the nuclei are formed within gel particles, not in the liquid phase, and many experimental and modeling studies corroborate this assumption.102–107 A recent study of zeolite precursors in a high-alumina OSDAs-free system have shown that the gel particles exhibit a core-shell structure.102 The onset of nucleation was found to occur in the impenetrable shell part, which gradually gets transformed into zeolite crystals. Once the shell is mostly digested, the reactive species’ transfer between the mesoporous core and the bulk liquid phase is enabled. At this moment, the intensive zeolite crystal growth commences, and the mass

Fig. 3

Schematic representation of classical nucleation mechanisms.

Synthesis and application of (nano) zeolites

23

transformation ensues.102 Furthermore, porous precursor particles were observed in the Zn-containing system yielding RSN-type zeolite material.50 These and previous findings demonstrate that the precursor particles’ engineering affords the preparation of zeolite materials with predetermined properties.101,108–112 In OSDA containing systems for the preparation of high-silica zeolites, the majority of the studies have been performed in initially clear solutions,113–116 as MFI yielding systems were preferred for their simplicity29,117–123. The amorphous precursors of silicalite-1 comprise primary particles having the size 1–6 nm and exhibiting a core-shell structure. The core is silica-rich and recently was found that the shell is composed of silicate oligomers associated with TPAþ cations.124,125 When sodium was added to such a system, it partially displaced some TPAþ ions from the particles’ shell.126 Consequently, TPA-free terminal –OH and –O– groups get bounded upon the nanoparticles’ collision leading to their coalescence. Once a stable nucleus is formed, it grows into a crystal by transport-controlled processes or processes determined by the surface reactions.127,128 Recent AFM studies demonstrated that classical surface-controlled mechanism and non-classical aggregate growth (attachment) of the pre-formed nanoparticles occur during zeolite synthesis.129,130 It appears that particle attachment is the dominant pathway, while the surface reactions become significant once the population of nanoparticles has decreased. Besides, an additional distinct growth mechanism has been identified in the zeolite A formation, where gel-like islands from supersaturated aluminosilicate solutions are the key feature of this growth pathway.131 Furthermore, the intimate crystal structure analysis may provide insight into their formation mechanisms. For instance, a thorough study of micron-sized crystals of several zeolite types revealed they are composed of intergrown subunits.132 The unbiased framework Si and Al extraction in fluoride medium revealed the abundant presence of misaligned nanodomains in crystals.133 The zeolite crystals were found to be inherently built of numerous nanodomains with well-defined size, shape, and structure. It is most probably a direct consequence of the crystal growth by nanoparticle attachment to the growing face. The newly adhered particles can be tilted or rotated to the growing surface, and consequently, crystalline domains with discernible grain boundaries are formed. Seed crystals are added to a reaction mixture to induce zeolite crystallization.134 The benefits of the seeding procedure involve the shorter synthesis period, direct the formation of the desired zeolite, and prevent by-products. Besides, there is an environmentfriendly aspect due to the absence of OSDA. The calcination is not required, which prevents the aggregation of nanoparticles and Al’s release from the framework. Sometimes even a “catalytic” amount of seeds is sufficient to promote the desired material’s synthesis.135 In zeolite systems, nanosized zeolites smaller than 200 nm are the most advantageous seed crystals since they provide a larger yield and more uniform crystal size distribution, which is of particular importance in preparing membranes.136,137 Okubo and co-workers proposed the CBU hypothesis, i.e., “the target zeolite should be added as seeds to a gel that yields a zeolite containing the common composite building unit when the gel is heated without seeds.”138 The hypothesis was proved by adding MEL-type and MFI-type seeds to MOR-type zeolite yielding gel and preparing respective target materials. The common building unit in all three frameworks is mor.4 Similarly, PAU-type zeolite was obtained from MER-yielding gel.138 The seeds’ size and quantity affect the size and the morphology of the final crystals and sometimes their framework composition.139 The type of seeds, single crystals or aggregates, control the growth process in terms of single-crystal or aggregated growth.134 It should be noted that adding seeds is one of the crucial issues in ultrafast synthesis.140

6.02.4.3

Interzeolite conversion

The zeolite is a thermodynamically metastable phase and, under certain conditions, can be transformed into a more stable crystalline product. For this reason, the same reaction mixture can yield consecutively two or more different products, depending on the reaction conditions (temperature, agitation, aging, .), or, in some cases, two or more crystalline phases can be formed at the same time.141 Also, the transformation of one zeolite to another at elevated temperature is possible.142 The transformation of one zeolite to another upon heating in aqueous systems was observed in the early periods of zeolite research.51 The interzeolite conversion can be conducted in the absence/presence of organic templates as well as in systems with added crystalline seeds.143 Since the starting silica-alumina source is different it is expected that the obtained crystals will have different properties (crystal size and morphology, chemical composition) than the conventionally prepared ones. This is exemplified in a study by Tang et al. where zeolites L and beta were hydrothermally treated in reaction mixtures with similar chemical composition targeting zeolite SSZ-13 synthesis. In zeolite L system were formed uniform 2–3 mm SSZ-13 (CHA-type material) crystals, whereas zeolite Beta-derived SSZ-13 containing two populations of crystals with size 1–2 and 4–5 mm.144 The interzeolite conversion is also a way to obtain zeolites that are difficult to be synthesized from conventional hydrogel precursors. Subotic and co-workers were the first to study the kinetics of zeolite conversion.145,146 They concluded that the transformation is a solution-mediated process. The first step is the dissolution of the starting zeolite, and the new phase is formed without any intermediates, as demonstrated in Fig. 4. Similarly, Khodabandeh and Davis reported that the transformation of zeolite P “occurs via continuous and gradual dissolution of P1 and subsequent nucleation from the aluminosilicate species released into the solution” without any precursor (amorphous phase).147 Sano et al. showed that zeolite Beta’s formation is faster from FAU as a starting material than from amorphous silica and a- or g-alumina.148 The crystallization rate enhancement was associated with the small fragments (nanoparts) of ordered aluminosilicate species generated by the decomposition/dissolution of the initial zeolite and their reassembly.149 It is generally accepted that the similarity between the composite building units of the initial and resulting zeolites ensures the possibility of transformation. Yet, Rimer et al. succeeded in obtaining ZSM-5 zeolite from FAU-type material, demonstrating that a common building unit is not always necessary to convert one zeolite into another.150 Clearly, the interzeolite

24

Synthesis and application of (nano) zeolites

 c, Fig. 4 Interzeolite transformation of zeolite A into hydroxysodalite in alkaline media. Reproduced with permission from reference Subotic, B.; Skrti   D.; Smit, I.; Sekovanic, L. J. Cryst. Growth 1980, 50, 498–508.

conversions offer new opportunities for tuning zeolite properties in terms of phase and chemical composition, size, morphology, and acidic properties.

6.02.4.4

Zeolite synthesis by the assembly of pre-formed layers

Conversion of layered precursors into zeolites and the ADOR approach represents another avenue to new zeolitic materials composed of precursors with a certain degree of order.151,152 Two-dimensional layered zeolite precursors (2D) constitute a repetition of individual layers (sheets), bonded by weak forces such as hydrogen bonds or Van der Waals interactions. Their structure can be modified using the free space between the adjacent layers, leading to new materials, most commonly three-dimensional (3D) zeolites (Fig. 5). Pillared layered structures are constructed by inserting the organic molecules into the interlayer space (swelling) and further intercalating more thermally stable compounds that act as pillars between two sheets. The first example of this procedure was swelling the MCM-22 layers with hexadecyltrimethylammonium hydroxide followed by pillaring with tetraethylorthosilicate and calcination, which resulted in the 3D MCM-36 material.151 The method was further extended to FER-, MFI- as well as NSI-type materials.153–155 The individual layers can be completely separated, disordered, and randomly distributed in the “house of cards”-form by delamination process, as in the case of delaminated MCM-22 precursor (ITQ-2 material), FER-type precursor (ITQ-6 material), and NSI-type precursor (ITQ-18 material).28,153,156 This avenue in zeolite synthesis was enriched by the ADOR approach involving assembly (A), disassembly (D), organization (O), and reassembly (R) steps.157 Assembly step connotes the synthesis of crystalline zeolite material that usually contains germanium in double 4-membered rings (D4Rs), e.g., UTL-, IWW-, ITR- and ITH-type materials. Acid hydrolysis of D4Rs due to the lower stability of the chemical bonds GeeO than SieO is conducted during the disassembly step. There are many possibilities for the organization of the layers formed during the disassembly stepdintercalation of an organizing agent, self-organization of the layers upon heating, de-intercalation of any residual species remaining between the layers together with their alignment and rearrangement of the silica within the layers. Finally, reassembly into a fully connected framework occurs during the calcination of the new material. This method enabled us to achieve several new zeolite frameworks, even some that were previously considered unfeasible.158–160 The ADOR method also allows functionalizing some materials and controlling the acid sites distribution within the zeolite framework.161,162

Synthesis and application of (nano) zeolites

25

Fig. 5 Transformation of layered precursors into zeolites employing various post-synthesis methods. Reprinted with permission from Roth, W. J.;  Nachtigall, P.; Morris, R. E.; Cejka, J. Chem. Rev. 2014, 114, 4807–4837. https://pubs.acs.org/doi/10.1021/cr400600f. Copyright (2014) American Chemical Society.

6.02.4.5

The synthesis of nanosized zeolites

In the early days of the investigation of nanosized zeolites, they were designated as crystals having a size smaller than 1000 nm.163 Namely, nanosized zeolites were firstly studied as colloidal suspensions, which by definition comprise particles smaller than 1 mm.164 Nowadays, nanosized zeolites are considered as materials having crystals smaller than 100 nm.165 Besides the conventional synthesis approach, various methods to obtain nanosized zeolites have emerged. In the following section, we will address both traditional and non-conventional synthesis routes to nanosized zeolites.

6.02.4.5.1

The conventional synthesis of nanosized zeolites

The conventional synthesis of zeolites assumes heating a hydrogel precursor in a closed reaction vessel. In order to prepare small crystals with a narrow particle size distribution, nucleation needs to be favored over crystal growth, and a uniform nuclei distribution has to be achieved. Hence, the simultaneity of the nucleation events is a prerequisite. This can be attained using uniform initial reaction mixturesdthe so-called “clear solutions”. Clear solutions are defined as optically clear sols that contain only discrete gel particles.80 The colloidal stability is maintained by preventing the aggregation between the particles via (1) using abundant amounts of OSDAs (primarily tetraalkylammonium cations);166 (2) controlling the water and alkali cations content in the reaction mixture;167,168 (3) lowering the crystallization temperature;169 and (4) pre-aging of the synthesis mixture170. In general, in the OSDA-containing systems, it is beneficial to keep the alkali cations as low as possible, whereas the dilution degree should be rather high.171 On the other hand, in the OSDA-free systems that yield zeolites with the nano-range size, a large excess of alkali cations is typical.172 In a similar manner as in other fields of science, the tendency in zeolites synthesis, and likewise in the preparation of nanosized zeolites, is to develop more environment-friendly as well as economically efficient processes. In this respect, it should be noted that a range of low-silica nanosized zeolites (LTA, LTL, EMT, SOD, FAU, RHO, etc.) has been prepared from OSDA-free systems having high concentrations of alkali hydroxides.141,169,172–175 In these systems, the colloidal precursors, were stabilized owing to the high concentration of alkali cations, while the crystal growth was decelerated at low temperatures. A sound example is a preparation of ultra-small FAU-type crystals (10 nm, Fig. 6) produced from the reaction mixture of molar oxide composition 9 Na2O : 0.7 Al2O3 : 10 SiO2 : 160 H2O at 50  C. Reducing the alkalinity (8 Na2O : 0.7 Al2O3 : 10 SiO2 : 160 H2O) and increasing the synthesis temperature resulted in zeolite Y having the size of the crystals of 70 nm whereas further dilution and an additional quantity of Al (8 Na2O : 1 Al2O3 : 10 SiO2 : 400 H2O) bring about zeolite Y crystals of 400 nm in size.172 The obtained materials present high crystallinity and micropore volume matching the micron-sized crystals. Besides, compared to 70 and 400 nm samples, significantly large external surface area values are reported for 10 nm material.

26

Synthesis and application of (nano) zeolites

Fig. 6 (A) DLS curves of the initial and crystalline suspensions of 10 nm nanosized zeolite Y and photographs of the corresponding suspensions. (B) TEM images of 10 nm nanosized zeolite Y crystals. Reproduced with permission from reference Awala, H.; Gilson, J.-P.; Retoux, R.; Boullay, P.; Goupil, J.-M.; Valtchev, V.; Mintova, S. Nat. Mater. 2015, 14, 447–451.

The current trend in the synthesis of nanosized zeolites is to become more sustainable and greener. Usually, this is achieved by using less water, combining OSDAs and alkali cations to reduce the usage of organics, investigating alternative OSDAs as well as adding certain quantities of growth modifiers, very often surfactants.176–180 An example of such a study is Catizzone et al.’s work where the addition of sodium lauryl sulfate (SLS; SLS/Al2O3 ¼ 0.3) during the ferrierite synthesis resulted in the formation of non-stacked plate-like crystals with size ranging between 300 and 500 nm.180 On the other hand, the systems without SLS produced elongated stacked plate-like crystals having the length 8–12 mm. Further, Zhang et al. accelerated the zeolite nucleation rate by increasing nutrient concentration in the synthesis gels by reducing the water content. The crystal size of ZSM-5 materials decreased from 5–6 mm to 100–200 nm when the H2O/SiO2 ratio dropped from 18 to 2. Also, zeolite beta with size 120– 130 nm is obtained for H2O/SiO2 ¼ 2.5, whereas H2O/SiO2 ¼ 6.25 yields crystals having the size 610 nm.176

6.02.4.5.2

Synthesis of nanosized zeolites via interzeolite conversion

A range of nanosized zeolites, particularly small-pore zeolites, was obtained by the interzeolite conversion approach. The nucleation is usually abundant and fast. In many cases, the final product’s crystals are smaller than the crystals of the starting zeolites, which indicates a solution mediated zeolite transformation process rather than direct solid-solid transformation. Chabazite is an archetype material for interzeolite conversion,181 and accordingly, there are many procedures developed to prepare nanosized CHA-type materials using various initial zeolites such as FAU-, LTL-, BEA-, MFI- and LEV-type materials.182–186 The reaction takes place in the presence of the common OSDA that yields CHA-type zeolites, N,N,N-trimethyl-1-adamantanamine hydroxide. Tetraethylammonium hydroxide and benzyltrimethylammonium were also employed. The first report on nanosized SSZ-39 crystals (AEI-type material; 40–50 nm) involved using high silica FAU-type zeolite precursor in the presence of tetraethylphosphonium or N,Ndimethyl-3,5-dimethylpiperidinium as organic structure-directing agents (OSDAs).187 Further, nanosized BEA-, MFI-, LTL-, LEV-, OFF-, MAZ-, ANA-, ERI- and AFX-type zeolite materials were synthesized by interzeolite conversion.188–194 It should be underlined that the most studied source zeolite material for performing interzeolite conversion is faujasite since it is available in a range of framework chemical compositions. The final product and its properties are a function of the synthesis conditions, particularly OSDA, alkali cations, and the seeds present in the reaction mixture.188,192,193

6.02.4.5.3

Seed-assisted synthesis of nanosized zeolites

The seed-induced synthesis provides particular opportunities in preparing nanosized zeolites in terms of size and chemical composition, usually with a reduced synthesis period. For instance, 400–500 nm zeolite beta can be prepared by seeded ultrafast synthesis within just 10 min.195 Furthermore, in the already mentioned study by Catizzone et al., the FER-type crystals obtained in the SLSsystem were added as seeds (3 wt%) in the otherwise identical reaction mixture and lead to the formation of the 100 nm ferrierite crystals.180 An extensive study of the synthesis of nanosized ZSM-5 crystals (140–230 nm) from OSDA-free system using a suspension of 80 nm non-calcined silicalite-1 seeds demonstrated that the ultimate size of the crystals is a function of seeds’ concentration. Besides, the crystals become smaller as the reaction temperature was getting lower.134 Employing the silicalite-1 seeds enabled Ren and al. to prepare 270 nm ZSM-5 crystals in less than 2 h. In addition, the Si/Al ratio of the end products increased with a higher amount of the added seeds.196 Furthermore, despite the common outcome of zeolite preparation in fluoride medium, where are mostly produced large zeolite crystals, conducting the synthesis of zeolites in seeded fluoride medium may yield nanosized zeolites. Thus, 300–400 nm zeolite beta and ZSM-5 with particle size below 200 nm (Fig. 7) were recovered using 10 wt% of seeds. These zeolite beta and ZSM-5 materials present fewer defects and different acidic properties than their hydroxide medium synthesized

Synthesis and application of (nano) zeolites

27

Fig. 7 SEM micrographs of ZSM-5 crystals synthesized in fluoride medium: (A) without seeds and with (B) 2.5 wt%, (C) 5 wt%. and (D) 10 wt% of added seeds. Reproduced from reference Qin, Z.; Lakiss, L.; Tosheva, L.; Gilson, J.-P.; Vicente, A.; Fernandez, C.; Valtchev, V. Adv. Funct. Mater. 2014, 24, 257–264, with permission.

counterparts. They were found more active and stable catalysts in transforming m-xylene and converting methanol into hydrocarbons, respectively.56,197 Most of the seed-assisted zeolite synthesis routes reported in the literature are based on utilizing as-prepared or calcined crystals. A particular top-down seeding procedure was developed where zeolite crystals larger than 1 mm are downsized by milling and subsequently transformed to nanosized materials. Milling inevitably causes the amorphization of the materials to a certain extent, especially at the outer surface of the particles generated during this process. The milled particles’ recrystallization resulted in highly crystalline nanosized zeolites since the amorphous parts were digested and converted into zeolite upon the hydrothermal treatment.198,199 This approach was further extended by designing a milling apparatus working at different temperatures under alkaline conditions, which enabled the in situ recrystallization of nanosized zeolites.200

6.02.4.5.4

Synthesis of nanosized zeolites by modifying the initial precursor

The preparation of classical zeolite precursor, removing the liquid phase, and performing the hydrothermal treatment of the remaining dried solid amorphous precursor in alkaline solution or exposing it to solvent vapors represents another avenue for obtaining nanosized zeolites. These reactions are often referred to as steam-assisted synthesis or dry gel conversions. Typically, the reaction period is reduced in comparison to the conventional hydrogel system, but the proximity of the nuclei in a concentrated system generally provides aggregated particles.201,202 Recently, this issue was circumvented by washing and subsequent freeze-drying of the amorphous precursors. The obtained crystals were discrete with preserved size and morphology of the precursor particles.203

6.02.4.5.5

Special cases of nanosized zeolites

In addition to classical nanosized zeolite crystals, several special cases of nanosized materials preparation zeolites were reported. The objective is to avoid the deficiencies related to nano zeolites’ post-synthesis processing and make the preparation similar to the conventional micron-sized synthesis. These materials preserve nanosized zeolites’ properties, i.e., decreased mass transportation limitations and enhanced accessibility to active sites, yet present some additional features. For instance, finned zeolites represent a special case of nanosized zeolite preparation. The particles are produced by epitaxial growth of fin-like protrusions on micronsized crystals, and when employed as catalysts in methanol to hydrocarbons reactions, they behave as pseudo-nanocrystals with

28

Synthesis and application of (nano) zeolites

sizes commensurate to that of the fin.204 However, it should be noted that the mass of nanosized particles (fins) with respect to total zeolite mass is negligible, and the impact on the catalytic performance is marginal. 6.02.4.5.5.1 Bi-dimensional zeolitesdzeolite nanosheets Bi-dimensional zeolites are usually in micron-sized range, yet they are nanosized in one dimension having the thickness of one to several unit cells. Thus they can be processed as conventional zeolite crystals but still offer enhanced transport properties.205,206 The delamination of zeolites, i.e., the disassembly step in the ADOR approach, represents a top-down method for obtaining zeolite nanosheets. Corma and co-workers reported the first example of a delaminated zeolite.28 Further, Ryoo et al. developed a bottom-up strategy for direct preparation of MFI-type zeolite nanosheets employing specific molecules: organic surfactant functionalized with a diquaternary ammonium group in the head.207,208 The ammonium head parts direct the zeolite framework’s formation, while the long hydrophobic chains self-assembled into a supramolecular structure dictate the lamellar crystal morphology. Altering the synthesis conditions enables the control of the degree of the disorder between the layers as well as interlamellar spacing. Besides, the extensive research efforts resulted in the preparation of MFI-type zeolite nanosheets from seeded surfactant-free system.206,209 6.02.4.5.5.2 Embryonic zeolites Embryonic zeolites (EZ) are ultra-small (3–5 nm) X-ray amorphous zeolitic units.210 In embryonic zeolites, the mass transfer limitations present in the zeolites are evaded by decreasing their size to a few nanometers.211 Akouche et al. have shown that their porosity, acidity, and substrate molecules’ accessibility can be tuned.212 The EZ exhibit N2 adsorption isotherm type Ib, characteristic of extra-large pore (1–2 nm) microporous materials (Fig. 8). Combining this feature with their acid sites of moderate strength opens possibilities for some new catalytic applications of zeolitic materials.213

6.02.4.5.6

Alternative reaction conditions for the synthesis of nanosized zeolites

Previously addressed synthesis routes concern performing the zeolite synthesis under conventional reaction conditions. This section will discuss the possibilities of harvesting nanosized zeolites from systems different from the typical ones in terms of heat transfer, solvents, and the available free space to transport reactive species/or crystal growth. In general, microwave yields smaller crystals of narrower particle size distributions than does conventional heating since the heating of the reaction vessel is uniform and faster, thus leading to the accelerated nucleation. Consequently, a shorter reaction

Fig. 8 N2 adsorption/desorption isotherms of embryonic zeolites prepared with various OSDAs: TMA, tetramethylammonium hydroxide; TPA, tetrapropylammonium hydroxide; TMAd, trimethyladamantylammonium hydroxide; TBA, tetrabutylammonium hydroxide; and THA, tetrahexylammonium hydroxide; ZSM-5 represents a reference. The inset is plotted in the logarithmic x-axis scale and displays the results in the low ( K > Cs. Likewise, the shape of the adsorption isotherms and the kinetics of adsorption depend on the cation form. Besides, in single-component adsorption measurements, all studied MER-type materials show very low methane uptakes. Zeolites are used in the separation of various gases. For example, the pore openings of SOD-type zeolite are limited by 6MRs (2.9 Å), which makes it a promising candidate for the separation of hydrogen from gas mixtures.292 A membrane comprising graphene matrix and nanosized sodalite crystals demonstrated high gas permeance and enhanced selectivity towards H2 from a mixture containing CO2 and H2O.293 Furthermore, silanated nanosized LTA zeolite as a component of a mixed-matrix membrane with poly(vinyl acetate) demonstrated high selectivity for separation of H2, and He mixed with N2.294 Besides, 100 nm zeolite X exchanged with Ba displayed higher selectivity towards para-xylene in a mix with other xylene isomers.295 The large external surface area brings about the possibilities of using nano zeolites to absorb various organic molecules, an issue particularly important for environmental protection. Thus, zeolites are suitable materials for removing volatile organic compounds emitted from different sources. For instance, mixed-matrix membranes of nanosized silicalite-1 and Ti-silicalite-1 dispersed in poly(vinylidenefluoride) present enhanced hydrophobic character with respect to pure polymer membrane and consequently

34

Synthesis and application of (nano) zeolites

Exposed to methane

Non porous

Zeolite Merlinoite CO2/CH4 mixture

CO2 triggered opening

Time on stream

Selective CO2 uptake

Fig. 13 Triggered gate opening and breathing effects during preferred CO2 adsorption over methane by merlinoite zeolite. Reprinted with permission from Georgieva, V. M.; Bruce, E. L.; Verbraeken, M. C.; Scott, A. R.; Casteel Jr., W. J.; Brandani, S.; Wright, P. A. J. Am. Chem. Soc. 2019, 141, 12744–12759.

improved hexane adsorption capacity.296 There are myriad examples of possible zeolite applications for eliminating hazardous organic compounds, such as the adsorption of pharmaceutically active substances by ZSM-5 zeolite, the selective removal of phenol from synthetic 2nd generation biofuels over commercial zeolite Y, etc.297,298

6.02.5.3

Ion-exchange

The greatest share of large-scale zeolite production is associated with their ion-exchange function. Namely, most of the synthetic zeolites produced nowadays are used as a component of solid detergents.235 The low-silica zeolites (LTA, GIS, FAU) possessing high ion-exchange capacity act as water softeners removing calcium and magnesium cations. They efficiently substitute phosphate-based detergent with additional advantages of low degree of abrasiveness and good dispersibility. Furthermore, the ion-exchange capacity combined with radiolytic stability, makes zeolites suitable for removing radionuclides and are thus employed in nuclear waste treatment. Zeolites were employed in clean-up after nuclear accidents such as Three Mile island, Chernobyl and Fukushima.299,300 Other ion-exchange related zeolites applications include wastewater remediation (elimination of toxic metal cations) and ammonia removal from animal wastes.299 In all these applications, the nano zeolites offer faster adsorption kinetics, which is advantageous when large water volumes are processed. However, in a few applications, zeolites can be used in powder form. The possible solutions are preparing zeolite monoliths or agglomerates using a binder.301

6.02.5.4

Other application fields

Over the years, zeolites’ utilization has been expanded to areas apart from the traditional application fields (petrochemistry, environmental) to agriculture, food industry, medicine, electronics, optics.165 Natural zeolites are mostly employed in agriculture. They may serve as either carrier of nutrients or medium to free nutrients, as soil ameliorants, in the remediation of heavy metal contaminated soils, for water retention, in animal nutrition, for odor control, storage pest management, for mycotoxin control, NH4þ removal, etc. Further, zeolites are extensively studied as ethylene scavengers added to food packaging.302,303 The nanosized zeolite X is an efficient additive to palm oil, inhibiting its oxidation.304 In addition, nano zeolites are used to eradicate pathogens and thus exhibit the potential to prevent oral diseases, inflammatory diseases, and hospital infections.305–307 Moreover, Gd-exchanged nanosized FAU-type zeolite could be used in cancer therapy as a gas-delivery agent for decreasing hypoxia in glioblastoma.308 It should be noted that zeolites are clinically already utilized in hemostatics as gastroprotective drugs and antioxidative agents.309 Lastly, (nanosized) zeolites are promising candidates for chemical sensing of various gases, ions, and biomolecules.310–313

6.02.6

Summary and outlook

This article reviews the large diversity of (nano)zeolite synthesis procedures and their applications’ versatility. The zeolite framework is microporous, which means that they have voids up to 2 nm in size. Several chemical elements (Si, Al, Ge, P, Ga, etc.) can reside in the center of tetrahedra and form a zeolite framework. Consequently, each T-atom brings about particular features of zeolite materials. Nevertheless, the most frequently used zeolitic materials are built of Si and Al. All-silica zeolites present a neutral framework and are hydrophobic. In contrast, rising Al content generates a negative framework charge, and thus the materials become hydrophilic and possess the ion-exchange ability. Further, zeolites can have Lewis acid sites and in H-form may present Brønsted acidity. In nanosized zeolites, the fraction of framework atoms at the external surface is higher than in micron-sized crystals. Besides, smaller crystals exhibit a larger external surface area and shorter intra-crystal channels. Consequently, this shortens the diffusion path lengths, renders more accessible active sites, and makes the catalytic, adsorption, and ion-exchange processes more efficient.

Synthesis and application of (nano) zeolites

35

The properties of zeolites dictate their utilization. In turn, the properties of zeolites ensue as an aftermath of their preparation process. The synthesis of zeolites is guided by the processes occurring during the genesis of the natural zeolites. The development of zeolite synthesis methods started in the 1930s with Barrer treating known minerals in strong salt solution under high-temperature hydrothermal conditions. Milton introduced freshly precipitated alumosilicate gels and low-temperature hydrothermal synthesis. Barrer and Denny used quaternary ammonium cations that lead to the preparation of high-silica zeolites. These pioneering works continue to be the fundament of modern zeolite production. Further advances in zeolite synthesis resulted in zeotypes with various T atoms, ordered mesoporous materials, extra-large pore structures, nanosized, chiral, and 2D zeolites. Today, the goal is the customized synthesis where the zeolite framework type, chemical composition, acidity, hydrophilicity/hydrophobicity, morphology, and size is predetermined. Control of zeolite formation, especially in terms of nuclei generation and growth of the crystals, represents a prerequisite for tailoring materials with target features. This is remarkably evident in harvesting nanosized zeolite crystals where the nucleation should be favored over crystal growth, and the uniformity of viable nuclei distribution is essential. Employing uniform zeolite yielding systems, i.e., clear solutions where the aggregation between the particles is blocked in one way to obtain nanosized zeolites. In the OSDA-containing systems, the formation of nano zeolites is favored by the low alkali cations content and high dilution. On the contrary, in the OSDA-free systems, the concentration of alkali cations is high. Other approaches to prepare nanosized zeolites, and zeolites in general, concern the interzeolite conversions, assembly of pre-formed layers as well as the two-step reaction involving the modification of the precursor. Besides, conducting the reaction under alternative conditions such as in the microwave or microfluidic reactor was found beneficial in attaining nanosized zeolite crystals. The synthetic approaches portfolio enabled the preparation of a range of zeolite materials with size below 100 nm (LTA, LTL, EMT, SOD, FAU, RHO, MFI, BEA, MEL, CHA, FER, ERI, AFX, etc.). Furthermore, materials with nanosized zeolite properties, but particular structural features were reported. In this respect, finned zeolites, bi-dimensional materials, and embryonic zeolites represent special cases of nanosized zeolites. The principal uses of zeolites continue to be the heterogeneous catalysis, adsorption/separation, and ion-exchange. The social demand for environmental protection and the quest for new sustainable energy resources represent the motivation to extend the zeolite uses. The studies of (nano) zeolites’ application aim towards the improvement of the established processes (MTO, NH3SCR, HC, separation/adsorption) and developing new applications such as biomass valorization, methane conversion, CO2 alleviation, medical uses, sensors, food industry, etc. Nanosized zeolites demonstrate promising performance in all mentioned application fields, and in some of them, as medicine and electronics, they are particularly promising. This overview revealed the high research activity in the nano zeolite field and a number of challenges. Only about 20 zeolite structures are obtained in the nanosized form. Thus the number of structures prepared in nanosized form has to be extended to increase nano zeolites’ portfolio. Moreover, many nano zeolites crystallize in a limited chemical composition range, which restricts their application. Hence, a major goal is to prepare nanosized crystals of relevant zeolites with pre-determined physical and chemical properties. Reaching this goal would broaden the prospects of nano zeolite uses and meet modern society’s needs from microporous materials for environmental protection and sustainable energy solutions.

References 1. Thommes, M.; Kaneko, K.; Neimark, A. V.; Olivier, J. P.; Rodriguez-Reinoso, F.; Rouquerol, J.; Sing, K. S. W. Pure Appl. Chem. 2015, 87, 1051–1069. 2. Coombs, D. S.; Alberti, A.; Armbruster, T.; Artioli, G.; Colella, C.; Galli, E.; Grice, J. D.; Liebau, F.; Mandarino, J. A.; Minato, H.; Nickel, E. H.; Passaglia, E.; Peacor, D. R.; Quartieri, S.; Rinaldi, R.; Ross, M.; Sheppard, R. A.; Tillmans, E.; Vezzalini, G. Can. Mineral. 1997, 35, 1571–1606. 3. Hypothetical Zeolites Database. http://www.hypotheticalzeolites.net (accessed Nov 22, 2020). 4. Database of Zeolite Structures. http://www.iza-structure.org/databases/ (accessed Nov 22, 2020). 5. International Zeolite Association. http://www.iza-online.org/default.htm (accessed Nov 22, 2020). 6. Report on Specialty Zeolites Market. https://www.mordorintelligence.com/ (accessed Nov 22, 2020). 7. Report on Synthetic Zeolites Market. https://www.marketsandmarkets.com (accessed Nov 22, 2020). 8. Baerlocher, C., McCusker, L., Olson, D. H., Eds.; Atlas of Zeolite Framework Types, 6th edn.; Elsevier: Amsterdam, 2007. 9. Palcic, A.; Valtchev, V. Appl. Catal. A. Gen. 2020, 606, 117795. 10. Rabo, J. A.; Schoonover, M. W. Appl. Catal. A. Gen. 2001, 222, 261–275. 11. Brändle, M.; Sauer, J. J. Am. Chem. Soc. 1998, 120, 1556–1570. 12. Jones, A. J.; Iglesia, E. ACS Catal. 2015, 5, 5741–5755. 13. Babic, V.; Tang, L.; Qin, Z.; Hafiz, L.; Gilson, J.-P.; Valtchev, V. Adv. Mater. Interfaces 2020, 2000348. https://doi.org/10.1002/admi.202000348. 14. Valtchev, V.; Majano, G.; Mintova, S.; Pérez-Ramírez, J. Chem. Soc. Rev. 2013, 42, 263–290. 15. Cronstedt, A. F. Akad. Handl. Stockholm 1756, 18, 120–130. 16. Tschernich, R. W. Zeolites of the World, Geoscience Press: Phoenix, Arizona, 1992. 17. de St Claire Deville, H. Compt. Rend. Se´ances Acad. Sci. 1862, 54, 324. 18. Cundy, C. S.; Cox, P. A. Micropor. Mesopor. Mater. 2005, 82, 1–78. 19. Wilson, S. T.; Lok, B. M.; Messina, C. A.; Cannan, T. R.; Flanigen, E. M. J. Am. Chem. Soc. 1982, 104, 1146–1147. 20. Wilson, S. T.; Lok, B. M.; Flanigen, E. M. (Union Carbide Co.) Crystalline Metallophosphate Compositions. US Patent 4,310,440, 1982. 21. Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck, J. S. Nature 1992, 359, 710–712. 22. Beck, J. S.; Varuli, J. C.; Roth, W. J.; Leonowicz, M. E.; Kresge, C. T.; Schmitt, K. D.; Chu, C. T. W.; Olson, D. H.; Sheppard, E. W.; McCullen, S. B.; Higgins, J. B.; Schlenker, J. L. J. Am. Chem. Soc. 1992, 114, 10834–10843. 23. Davis, M. E.; Saldarriaga, C.; Montes, C.; Garces, J. M.; Crowder, C. Nature 1988, 331, 698–699. 24. Corma, A.; Díaz-Cabaas, M. J.; Jord, J. L.; Martínez, C.; Moliner, M. Nature 2006, 443, 842–845. 25. Brand, S. K.; Schmidt, J. E.; Deem, M. W.; Daeyaert, F.; Ma, Y; Terasaki, O.; Orazov, M; Davis, M. E., Proc. Natl. Acad. Sci. 2017, 114, 5101-5106.

36 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91.

Synthesis and application of (nano) zeolites Lu, T.; Zhu, L.; Wang, X.; Yan, W.; Shi, W.; Xu, R. Inorg. Chem. Front. 2018, 5, 802–805. Blake, A. J.; Franklin, K. R.; Lowe, B. M. J. Chem. Soc. Dalton Trans. 1988, 2513–2517. Corma, A.; Fornes, V.; Pergher, S. B.; Maesen, T. L. M.; Buglass, J. G. Nature 1998, 396, 353–356. Schoeman, B. J.; Sterte, J.; Otterstedt, J. E. Zeolites 1994, 14, 568–575. Persson, A. E.; Schoeman, B. J.; Sterte, J.; Otterstedt, J. E. Zeolites 1994, 14, 557–567. Beck, L. W.; Davis, M. E. Micropor. Mesopor. Mater. 1998, 22, 107–114. Lai, Z. P.; Bonilla, G.; Diaz, I.; Nery, J. G.; Sujaoti, K.; Amat, M. A.; Kokkoli, E.; Terasaki, O.; Thompson, R. W.; Tsapatsis, M.; Vlachos, D. G. Science 2003, 300, 456–460. Gounder, R.; Davis, M. E. J. Catal. 2013, 308, 176–188. Blasco, T.; Camblor, M. A.; Corma, A.; Esteve, P.; Martinez, A.; Prieto, C.; Valencia, S. Chem. Commun. 1996, 20, 2367–2368. Gounder, R.; Iglesia, E. Angew. Chem. Int. Ed. 2010, 49, 808–811. Pérez-Ramírez, J.; Mitchell, S.; Verboekend, D.; Milina, M.; Michels, N.-L.; Krumeich, F.; Marti, N.; Erdmann, M. ChemCatChem 2011, 3, 1731–1734. Valtchev, V.; Balanzat, E.; Mavrodinova, V.; Diaz, I.; El Fallah, J.; Goupil, J.-M. J. Am. Chem. Soc. 2011, 133, 18950–18956. Zhang, L.; Chen, K.; Chen, B.; White, J. L.; Resasco, D. E. J. Am. Chem. Soc. 2015, 137, 11810–11819. Usui, T.; Liu, Z.; Ibe, S.; Zhu, J.; Anand, C.; Igarashi, H.; Onaya, N.; Sasaki, Y.; Shiramata, Y.; Kusamoto, T.; Wakihara, T. ACS Catal. 2018, 8, 9165–9173. Verheyen, E.; Sree, S. P.; Thomas, K.; Dendooven, J.; De Prins, M.; Vanbutsele, G.; Breynaert, E.; Gilson, J.-P.; Kirschhock, C. E. A.; Detavernier, C.; Martens, J. A. Chem. Commun. 2014, 50, 4610–4612. Gao, F.; Jaber, M.; Bozhilov, K.; Vicente, A.; Fernandez, C.; Valtchev, V. J. Am. Chem. Soc. 2009, 131, 16580–16586. Rey, F.; Simancas, J. Beyond Nitrogen OSDAs. In Insights into the Chemistry of Organic Structure-Directing Agents in the Synthesis of Zeolitic Materials. Structure and Bonding; Gómez-Hortigüela, L., Ed.; vol. 175; Springer: Cham, 2017; pp 103–138. Chapter 4. Zones, S. I.; Hwang, S.-J. Micropor. Mesopor. Mater. 2011, 146, 48–56. Moliner, M.; Corma, A. Chem. Mater. 2012, 24, 4371–4375. Morris, R. E.; Cejka, J. Nat. Chem. 2015, 7, 381–388. Corma, A.; Díaz-Cabanas, M. J.; Moliner, M.; Martínez, C. J. Catal. 2006, 241, 312–318. Liu, Z.; Zhu, J.; Wakihara, T.; Okubo, T. Inorg. Chem. Front. 2019, 6, 14–31. Qin, Z.; Pinard, L.; Benghalem, M. A.; Daou, T. J.; Melinte, G.; Ersen, O.; Asahina, S.; Gilson, J.-P.; Valtchev, V. Chem. Mater. 2019, 31, 4639–4648. Carceller, J. M.; Martínez Galán, J. P.; Monti, R.; Bassan, J. C.; Filice, M.; Yu, J.; Climent, M. J.; Iborra, S.; Corma, A. ChemCatChem 2020, 12, 4502–4511.  Palcic, A.; Szyja, B. M.; Micetic, M.; Cendak, T.; Akouche, M.; Juraic, K.; Cargonja, M.; Mekterovic, D.; Vusak, V.; Valtchev, V. Inorg. Chem. Front. 2019, 6, 2279–2290. Breck, D. W. Zeolite Molecular Sieves, Structure, Chemistry and Use, Wiley: New York, 1974. Cooper, E. R.; Andrews, C. D.; Wheatley, P. S.; Webb, P. B.; Wormald, P.; Morris, R. E. Nature 2004, 430, 1012–1016. Bibby, D. M.; Dale, M. P. Nature 1985, 317, 157–158. Jianquan, L.; Jinxiang, D.; Guanghuan, L.; Shunquan, G.; Feng, W. J. Chem. Soc. Chem. Commun. 1993, 659–660. Huo, Q.; Feng, S.; Xu, R. J. Chem. Soc. Chem. Commun. 1988, 1486–1487. Qin, Z.; Lakiss, L.; Tosheva, L.; Gilson, J.-P.; Vicente, A.; Fernandez, C.; Valtchev, V. Adv. Funct. Mater. 2014, 24, 257–264. Camblor, M. A.; Corma, A.; Valencia, S. Chem. Commun. 1996, 2365–2366. Zones, S. I.; Hwang, S.-J.; Elomari, S.; Ogino, I.; Davis, M. E.; Burton, A. W. C. R. Chim. 2005, 8, 267–282. Estermann, M.; McCusker, L. B.; Baerlocher, C.; Merrouche, A.; Kessler, H. Nature 1991, 352, 320–323. Corma, A.; Diaz-Cabanas, M. J.; Martinez-Triguero, J.; Rey, F.; Rius, J. A. Nature 2002, 418, 514–517. Kessler, H.; Patarin, J.; Schott-Darie, C. Stud. Surf. Sci. Catal. 1994, 85, 75–113. Zones, S. I.; Davis, M. Curr. Opin. Solid State Mater. Sci. 1996, 1, 107–117. Wright, P. A.; Lozinska, M. Structural Chemistry and Properties of Zeolites. In Zeolites and Ordered Porous Solids: Fundamentals and Applications; Martínez, C., PérezPariente, J., Eds., Universitat Politenica de Valencia: Valencia, 2011; pp 1–36. Chapter 1. Lozinska, M. M.; Miller, D. N.; Brandani, S.; Wright, P. A. J. Mater. Chem. A 2020, 8, 3280–3292. Lee, H.; Zones, S. I.; Davis, M. E. Nature 2003, 425, 385–388. Szostak, R. Handbook of Molecular Sieves, Van Nostrand Reinhold: New York, 1992. Lawton, S. L.; Rohrbaugh, W. J. Science 1990, 247, 1319–1322. Sastre, G.; Leiva, S.; Sabater, M. J.; Gimenez, I.; Rey, F.; Valencia, S.; Corma, A. J. Phys. Chem. B 2003, 107, 5432–5440. Freyhardt, C. C.; Tsapatsis, M.; Lobo, R. F.; Balkus, K. J., Jr.; Davis, M. E. Nature 1996, 381, 295–298. Gallego, E. M.; Portilla, M. T.; Paris, C.; León-Escamilla, A.; Boronat, M.; Moliner, M.; Corma, A. Science 2017, 355, 1051–1054. Lu, P.; Gómez-Hortigüela, L.; Camblor, M. A. Dalton Trans. 2018, 47, 7498–7504. Gies, H.; Marler, B. Zeolites 1992, 12, 42–49. Kubota, Y.; Helmkamp, M. M.; Zones, S. I.; Davis, M. E. Micropor. Mater. 1996, 6, 213–229. Testa, F.; Szostak, R.; Chiappetta, R.; Aiello, R.; Fonseca, A.; Nagy, J. B. Zeolites 1997, 18, 106–114. Kim, S. H.; Park, M. B.; Min, H.-K.; Hong, S. B. Micropor. Mesopor. Mater. 2009, 123, 160–168. Paris, C.; Martín, N.; Martínez-Triguero, J.; Moliner, M.; Corma, A. New J. Chem. 2016, 40, 4140–4145. Di Iorio, J. R.; Gounder, R. Chem. Mater. 2016, 28, 2236–2247. Dang, L. V.; Le, S. T.; Lobo, R. F.; Pham, T. D. J. Porous Mater. 2020, 27, 1481–1489. Pinar, A. B.; Gómez-Hortigüela, L.; McCusker, L. B.; Pérez-Pariente, J. Chem. Mater. 2013, 25, 3654–3661. Valtchev, V.; Tosheva, L. Chem. Rev. 2013, 113, 6734–6760. Subotic, B.; Bronic, J.; Antonic-Jelic, T. Theoretical and Practical Aspects of Zeolite Nucleation. In Ordered Porous Materials; Valtchev, V., Mintova, S., Tsapatsis, M., Eds., Elsevier: Amsterdam, 2009; pp 127–185. Chapter 6. Muraoka, K.; Sada, Y.; Miyazaki, D.; Chaikittisilp, W.; Okubo, T. Nat. Commun. 2019, 10, 4459. Knight, C. T. G.; Balec, R. J.; Kinrade, S. D. Angew. Chem. Int. Ed. 2007, 46, 8148–8152. Harvey, G.; Dent Glaser, L. S. Structure and Properties of Aluminosilicate Solutions and Gels. In Ocelli, M. L., Robson, H. E., EdsZeolite Synthesis, Symposium Series 398, American Chemical Society: Washington, DC, 1989; pp 49–65. Bell, A. T. Applications of NMR Spectroscopy to the Study of Zeolite Synthesis. In Ocelli, M. L., Robson, H. E., EdsZeolite Synthesis, Symposium Series 398, American Chemical Society: Washington, DC, 1989; pp 66–82. Epping, J. D.; Chmelka, B. F. Curr. Opin. Colloid Interface Sci. 2006, 11, 81–117. Petrovic, I.; Navrotsky, A.; Davis, M. E.; Zones, S. I. Chem. Mater. 1993, 5, 1805–1813. Smaihi, M.; Barida, O.; Valtchev, V. Eur. J. Inorg. Chem. 2003, 4370–4377. Navrotsky, A.; Trofymluk, O.; Levchenko, A. A. Chem. Rev. 2009, 109, 3885–3902. Breck, D. W. J. Chem. Educ. 1964, 41, 678–689. Zhdanov, S. P. Some Problems of Zeolite Crystallization. In Molecular Sieve Zeolites – I; Flanigen, E. M., Sand, L. B., Eds. Advances in Chemistry Series 101, American Chemical Society: Washington, DC, 1971; pp 20–43.

Synthesis and application of (nano) zeolites 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157.

37

Subotic, B.; Graovac, A. Stud. Surf. Sci. Catal. 1985, 24, 199–206. Guth, J.-L.; Kessler, H. In Catalysis and Zeolites; Weitkamp, J., Puppe, L., Eds., Springer-Verlag: Berlin, 1999; pp 1–52. Chapter 1. Brunner, G. O. Zeolites 1992, 12, 428–430. Thompson, R. W.; Ying, T.-C. Zeolites 1984, 4, 353–360. Burkett, S. L.; Davis, M. E. J. Phys. Chem. 1994, 98, 4647–4653. de Moor, P.-P. E. A.; Beelen, T. P. M.; Komanschek, B. U.; Diat, O.; van Santen, R. A. J. Phys. Chem. B 1997, 101, 11077–11086. Auerbach, S. M.; Ford, M. H.; Monson, P. A. Curr. Opin. Colloid Interface Sci. 2005, 10, 220–225. Serrano, D. P.; van Grieken, R. J. Mater. Chem. 2001, 11, 2391–2407. Palcic, A.; Subotic, B.; Valtchev, V.; Bronic, J. CrstEngComm 2013, 15, 5784–5791. Wakihara, T.; Kohara, S.; Sankar, G.; Saito, S.; Sanchez-Sanchez, M.; Overweg, A. R.; Fan, W.; Ogura, M.; Okubo, T. Phys. Chem. Chem. Phys. 2006, 8, 224–227. Melinte, G.; Georgieva, V.; Springuel-Huet, M.-A.; Nossov, A.; Ersen, O.; Guenneau, F.; Gedeon, A.; Palcic, A.; Bozhilov, K. N.; Pham-Huu, C.; Qiu, S.; Mintova, S.; Valtchev, V. Chem. A Eur. J. 2015, 21, 18316–18327. Cundy, C. S.; Lowe, B. M.; Sinclair, D. M. J. Cryst. Growth 1990, 100, 189–202. Gonthier, S.; Gora, L.; Güray, I.; Thompson, R. W. Zeolites 1993, 13, 414–418. Nikolakis, V.; Vlachos, D. G.; Tsapatsis, M. Micropor. Mesopor. Mater. 1998, 21, 34–37. Golemme, G.; Nastro, A.; Nagy, J. B.; Subotic, B.; Crea, F.; Aiello, R. Zeolites 1991, 11, 776–783. Auerbach, S. M.; Fan, W.; Monson, P. A. Int. Rev. Phys. Chem. 2014, 34, 35–70. Valtchev, V. P.; Bozhilov, K. N. J. Am. Chem. Soc. 2005, 127, 16171–16177. Valtchev, V. P.; Bozhilov, K. N. J. Phys. Chem. B 2004, 108, 15587–15598. Itani, L.; Liu, Y.; Zhang, W.; Bozhilov, K. N.; Delmotte, L.; Valtchev, V. J. Am. Chem. Soc. 2009, 131, 10127–10139. Maldonado, M.; Oleksiak, M. D.; Chinta, S.; Rimer, J. D. J. Am. Chem. Soc. 2013, 135, 2641–2652. Palcic, A.; Bosnar, S.; Bosnar, D.; Kontrec, J.; Bronic, J. Acta Chim. Slov. 2015, 62, 130–135. Eilertsen, E. A.; Haouas, M.; Pinar, A. B.; Hould, N. D.; Lobo, R. F.; Lillerud, K. P.; Taulelle, F. Chem. Mater. 2012, 24, 571–578. Haouas, M.; Lakiss, L.; Martineau, C.; El Fallah, J.; Valtchev, V.; Taulelle, F. Micropor. Mesopor. Mater. 2014, 198, 35–44. Pelster, S. A.; Kalamajka, R.; Schrader, W.; Schüth, F. Angew. Chem. Int. Ed. 2007, 46, 2299–2302. Castro, M.; Haouas, M.; Taulelle, F.; Lim, I.; Breynaert, E.; Brabants, G.; Kirschhock, C. E. A.; Schmidt, W. Micropor. Mesopor. Mater. 2014, 189, 158–162. Kragten, D. D.; Fedeyko, J. M.; Sawant, K. R.; Rimer, J. D.; Vlachos, D. G.; Lobo, R. F. J. Phys. Chem. B 2003, 107, 10006–10016. Dokter, W. H.; van Garderen, H. F.; Beleen, T. P. M.; van Santen, R. A.; Bras, V. Angew. Chem. Int. Ed. 1995, 34, 73–75. Mintova, S.; Valtchev, V.; Bein, T. Colloids Surf. A Physicochem. Eng. Asp. 2003, 217, 153–157. Kirschhock, C. E. A.; Buschmann, V.; Kremer, S.; Ravishankar, R.; Houssin, C. J. Y.; Mojet, B. L.; van Santen, R. A.; Grobet, P. J.; Jacobs, P. A.; Martens, J. A. Angew. Chem. Int. Ed. 2001, 40, 2637–2640. Davis, T. M.; Drews, T. O.; Ramanan, H.; He, C.; Dong, J.; Schnablegger, H.; Katsoulakis, M. A.; Kokkoli, E.; Mccormick, A. V.; Lee, R.; Tsapatsis, M. Nat. Mater. 2006, 5, 400–408. Aerts, A.; Haouas, M.; Caremans, T. P.; Follens, L. R. A.; van Erp, T. S.; Taulelle, F.; Vermant, J.; Martens, J. A.; Kirschhock, C. E. A. Chem. A Eur. J. 2010, 16, 2764–2774. Cheng, C.-H.; Shantz, D. F. J. Phys. Chem. B 2005, 109, 13912–13920. Rivas-Cardona, A.; Chovanetz, M.; Shantz, D. F. Micropor. Mesopor. Mater. 2012, 155, 56–64.  z, V.; Smrecki, V.; Palcic, A.; Subotic, B. J. Phys. Chem. C 2018, 122, 9441–9494. Bosnar, S.; Antonic Jelic, T.; Bronic, J.; Dutour Sikiric, M.; Segota, S.; Cade Bosnar, S.; Dutour Sikiric, M.; Smrecki, V.; Bronic, J.; Segota, S.; Strasser, V.; Antonic Jelic, T.; Palcic, A.; Subotic, B. Inorg. Chem. Front. 2019, 6, 1639–1653. Nielsen, A. E. Croat. Chem. Acta 1970, 42, 319–333. Nielsen, A. E. Croat. Chem. Acta 1987, 60, 531–539.  Cubillas, P.; Anderson, M. W. Synthesis Mechanism: Crystal Growth and Nucleation. In Zeolites and Catalysis, Synthesis, Reactions and Applications; Cejka, J., Corma, A., Zones, S., Eds.; vol. 1; Wiley-VCH: Weinheim, 2011; pp 1–55. Chapter 1. Kumar, M.; Luo, H.; Román-Leshkov, Y.; Rimer, J. D. J. Am. Chem. Soc. 2015, 137, 13007–13017. Kumar, M.; Choudhary, M. K.; Rimer, J. D. Nat. Commun. 2018, 9, 2129. Karwacki, L.; Stavitski, E.; Kox, M. H. F.; Kornatowski, J.; Weckhuysen, B. M. Angew. Chem. Int. Ed. 2007, 46, 7228–7231. Qin, Z.; Melinte, G.; Gilson, J.-P.; Jaber, M.; Bozhilov, K.; Boullay, P.; Mintova, S.; Ersen, O.; Valtchev, V. Angew. Chem. Int. Ed. 2016, 55, 15049–15052. Majano, G.; Darwiche, A.; Mintova, S.; Valtchev, V. Ind. Eng. Chem. Res. 2009, 48, 7084–7091. Tang, L.; Haw, K.-G.; He, P.; Fang, Q.; Qiu, S.; Valtchev, V. Inorg. Chem. Front. 2019, 6, 3097–3103. Kasahara, S.; Itabashi, K.; Igawa, K. Stud. Surf. Sci. Catal. 1986, 28, 185–192. Zhang, X.; Liu, H.; Yeung, K. Mater. Chem. Phys. 2006, 96, 42–50. Itabashi, K.; Kamimura, Y.; Iyoki, K.; Shimojima, A.; Okubo, T. J. Am. Chem. Soc. 2012, 134, 11542–11549. Ren, N.; Subotic, B.; Bronic, J. Crystallization of Sub-Micrometer Sized ZSM-5 Zeolites in SDA-Free Systems. In Advances in Crystallization Processes; Mastai, Y., Ed., InTech: Rijeka, 2012; pp 259–284. Chapter 10. Chokkalingam, A.; Iyoki, K.; Hoshikawa, N.; Onozuka, H.; Chaikittisilp, W.; Tsutsuminai, S.; Takewaki, T.; Wakihara, T.; Okubo, T. Annu. Rev. Chem. Biomol. Eng. 2020. https://doi.org/10.1039/D0RE00309C. Georgieva, V.; Vicente, A.; Fernandez, C.; Retoux, R.; Palcic, A.; Valtchev, V.; Mintova, S. Cryst. Growth Des. 2015, 15, 1898–1906. Alberti, A.; Martucci, A. Micropor. Mesopor. Mater. 2011, 141, 192–198. Goel, S.; Zones, S. I.; Iglesia, E. Chem. Mater. 2015, 27, 2056–2066. Tang, L.; Haw, K.-G.; Zhang, Y.; Fang, Q.; Qiu, S.; Valtchev, V. Micropor. Mesopor. Mater. 2019, 280, 306–314. Subotic, B.; Skrtic, D.; Smit, I.; Sekovanic, L. J. Cryst. Growth 1980, 50, 498–508. Subotic, B.; Sekovanic, L. J. Cryst. Growth 1986, 75, 561–572. Khodabandeh, S.; Davis, M. E. Micropor. Mater. 1997, 9, 149–160. Jon, H.; Nakahata, K.; Lu, B.; Oumi, Y.; Sano, T. Micropor. Mesopor. Mater. 2006, 96, 72–78. Jon, H.; Ikawa, N.; Oumi, Y.; Sano, T. Chem. Mater. 2008, 20, 4135–4141. Qin, W.; Jain, R.; Robles Hernández, F. C.; Rimer, J. D. Chem. A Eur. J. 2019, 25, 5893–5898. Roth, W. J.; Kresge, C. T.; Vartuli, J. C.; Leonowicz, M. E.; Fung, A. S.; McCullen, S. B. Stud. Surf. Sci. Catal. 1995, 94, 301–308. Roth, W. J.; Nachtigall, P.; Morris, R. E.; Wheatley, P. S.; Seymour, V. R.; Ashbrook, S. E.; Chlubná, P.; Grajciar, L.; Polozij, M.; Zukal, A.; Shvets, O.;  Cejka, J. Nat. Chem. 2013, 5, 628–633. Corma, A.; Diaz, U.; Domine, M. E.; Fornés, V. Angew. Chem. Int. Ed. 2000, 39, 1499–1501. Na, K.; Choi, M.; Park, W.; Sakamoto, Y.; Terasaki, O.; Ryoo, R. J. Am. Chem. Soc. 2010, 132, 4169–4177. Kresge, C. T.; Roth, W. J. (Exxon Mobile Co.) Crystalline Oxide Material. US Patent 5,266,541, 1993. Corma, A.; Fornés, V.; Díaz, U. Chem. Commun. 2001, 2642–2643.  Eliásová, P.; Opanasenko, M.; Wheatley, P. S.; Shamzhy, M.; Mazur, M.; Nachtigall, P.; Roth, W. J.; Morris, R. E.; Cejka, J. Chem. Soc. Rev. 2015, 44, 7177–7206.

38

Synthesis and application of (nano) zeolites

 158. Chlubná-Eliásová, P.; Tian, Y.; Pinar, A. B.; Kubu, M.; Cejka, J.; Morris, R. E. Angew. Chem. Int. Ed. 2014, 53, 7048–7052. 159. Kasneryk, V.; Shamzhy, M.; Opanasenko, M.; Wheatley, P. S.; Morris, S. A.; Russell, S. E.; Mayoral, A.; Trachta, M.;  Cejka, J.; Morris, R. E. Angew. Chem. Int. Ed. 2017, 56, 4324–4327.  160. Mazur, M.; Wheatley, P. S.; Navarro, M.; Roth, W. J.; Polozij, M.; Mayoral, A.; Eliásová, P.; Nachtigall, P.; Cejka, J.; Morris, R. E. Nat. Chem. 2016, 8, 58–62.  161. Mazur, M.; Kasneryk, V.; Prech, J.; Brivio, F.; Ochoa-Hernández, C.; Mayoral, A.; KubŮ, M.; Cejka, J. Inorg. Chem. Front. 2018, 5, 2746–2755.  162. Zhou, Y.; Kadam, S. A.; Shamzhy, M.; Cejka, J.; Opanasenko, M. ACS Catal. 2019, 9, 5136–5146. 163. Tosheva, L.; Valtchev, V. P. Chem. Mater. 2005, 17, 2494–2513. 164. Schoeman, B. J.; Sterte, J.; Otterstedt, J.-E. Zeolites 1994, 14, 110–116. 165. Mintova, S.; Jaber, M.; Valtchev, V. Chem. Soc. Rev. 2015, 44, 7207–7233. 166. Mintova, S.; Olson, N.; Valtchev, V.; Bein, T. Science 1999, 283, 958–960. 167. Van Grieken, R.; Sotelo, J. L.; Menéndez, J. M.; Melero, J. A. Micropor. Mesopor. Mater. 2000, 39, 135–147. 168. Smaïhi, M.; Gavilan, E.; Durand, J.-O.; Valtchev, V. P. J. Mater. Chem. 2004, 14, 1347–1351. 169. Valtchev, V. P.; Tosheva, L.; Bozhilov, K. N. Langmuir 2005, 21, 10724–10729. 170. Li, R.; Linares, N.; Sutjianto, J. G.; Chawla, A.; Garcia-Martinez, J.; Rimer, J. D. Angew. Chem. Int. Ed. 2018, 57, 11283–11288. 171. Morales-Pacheco, P.; Alvarez, F.; Bucio, L.; Domínguez, J. M. J. Phys. Chem. C 2009, 113, 2247–2255. 172. Awala, H.; Gilson, J.-P.; Retoux, R.; Boullay, P.; Goupil, J.-M.; Valtchev, V.; Mintova, S. Nat. Mater. 2015, 14, 447–451. 173. Itani, L.; Bozhilov, K. N.; Clet, G.; Delmotte, L.; Valtchev, V. Chem. A Eur. J. 2011, 17, 2199–2210. 174. Ng, E.-P.; Chateigner, D.; Bein, T.; Valtchev, V.; Mintova, S. Science 2012, 335, 70–73. 175. Grand, J.; Barrier, N.; Debost, M.; Clatworthy, E. B.; Laine, F.; Boullay, P.; Nesterenko, N.; Dath, J.-P.; Gilson, J.-P.; Mintova, S. Chem. Mater. 2020, 32 (14), 5985–5993. 176. Zhang, C.; Wu, Q.; Lei, C.; Han, S.; Zhu, Q.; Maurer, S.; Dai, D.; Parvulescu, A.-N.; Müller, U.; Meng, X.; Xiao, F.-S. J. Mater. Chem. A 2018, 6, 21156–21161. 177. Iwakai, K.; Tago, T.; Konno, H.; Nakasaka, Y.; Masuda, T. Micropor. Mesopor. Mater. 2011, 141, 167–174. 178. Zhang, Q.; Xiang, S.; Zhang, Q.; Wang, B.; Mayoral, A.; Liu, W.; Wang, Y.; Liu, Y.; Shi, J.; Yang, G.; Luo, J.; Chen, X.; Terasaki, O.; Gilson, J.-P.; Yu, J. Chem. Mater. 2020, 32, 751–758. 179. Zhang, Q.; Chen, G.; Wang, Y.; Chen, M.; Guo, G.; Shi, J.; Luo, J.; Yu, J. Chem. Mater. 2018, 30, 2750–2758. 180. Catizzone, E.; Van Daele, S.; Bianco, M.; Di Michele, A.; Aloise, A.; Migliori, M.; Valtchev, V.; Giordano, G. Appl Catal B 2019, 243, 273–282. 181. Robson, H., Ed.; Verified Synthesis of Zeolitic Materials, Elsevier: Amsterdam, 2001, 2nd revised edition. 182. Takata, T.; Tsunoji, N.; Takamitsu, Y.; Sadakane, M.; Sano, T. Micropor. Mesopor. Mater. 2016, 225, 524–533. 183. Li, Y.; Zhang, Y.; Lan, A.; Bian, H.; Liu, R.; Li, X.; Han, P.; Dou, T. Micropor. Mesopor. Mater. 2019, 279, 1–9. 184. Khan, N. A.; Yoo, D. K.; Bhadra, B. N.; Jun, J. W.; Kim, T.-W.; Kim, C.-U.; Jhung, S. H. Chem. Eng. J. 2019, 377, 119546. 185. Li, D.; Chen, Y.; Hu, J.; Deng, B.; Cheng, X.; Zhang, Y. Appl Catal B 2020, 270, 118881. 186. Goto, I.; Itakura, M.; Shibata, S.; Honda, K.; Ide, Y.; Sadakane, M.; Sano, T. Micropor. Mesopor. Mater. 2012, 158, 117–122. 187. Martín, N.; Li, Z.; Martínez-Triguero, J.; Yu, J.; Moliner, M.; Corma, A. Chem. Commun. 2016, 52, 6072–6075. 188. Tanigawa, T.; Tsunoji, N.; Sadakane, M.; Sano, T. Micropor. Mesopor. Mater. 2019, 277, 115–123. 189. Xie, D.; Zones, S. I.; Saxton, R. J. (Chevron Co.) Small Crystal LTL Framework Type Zeolites, WO2016/032565, 2016. 190. Funase, N.; Tanigawa, T.; Yamasaki, Y.; Tsunoji, N.; Sadakane, M.; Sano, T. J. Mater. Chem. A 2017, 5, 19245–19254. 191. Itakura, M.; Oumi, Y.; Sadakane, M.; Sano, T. Mater. Res. Bull. 2010, 45, 646–650. 192. Honda, K.; Yashiki, A.; Sadakane, M.; Sano, T. Micropor. Mesopor. Mater. 2014, 196, 254–260. 193. Van Tendeloo, L.; Gobechiya, E.; Breynaert, E.; Martens, J. A.; Kirschhock, C. E. A. Chem. Commun. 2013, 49, 11737–11739. 194. Martín, N.; Paris, C.; Vennestrøm, P. N. R.; Thøgersen, J. R.; Moliner, M.; Corma, A. Appl Catal B 2017, 217, 125–136. 195. Zhu, J.; Liu, Z.; Sukenaga, S.; Ando, M.; Shibata, H.; Okubo, T.; Wakihara, T. Micropor. Mesopor. Mater. 2018, 268, 1–8. 196. Ren, N.; Yang, Z.-J.; Lv, X.-C.; Shi, J.; Zhang, Y.-H.; Tang, Y. Micropor. Mesopor. Mater. 2010, 131, 103–114. 197. Kalvachev, Y.; Jaber, M.; Mavrodinova, V.; Dimitrov, L.; Nihtianova, D.; Valtchev, V. Micropor. Mesopor. Mater. 2013, 177, 127–134. 198. Wakihara, T.; Ichikawa, R.; Tatami, J.; Endo, A.; Yoshida, K.; Sasaki, Y.; Komeya, K.; Meguro, T. Cryst. Growth Des. 2011, 11, 955–958. 199. Wakihara, T.; Sato, K.; Inagaki, S.; Tatami, J.; Komeya, K.; Meguro, T.; Kubota, Y. ACS Appl. Mater. Interfaces 2010, 2, 2715–2718. 200. Anand, C.; Yamaguchi, Y.; Liu, Z.; Ibe, S.; Elangovan, S. P.; Ishii, T.; Ishikawa, T.; Endo, A.; Okubo, T.; Wakihara, T. Sci. Rep. 2016, 6, 29210. 201. Zeng, L.; Yu, Z.; Sun, Z.; Han, Y.; Xu, Y.; Wu, J.; Liang, Z.; Wang, Z. Micropor. Mesopor. Mater. 2020, 293, 109789. 202. Inagaki, S.; Nakatsuyama, K.; Saka, Y.; Kikuchi, E.; Kohara, S.; Matsukata, M. J. Phys. Chem. C 2007, 111, 10285–10293. 203. Guo, H.; Zhao, L.; Martineau-Corcos, C.; Fayon, F.; Viger-Gravel, J.; Awala, H.; Boullay, P.; Grand, J.; Vicente, A.; Gilson, J.-P.; Mintova, S. Adv. Mater. Interfaces 2020, 2000634. https://doi.org/10.1002/admi.202000634. 204. Dai, H.; Shen, Y.; Yang, T.; Lee, C.; Fu, D.; Agarwal, A.; Le, T. T.; Tsapatsis, M.; Palmer, J. C.; Weckhuysen, B. M.; Dauenhauer, P. J.; Zou, X.; Rimer, J. D. Nat. Mater. 2020, 19, 1074–1080. 205. Seo, Y.; Cho, K.; Jung, Y.; Ryoo, R. ACS Catal. 2013, 3, 713–720. 206. Jeon, M. Y.; Kim, D.; Kumar, P.; Lee, P. S.; Rangnekar, N.; Bai, P.; Shete, M.; Elyassi, B.; Lee, H. S.; Narasimharao, K.; Basahel, S. N.; Al-Thabaiti, S.; Xu, W.; Cho, H. J.; Fetisov, E. O.; Thyagarajan, R.; DeJaco, R. F.; Fan, W.; Mkhoyan, K. A.; Siepmann, J. I.; Tsapatsis, M. Nature 2017, 543, 690–694. 207. Choi, M.; Na, K.; Kim, J.; Sakamoto, Y.; Terasaki, O.; Ryoo, R. Nature 2009, 461, 246–249. 208. Park, W.; Yu, D.; Na, K.; Jelfs, K. E.; Slater, B.; Sakamoto, Y.; Ryoo, R. Chem. Mater. 2011, 23, 5131–5137. 209. Zhang, X.; Liu, D.; Xu, D.; Asahina, S.; Cychosz, K. A.; Agrawal, K. V.; Al Wahedi, Y.; Bhan, A.; Al Hashimi, S.; Terasaki, O.; Thommes, M.; Tsapatsis, M. Science 2012, 336, 1684–1687. 210. Haw, K.-G.; Goupil, J.-M.; Gilson, J.-P.; Nesterenko, N.; Minoux, D.; Dath, J.-P.; Valtchev, V. New J. Chem. 2016, 40, 4307–4313. 211. Haw, K.-G.; Goupil, J.-M.; Gilson, J.-P.; Valtchev, V.; Nesterenko, N.; Minoux, D.; Dath J.-P. Catalyst Compositions Comprising Ultra-Small Size Molecular Sieves Crystals Deposited on a Porous Material. WO2015001004, 2015. 212. Akouche, M.; Gilson, J.-P.; Nesterenko, N.; Moldovan, S.; Chateigner, D.; Siblani, H. E.; Minoux, D.; Dath, J.-P.; Valtchev, V. Chem. Mater. 2020, 32, 2123–2132. 213. Haw, K.-G.; Gilson, J.-P.; Nesterenko, N.; Akouche, M.; El Siblani, H.; Goupil, J.-M.; Rigaud, B.; Minoux, D.; Dath, J.-P.; Valtchev, V. ACS Catal. 2018, 8, 8199–8212. 214. Tompsett, G. A.; Conner, W. C.; Yngvesson, K. S. ChemPhysChem 2006, 7, 296–319. 215. Cheng, Z.; Han, S.; Sun, W.; Qin, Q. Mater. Lett. 2013, 95, 193–196. 216. Nasser, G. A.; Muraza, O.; Nishitoba, T.; Malaibari, Z.; Yamani, Z. H.; Al-Shammari, T. K.; Yokoi, T. Ind. Eng. Chem. Res. 2019, 58, 60–68. 217. Jawor, A.; Jeong, B.-H.; Hoek, E. M. V. J. Nanopart. Res. 2009, 11, 1795. 218. Hu, Y.; Liu, C.; Zhang, Y.; Ren, N.; Tang, Y. Micropor. Mesopor. Mater. 2009, 119, 306–314. 219. van Heyden, H.; Mintova, S.; Bein, T. J. Mater. Chem. 2006, 16, 514–518. 220. Álvaro-Muñoz, T.; Sastre, E.; Márquez-Álvarez, C. Cat. Sci. Technol. 2014, 4, 4330–4339. 221. Lin, S.; Li, J.; Sharma, R. P.; Yu, J.; Xu, R. Top. Catal. 2010, 53, 1304–1310. 222. Ng, E.-P.; Awala, H.; Komaty, S.; Mintova, S. Micropor. Mesopor. Mater. 2019, 280, 256–263. 223. Pan, Y.; Yao, J.; Zhang, L.; Xu, N. Ind. Eng. Chem. Res. 2009, 48, 8471–8477.

Synthesis and application of (nano) zeolites 224. 225. 226. 227. 228. 229. 230. 231. 232. 233. 234. 235. 236. 237. 238. 239. 240. 241. 242. 243. 244. 245. 246. 247. 248. 249. 250. 251. 252. 253. 254. 255. 256. 257. 258. 259. 260. 261. 262. 263. 264. 265. 266. 267. 268. 269. 270. 271. 272. 273. 274. 275. 276. 277. 278. 279. 280. 281. 282. 283. 284. 285. 286. 287. 288.

39

Pan, Y.; Ju, M.; Yao, J.; Zhang, L.; Xu, N. Chem. Commun. 2009, 7233–7235. Hoang, P. H.; Yoon, K. B.; Kim, D. P. RSC Adv. 2012, 2, 5323–5328. Askari, S.; Alipour, S. M.; Halladj, R.; Farahani, M. H. D. A. J. Porous Mater. 2013, 20, 285–302. Ng, E.-P.; Awala, H.; Ghoy, J.-P.; Vicente, A.; Ling, T. C.; Ng, Y. H.; Mintova, S.; Adam, F. Mater. Chem. Phys. 2015, 159, 38–45. Pajaie, H. S.; Taghizadeh, M. J. Ind. Eng. Chem. 2015, 24, 59–70. Welton, T. Chem. Rev. 1999, 99, 2071–2084. Ng, E. P.; Itani, L.; Sekhon, S. S.; Mintova, S. Chem. A Eur. J. 2010, 16, 12890–12897. Madsen, C.; Jacobsen, C. J. H. Chem. Commun. 1999, 673–674. Wang, H. T.; Holmberg, B. A.; Yan, Y. S. J. Am. Chem. Soc. 2003, 125, 9928–9929. Wang, B.; Ma, H. Z.; Shi, Q. Z. Chin. Chem. Lett. 2002, 13, 385–388. Sharma, P.; Han, M. H.; Cho, C. H. J. Colloid Interface Sci. 2014, 422, 45–53. Report on Zeolite Market. https://www.fortunebusinessinsights.com (accessed November 10 2020). Vogt, E. T. C.; Weckhuysen, B. M. Chem. Soc. Rev. 2015, 44, 7342–7370. Vermeiren, W.; Gilson, J.-P. Top. Catal. 2009, 52, 1131–1161. Tamm, P. W.; Mohr, D. H.; Wilson, C. R. Surf. Sci. Catal. 1987, 38, 335–353. Degnan, T. F. Top. Catal. 2000, 13, 349–356. Vogt, E. T. C.; Whiting, G. T.; Chowdhury, A. D.; Weckhuysen, B. M. Zeolites and Zeotypes for Oil and Gas Conversion. In Advances in Catalysis; Jentoft, F. C., Ed.; vol. 58; Academic Press, 2015; pp 143–314. Chapter 2. Tian, P.; Wei, Y.; Ye, M.; Liu, Z. ACS Catal. 2015, 5, 1922–1938. Perego, C.; Ingallina, P. Catal. Today 2002, 73, 3–22. Barthe, P.; Chaugny, M.; Roudier, S.; Delgado Sancho, L. Best Available Techniques (BAT) Reference Document for the Refining of Mineral Oil and Gas, European Commission, 2015; pp 158–159. Industrial Emissions Directive 2010/75/EU. Yilmaz, B.; Müller, U. Top. Catal. 2009, 52, 888–895. Lei, C.; Dong, Z.; Martínez, C.; Martínez-Triguero, J.; Chen, W.; Wu, Q.; Meng, X.; Parvulescu, A.-N.; De Baerdemaeker, T.; Müller, U.; Zheng, A.; Ma, Y.; Zhang, W.; Yokoi, T.; Marler, B.; De Vos, D. E.; Kolb, U.; Corma, A.; Xiao, F.-S. Angew. Chem. Int. Ed. 2020, 59, 15649–15655. Ferella, F.; Leone, S.; Innocenzi, V.; De Michelis, I.; Taglieri, G.; Gallucci, K. J. Clean. Prod. 2019, 230, 910–926. Xu, J. S.; Zhang, T.; Zhang, J. Sci. Rep. 2020, 10, 12730. Lu, G. J.; Lu, X. Y.; Liu, P. J. Ind. Eng. Chem. 2020, 92, 236–242. Vuong, G.-T.; Hoang, V.-T.; Nguyen, D.-T.; Do, T.-O. Appl. Catal. A 2010, 382, 231–239. Koempel, W. Stud. Surf. Sci. Catal. 2007, 167, 261–267. Olsbye, U.; Svelle, S.; Lillerud, K. P.; Wei, Z. H.; Chen, Y. Y.; Li, J. F.; Wang, J. G.; Fan, W. B. Chem. Soc. Rev. 2015, 44, 7155–7176. Ferri, P.; Li, C.; Millán, R.; Martínez-Triguero, J.; Moliner, M.; Boronat, M.; Corma, A. Angew. Chem. Int. Ed. 2020, 59, 19708–19715. Palcic, A.; Ordomsky, V. V.; Qin, Z.; Georgieva, V.; Valtchev, V. Chem. A Eur. J. 2018, 24, 13136–13149. Gallego, E. M.; Li, C.; Paris, C.; Martín, N.; Martínez-Triguero, J.; Boronat, M.; Moliner, M.; Corma, A. Chem. A Eur. J. 2018, 24, 14631–14635. Klingstedt, F.; Arve, K.; Eraenen, K.; Murzin, D. Y. ChemInform 2006, 37, 273–282. Zhang, R.; Liu, N.; Lei, Z.; Chen, B. Chem. Rev. 2016, 116, 3658–3721. Moliner, M.; Corma, A. Annu. Rev. Chem. Biomol. Eng. 2019, 4, 223–234. Fickel, D. W.; D’Addio, E.; Lauterbach, J. A.; Lobo, R. F. Appl Catal B 2011, 102, 441–448. Moliner, M.; Franch, C.; Palomares, E.; Grill, M.; Corma, A. Chem. Commun. 2012, 48, 8264–8266. Ragauskas, A. J.; Williams, C. K.; Davison, B. H.; Britovsek, G.; Cairney, J.; Eckert, C. A.; Frederick, W. J., Jr.; Hallett, J. P.; Leak, D. J.; Liotta, C. L.; Mielenz, J. R.; Murphy, R.; Templer, R.; Tschaplinski, T. Science 2006, 311, 484–489. Milovanovic, J.; Luque, R.; Tschentscher, R.; Romero, A. A.; Li, H.; Shih, K.; Rajic, N. Biomass Bioenergy 2017, 103, 29–34. Graça, I.; Bacariza, M. C.; Fernandes, A.; Chadwick, D. Appl Catal B 2018, 224, 660–670. Ennaert, T.; Van Aelst, J.; Dijkmans, J.; De Clercq, R.; Schutyser, W.; Dusselier, M.; Verboekend, D.; Sels, B. F. Chem. Soc. Rev. 2016, 45, 584–611. Delidovich, I.; Palkovits, R. ChemSusChem 2016, 9, 547–561. Luo, H. Y.; Lewis, J. D.; Román-Leshkov, Y. Annu. Rev. Chem. Biomol. Eng. 2016, 7, 663–692. Grand, J.; Talapaneni, S. N.; Vicente, A.; Fernandez, C.; Dib, E.; Aleksandrov, H. A.; Vayssilov, G. N.; Retoux, R.; Boullay, P.; Gilson, J.-P.; Valtchev, V.; Mintova, S. Nat. Mater. 2017, 16, 1010–1015. Moliner, M.; Román-Leshkov, Y.; Davis, M. E. Proc. Natl. Acad. Sci. 2010, 107, 6164–6168. Luo, H. Y.; Consoli, D. F.; Gunther, W. R.; Román-Leshkov, Y. J. Catal. 2014, 320, 198–207. Agarwal, A.; Park, S.-J.; Park, J.-H. Fuel 2020, 271, 117630. Parulkar, A.; Joshi, R.; Deshpande, N.; Brunelli, N. A. Appl. Catal. A 2018, 566, 25–32. Parulkar, A.; Spanos, A. P.; Deshpande, N.; Brunelli, N. A. Appl. Catal. A 2019, 577, 28–34. Li, P.; Liu, G.; Wu, H.; Liu, Y.; Jiang, J.; Wu, P. J. Phys. Chem. C 2011, 115, 3663–3670. Liao, Y.; Koelewijn, S.-F.; Van den Bossche, G.; Van Aelst, J.; Van den Bosch, S.; Renders, T.; Navare, K.; Nicolaï, T.; Van Aelst, K.; Maesen, M.; Matsushima, H.; Thevelein, J. M.; Van Acker, K.; Lagrain, B.; Verboekend, D.; Sels, B. F. Science 2020, 367, 1385–1390. Camargo, M. D.; Pimenta, J. L. C. W.; Camargo, M. D.; Arroyo, P. A. Fuel 2020, 281, 118719. Montañez Valencia, M. K.; Padró, C. L.; Sad, M. E. Appl Catal B 2020, 278, 119317. Gushchin, P. A.; Lubimenko, V. A.; Petrova, D. A.; Ivanov, E. V.; Kolesnikov, I. M.; Vinokurov, V. A. Chem. Eng. Sci. 2020, 227, 115903. Zawadzki, B.; Kowalewski, E.; Asztemborska, M.; Matus, K.; Casale, S.; Dzwigaj, S.; Sre˛ bowata, A. Catal. Commun. 2020, 145, 106113. Deng, R.; You, K.; Ni, W.; Zhao, F.; Liu, P.; Luo, H. Appl. Catal. A 2020, 594, 117468. Spivey, J. J.; Hutchings, G. Chem. Soc. Rev. 2014, 43, 792–803. Schmidt, M. Environ. Iss. 1994, 23–30. Grundner, S.; Markovits, M. A. C.; Li, G.; Tromp, M.; Pidko, E. A.; Hensen, E. J. M.; Jentys, A.; Sanchez-Sanchez, M.; Lercher, J. A. Nat. Commun. 2015, 6, 7546. Han, S. J.; Kim, S. K.; Hwang, A.; Kim, S.; Hong, D.-Y.; Kwak, G.; Jun, K.-W.; Kim, Y. T. Appl Catal B 2019, 241, 305–318. Konnov, S. V.; Dubray, F.; Clatworthy, E. B.; Kouvatas, C.; Gilson, J.-P.; Dath, J.-P.; Minoux, D.; Aquino, C.; Valtchev, V.; Moldovan, S.; Koneti, S.; Nesterenko, N.; Mintova, S. Angew. Chem. Int. Ed. 2020, 59, 2–10. Frusteri, F.; Migliori, M.; Cannilla, C.; Frusteri, L.; Catizzone, E.; Aloise, A.; Giordano, G.; Bonura, G. J. CO2 Utiliz. 2017, 18, 353–361. Li, Z.; Qu, Y.; Wang, J.; Liu, H.; Li, M.; Miao, S.; Li, C. Joule 2019, 3, 570–583. Georgieva, V. M.; Bruce, E. L.; Verbraeken, M. C.; Scott, A. R.; Casteel, W. J., Jr.; Brandani, S.; Wright, P. A. J. Am. Chem. Soc. 2019, 141, 12744–12759. Flanigen, E. M. Stud. Surf. Sci. Catal. 2001, 137, 11–35. Payra, P.; Dutta, P. K. Zeolites: A Primer. In Handbook of Zeolite Science and Technology; Auerbach, S. M., Carrado, K. A., Dutta, P. K., Eds., Marcel Dekker: New York, 2003; pp 1–19. Chapter 1.

40

Synthesis and application of (nano) zeolites

289. Ridha, F. N.; Webley, P. A. Sep. Purif. Technol. 2009, 67, 336–343. 290. Shang, J.; Li, G.; Singh, R.; Gu, Q.; Nairn, K. M.; Bastow, T. J.; Medhekar, N.; Doherty, C. M.; Hill, A. J.; Liu, J. Z.; Webley, P. A. J. Am. Chem. Soc. 2012, 134, 19246– 19253. 291. Debost, M.; Klar, P. B.; Barrier, N.; Clatworthy, E. B.; Grand, J.; Laine, F.; Brázda, P.; Palatinus, L.; Nesterenko, N.; Boullay, P.; Mintova, S. Angew. Chem. Int. Ed. 2020. https://doi.org/10.1002/anie.202009397. 292. Xu, X.; Bao, Y.; Song, C.; Yang, W.; Liu, J.; Lin, L. Micropor. Mesopor. Mater. 2004, 75, 173–181. 293. Guo, H.; Kong, G.; Yang, G.; Pang, J.; Kang, Z.; Feng, S.; Zhao, L.; Fan, L.; Zhu, L.; Vicente, A.; Peng, P.; Yan, Z.; Sun, D.; Mintova, S. Angew. Chem. Int. Ed. 2020, 59, 6284–6288. 294. Esmaeili, N.; Boyd, S. E.; Brown, C. L.; Gray, E. M. A.; Webb, C. J. ChemPhysChem 2019, 20, 1590–1606. 295. Rasoulia, M.; Yaghobi, N.; Allahgholipour, F.; Atashi, H. Chem. Eng. Res. Design 2014, 92, 1192–1199. 296. Drobek, M.; Figoli, A.; Santoro, S.; Navascués, N.; Motuzas, J.; Simone, S.; Algieri, C.; Gaeta, N.; Querze, L.; Trotta, A.; Barbieri, G.; Mallada, R.; Julbe, A.; Drioli, E. Micropor. Mesopor. Mater. 2015, 207, 126–133. 297. Rac, V.; Rakic, V.; Stosic, D.; Pavlovic, V.; Bosnar, S.; Auroux, A. Arab. J. Chem. 2020, 13, 1945–1954. 298. Khalil, I.; Jabraoui, H.; Lebègue, S.; Kim, W. J.; Aguilera, L.-J.; Thomas, K.; Maugé, F.; Badawi, M. Chem. Eng. J. 2020, 402, 126264. 299. Dyer, A. Stud. Surf. Sci. Catal. 2007, 168, 525–553. 300. Yamagishi, I.; Nagaishi, R.; Kato, C.; Morita, K.; Terada, A.; Kamiji, Y.; Hino, R.; Sato, H.; Nishihara, K.; Tsubata, Y.; Tashiro, S.; Saito, R.; Satoh, T.; Nakano, J.; Ji, W.; Fukushima, H.; Sato, S.; Denton, M. J. Nucl. Sci. Technol. 2014, 51, 1044–1053. 301. Sachse, A.; Merceille, A.; Barré, Y.; Grandjean, A.; Fajula, F.; Galarneau, A. Micropor. Mesopor. Mater. 2012, 164, 251–258. 302. Villa, C. C.; Galus, S.; Nowacka, M.; Magri, A.; Petriccione, M.; Gutiérrez, T. J. Trends Food Sci. Technol. 2020, 102, 102–122. 303. Wei, H.; Seidi, F.; Zhanga, T.; Jin, Y.; Xiao, H. Food Chem. 2021, 337, 127750. 304. Tan, K.-H.; Awala, H.; Mukti, R. R.; Wong, K.-L.; Rigaud, B.; Ling, T. C.; Aleksandrov, H. A.; Koleva, I. Z.; Vayssilov, G. N.; Mintova, S.; Ng, E.-P. J. Agric. Food Chem. 2015, 63, 4655–4663. 305. Wan, Y.; Xu, W.; Ren, X.; Wang, Y.; Dong, B.; Wang, L. Front. Bioeng. Biotechnol. 2020, 8, 628. 306. Vernero, M.; Di Leo, I.; Givone, M.; Adriani, A.; Bergamaschi, A.; Fanelli, M.; Astegiano, M. Minerv. Gastroenterolog. Dietolog. 2020, 66, 113–116. 307. Hrenovic, J.; Milenkovic, J.; Goic-Barisic, I.; Rajic, N. Micropor. Mesopor. Mater. 2013, 169, 148–152. 308. Anfray, C.; Komaty, S.; Corroyer-Dulmont, A.; Zaarour, M.; Helaine, C.; Ozcelik, H.; Allioux, C.; Toutain, J.; Goldyn, K.; Petit, E.; Bordji, K.; Bernaudin, M.; Valtchev, V.; Touzani, O.; Mintova, S.; Valable, S. Biomaterials 2020, 257, 120249. 309. Bacakova, L.; Vandrovcova, M.; Kopova, I.; Jirka, I. Biomater. Sci. 2018, 6, 974–989. 310. Georgieva, V.; Retoux, R.; Ruaux, V.; Valtchev, V.; Mintova, S. Front. Chem. Sci. Eng. 2018, 12, 94–102. 311. Yang, B. X.; Zhang, Z.; Tian, C.; Yuan, W. J.; Hua, Z. Q.; Fan, S. R.; Wu, Y.; Tian, X. M. Sensor. Actuat. B 2020, 321, 128567. 312. Gul, S. E.; Cody, D.; Kharchenko, A.; Martin, S.; Mintova, S.; Cassidy, J.; Naydenova, I. Micropor. Mesopor. Mater. 2018, 261, 268–274. 313. Kalambate, P. K.; Rao, Z.; Wu, J.; Shen, Y.; Boddula, R.; Huang, Y. Biosens. Bioelectron. 2020, 163, 112270.

6.03

Mesostructured materials

Feng Yu and Feng-Shou Xiaoa,b, a Key Lab of Applied Chemistry of Zhejiang Province, Department of Chemistry, Zhejiang University, Hangzhou, China.; and b Key Lab of Biomass Chemical Engineering of Ministry of Education, College of Chemical and Biological Engineering, Zhejiang University, Hangzhou, China. a

© 2023 Elsevier Ltd. All rights reserved.

6.03.1 6.03.2 6.03.2.1 6.03.2.2 6.03.2.3 6.03.2.3.1 6.03.2.3.2 6.03.2.4 6.03.2.5 6.03.2.5.1 6.03.2.5.2 6.03.2.5.3 6.03.3 6.03.3.1 6.03.3.2 6.03.3.2.1 6.03.3.2.2 6.03.3.2.3 6.03.3.3 6.03.3.4 6.03.4 References

Introduction Synthesis of mesoporous materials Ordered mesoporous silicas (OMSs) Mesoporous metals and metal oxides Hybrid mesoporous materials Ordered mesoporous organosilicas (OMOSs) Mesoporous metal-organic frameworks (MOFs) Ordered mesoporous carbons (OMCs) Mesoporous zeolites Bottom-up zeolite synthetic strategies Top-down synthetic strategy via demetallization Mixed synthetic strategy Catalytic applications of mesoporous materials Mesoporous metals and metal oxides for catalysis Catalytic applications of representative ordered mesoporous materials Ordered mesoporous silicas for catalysis Ordered mesoporous organosilicas for catalysis Ordered mesoporous carbons for catalysis Mesoporous metal-organic frameworks for catalysis Mesoporous zeolites for catalysis Summary and perspective

41 42 42 43 45 45 46 47 48 49 51 52 53 53 56 56 56 57 59 60 62 62

Abstract The development of mesoporous materials has offered great opportunities for new applications in a variety of fields, including heterogeneous catalysis, thanks to their special intrinsic structural features. In this article, we focus on the main achievements of mesoporous materials in both synthesis and catalysis fields. The development of synthesis strategies, especially designed for generating improved physicochemical and textural properties of mesoporous materials, which include silicas, metals, metal oxides, organosilicas, metal-organic frameworks, carbons and zeolites, is briefly summarized with special emphasis on mesoporous zeolites. Additionally, prominent cases of recent advances of using these fabricated fascinating mesoporous materials as heterogeneous catalysts in various catalytic reactions are presented.

6.03.1

Introduction

Since the pioneering work of Mobil’s research on the synthesis of M41S,1 a new class of mesoporous materials, the research passion on developing materials with mesoporous architecture has gained an explosive growth.2–4 According to IUPAC definition, mesoporous materials are those possessing pore widths in the range from 2 to 50 nm, which is between micropores and macropores in size.5 Mesoporous materials are of great interest in both scientific and technological fields due to their many merits such as a large surface area and high pore volume, an improved accessibility and the ability to accommodate and interact with various chemical functionalities on their surface.6 Since the discovery of MCM-41, the development of mesoporous silica materials has progressed at a rapid pace, which covers several research topics including synthesis,7,8 characterization,9,10 morphological and alignment control of mesochannels.11,12 The incorporation of organic units and other metal atoms into the silica frameworks leads to diversified framework compositions for mesoporous materials. So far the scope of mesoporous materials has already been expanded to organosilicas, metals, metal oxides, carbons, metal-organic frameworks (MOFs), zeolites and so on. In order to create uniform ordered mesopores a number of templates have been utilized to prepare mesoporous materials, the synthesis of which can be categorized into three major routes based on the type of templates used: hard templating route (carbon black, carbon nanotubes, silica, colloidal crystals, particles, wood cells, etc.), soft templating route (polymers, surfactants), and non-templating route. Some of mesoporous materials, particularly zeolites and MOFs, exhibit hierarchical porosity, referring to at least two levels of pore size in their

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00006-6

41

42

Mesostructured materials

structure, which include micropores and mesopores. Mesoporous materials can be utilized in a wide spectrum of fields to meet demands of diverse applications ranging from adsorption,13 catalysis,14 chemical sensing,15 to drug delivery,16 energy storage and conversion devices.17 In this article, the first section will be highlighted with recent advances in the synthesis of mesoporous materials. The contents of this section are classified into the following categories on the basis of the types of materials: (i) silicas, (ii) metal and metal oxides, (iii) hybrid materials including organosilicas and MOFs, (iv) carbons, and (v) zeolites. Since the benefits of the mesopore in these materials are established today by an impressive number of publications in catalytic applications, prominent recent examples of the use of mesoporous materials mentioned above as the heterogeneous catalysts for catalytic transformation will be given in the second section.

6.03.2

Synthesis of mesoporous materials

6.03.2.1

Ordered mesoporous silicas (OMSs)

In 1992, Mobil Research and Development Corporation successfully synthesized ordered mesoporous silica, denoted as MCM-41, through self-assembly of silica species and cationic surfactants, which was marked as a real breakthrough in the field of nanoporous materials.1,18 In the synthesis processes, the positively charged quaternary ammonium surfactant micelles attracted negatively charged silica species through static interactions, which was able to induce the silica species mediating the hexagonal ordering. After removal of the organic micelles by calcination at 540
C, mesoporous silica with ordered and uniform as well as cylindrical structure was synthesized and further characterized by X-ray diffraction, transition electron microscopy and nitrogen adsorption analysis (Fig. 1A). Moreover, the study revealed that the pore dimensions of MCM-41 can be tuned by varying chain length of the surfactant employed or changing solution chemistry. Mesoporous materials in M41S family with different geometries, including cubic (MCM48) and lamellar structures (MCM-50), were prepared by increasing the surfactant to silicon ratio (Fig. 1B). In presence of surfactants, several OMSs featured with high surface area and pore volume as well as various crystallographic phases have been reported.19,20 However, the surfactant strategy shows an obvious limit to the mesopore size of these materials, usually resulting in the pore diameters in the range of 2–10 nm. Moreover, the amorphous nature of these materials causes a negative effect on their thermal and hydrothermal stability as well as acid tolerance properties. Besides surfactant templates, polymer was also successfully utilized as the template for synthesizing a new class of well-ordered mesoporous silicas by Stucky and Chmelka et al. in 1998.7 Silica SBA-15, indicating Santa Barbara Amorphous, as the most recognized representative in this class of materials, demonstrated a two-dimensional hexagonal network with uniform pore sizes up to 30 nm. Compared with MCM-41, SBA-15 exhibited larger pore dimensions, which was attributed to the formation of larger micelles from long-chained triblock copolymer P123 (EO20PO70EO20). It was also found that acidic condition (pH < 1) was necessary to

Fig. 1 (A) Schematic illustration of the synthesis of surfactant-templated ordered mesoporous silica MCM-41. (B) Materials examples with different geometries, including hexagonal MCM-41, cubic MCM-48 and lamellar MCM-50. Adapted with permission from Beck, J.S.; Vartuli, J.C.; Roth, W.J.; Leonowicz, M.E.; Kresge, C.T.; Schmitt, K.D.; Chu, C.T.W.; Olson, D.H.; Sheppard, E.W.; McCullen, S.B.; et al. A New Family of Mesoporous Molecular Sieves Prepared With Liquid Crystal Templates. J. Am. Chem. Soc. 1992, 114 (27), 10834–10843. Copyright©1992, American Chemical Society; Adapted with permission from ref. Van Der Voort, P.; Leus, K.; De Canck, E. Introduction to Porous Materials; John Wiley & Sons Ltd.: Hoboken, NJ, 2019. Copyright©2019, John Wiley and Sons.

Mesostructured materials

43

induce precipitation of ordered hexagonal mesostructured SBA-15. The subsequent studies conducted by Jaroniec and Ryoo et al. revealed that SBA-15 possessed interconnecting irregular micropores in the walls of ordered mesopores, which were attributed to the penetration of flexible amphiphilic polyethylene chains within the silica matrix (Fig. 2).21,22 The pore architecture of SBA-15 was found to be different from that of MCM-41, which was only featured with isolated ordered hexagonally mesopores. SBA-15 materials, compared to MCM-41, afforded higher thermal/hydrothermal stability due to the interconnected micro-meso hierarchical porosity and thicker mesopore walls. In the years that followed, cage-like mesostructured materials, such as SBA-16 and FDU-1, were synthesized by using other block copolymers as soft templates, such as EO106PO70EO106 and EO39BO47EO39.9,23 To date, the synthesis of SBA-type OMSs with different structures, which include hexagonal close-packed array (SBA-2),24 threedimensional hexagonal network (SBA-12),25,26 lamellar structure (SBA-14),25 two-dimensional hexagonal (SBA-15),7 and bodycentered cubic structure (SBA-16)25,26 have already been achieved. It is worth noting that the mesoporous silicas based on the polymer-templating strategy possess an additional microporosity and generally show thicker pore walls, larger mesopore size and higher thermal/hydrothermal stability when compared to the siliceous counterparts prepared via surfactant-templating method. The ability of surfactants and block copolymers to form micelles and subsequently ordered mesophases was found to be the key factor for the synthesis of OMSs. This finding directly led to rapid development of the soft templating strategy towards fabricating ordered mesoporous silicas by employing a variety of cationic, anionic and non-ionic surfactants or suitable polymers capable of forming micellar systems.19,25,27 Tetraethyl orthosilicate (TEOS) and sodium silicate, as silica precursors, were usually used for synthesizing OMSs. After the completion of solidifying inorganic siliceous species surrounding the organized micelles, the removal of micellar soft templates was conducted either by solvent extraction or thermal treatment. A fine-tuning of SBA-15 properties including pore size and thermal stability can be realized with the assistance of microwave irradiation via changing synthesis temperature and time.28 So far a wide spectrum of amorphous OMSs with distinct morphology and physicochemical properties have been explored, demonstrating different pore characteristics of mesostructures, which include two-dimensional hexagonal, bicontinuous cubic, face-centered cubic, body-centered cubic and lamellar symmetry (Fig. 3).8,29–33

6.03.2.2

Mesoporous metals and metal oxides

Two distinguishable methods, hard and soft templating routes, are conventionally used for the preparation of mesoporous metals and metal oxides. The hard templating strategy consists of three technical steps: (i) infiltrating appropriate precursors into the sacrificial hard templates, usually mesoporous carbon or silica; (ii) transforming the precursors to the desired materials; (iii) removing the original hard templates. The strategy based on hard template provides a versatile method for fabricating various classes of mesoporous metals and metal oxides. Nevertheless, the operation seems slightly cumbersome and it is generally considered to be impractical for production in a large scale.34 In the soft templating approach, lyotropic liquid crystals (LLCs) of surfactants usually function as direct and soft templates. By condensing or reducing the precursors that embedded in the confined LLC voids, the generation of the target materials occurs.35,36 Compared with the hard templating method, the soft templating approach presents several advantages. For instance, the construction of mesoporosity for metals and metal oxides can be realized in a rather efficient route, through a solvent-evaporation process. In 1997, Attard et al. successfully prepared platinum films with a well-defined long-ranged mesoporous nanostructure and high specific surface area via electrochemical reduction of platinum salts confined in the liquid crystalline plating solution, which opened a new way to fabricate various mesoporous metal films for different uses.36 In addition, Yamauchi et al. have provided a novel hard

Fig. 2 Schematic illustration of the synthesis of SBA-15 using PEO-PPO-PEO triblock copolymers. Reprinting with permission from Van Der Voort, P.; Leus, K.; De Canck, E. Introduction to Porous Materials; John Wiley & Sons Ltd.: Hoboken, NJ, 2019. Copyright©2019, John Wiley and Sons.

44

Mesostructured materials

Fig. 3 Examples of ordered mesoporous silicas structures. Adapted with permission from Wan, Y.; Zhao, D. On the Controllable Soft-Templating Approach to Mesoporous Silicates. Chem. Rev. 2007, 107(7), 2821–2860. Copyright©2007, American Chemical Society.

templating strategy for the preparation of mesoporous monocrystalline Pt nanoparticles, which was synthesized via controlled chemical reduction and employing 3D bicontinuous mesoporous silica (KIT-6) and 2D mesoporous silica (SBA-15) as hard templates.34 They further proposed a concept, namely “micelle assembly,” as a novel strategy to prepare mesoporous metal films by a simple electrodeposition technique.37 With the assistance of micelle assembly, the metallic mesoporous nanostructures, such as Pt films, were generated in a dilute surfactant electrolyte (Fig. 4). Very recently, they have demonstrated the synthesis of mesoporous palladium nanoparticles by employing a polymeric micelle assembly method.38 It was found that the pore construction of the mesoporous Pd nanoparticles can be tuned by varying solvent compositions. In 1995, Ying and Antonelli reported the successful synthesis of ordered mesoporous metal oxides for the first time.39 Relying on the strong interactions between titanium oxide framework and phosphorus as well as tetradecylphosphate as the soft template, they prepared hexagonally packed mesoporous titanium oxide, which exhibited a surface area of around 200 m2 g 1 after removing the used template by calcination. After this first report, several other ordered mesoporous metal oxides possessing amorphous or semicrystalline pore walls, including Al2O3, ZrO2, Nb2O5, Ta2O5, WO3, HfO2, SnO2 etc., were developed in the 10 years that followed.40–42 In 2006, a-Fe2O3 featured with crystalline wall, as the first example of ordered mesoporous crystalline metal oxides, was synthesized by Bruce et al. employing iron nitrate nonahydrate and KIT-6 as the metal oxide source and the hard-template respectively.43 In the synthesis of mesoporous crystalline oxides, collapse of mesoporous structure after thermal treatment for the crystallization of metal oxides is usually considered as a major challenge, due to the fact that a reduction in framework density and a subsequent shrinkage of mesopore walls occur during the transformation of amorphous walls into crystallized frameworks.44 To overcome this challenge, an interesting strategy was introduced by reinforcing mesoporous structures with the use of silicate- or carbon-based species. The general procedures of this strategy consist of the following steps: (i) the mesoporous oxides with amorphous walls are firstly prepared through traditional surfactant-directed sol-gel approaches; (ii) the amorphous walls of metal oxides are then protected by silicate or carbon coatings prior to crystallization at high temperature; (iii) the crystallization of amorphous walls into crystalline structures is processed under thermal treatment; (iv) the removal of the protective layers through treatment in alkali or HF solution for the cases of using silicate as the strengthening species, or through calcination in air for the cases of using

Fig. 4 Synthetic concept “micelle assembly” for the preparation of mesoporous Pt films. Inset image is photograph of the mesoporous Pt film deposited on Au substrate. The black area indicates the deposited Pt. Reprinting with permission from Wang, H.; Wang, L.; Sato, T.; Sakamoto, Y.; Tominaka, S.; Miyasaka, K.; Miyamoto, N.; Nemoto, Y.; Terasaki, O.; Yamauchi, Y. Synthesis of Mesoporous Pt Films With Tunable Pore Sizes From Aqueous Surfactant Solutions. Chem. Mater. 2012, 24(9), 1591–1598. Copyright©2012, American Chemical Society.

Mesostructured materials

45

carbonaceous protective layer, finally giving rise to the mesoporous crystalline metal oxides (Fig. 5). This strategy has been proven to be efficacious for the preparation of a series of mesoporous metal oxides, which include TiO2,45,46 Nb2O5,47 Ta2O5,47,48 and Al2O3.49 The avoidance of using high temperature for crystallization can be achieved by adding metal oxide nanocrystals into the synthesis gels. Bein et al. have presented a novel “brick and mortar” approach for creating highly crystalline mesoporous titania films.50 The approach was based on the assembly of preformed titania nanocrystalline “bricks” with surfactant-fused sol-gel titania “mortar.” Similar chemical composition of both bricks and mortar led to an enhancement in crystallization upon thermal treatment through a striking synergistic interaction of crystalline and amorphous components, resulting in the formation of highly mesoporous crystalline structures under mild conditions. With this concept, mesoporous titania-silica composite films and multilayered titania films were also successfully prepared.51,52 Related methods whereby the preformed metal oxide nanocrystals were assembled with the assistance of surfactants were also utilized for fabricating several mesoporous metal oxides, such as tin oxide,53 ceria,54 and alumina.55,56

6.03.2.3 6.03.2.3.1

Hybrid mesoporous materials Ordered mesoporous organosilicas (OMOSs)

The combination of inorganic and organic components often gives rise to the materials with special properties, which has attracted considerable research interest in assembly of inorganic solids with organic molecules.57 OMOSs represent an important class of mesoporous organic-inorganic hybrid materials, which are constructed on the basis of silica modified with functional organic groups Si–R on the pore walls, showing well-defined pore-structure and tunable function properties due to the presence of organic groups.58,59 As shown in Fig. 6, three main routes have been developed to introduce organic functional groups on silica materials: (i) via post-modification of the prepared silicas, (ii) via condensation of bridged alkoxysilanes, (iii) via co-condensation of silica precursors and bridged alkoxysilanes. It should be noted that OMOSs with bridging organic groups are usually known as periodic mesoporous organosilicas. They are comprised of a silica network containing organic functionalities that are homogeneously distributed throughout the entire materials. The synthesis of OMOSs was first and almost simultaneously reported in 1999 by three research groups, headed by Ozin, Ingaki, Stein, respectively.60–62 The discovery of this mesoporous nanomaterial and the subsequent tremendous developments in this research field made delicate-tuning the physicochemical properties of these functional materials possible.63–65 For example, in 2005 Jaroniec et al. successfully incorporated a large heterocyclic bridging group, that is, tris[3-(trimethoxysilyl)propyl]isocyanurate, into the silica framework while preserving the mesostructural ordering of this material.66 A high percentage of these heterocyclic chelating groups in the silica framework can be achieved for further practical applications. From a practical point of view, the incorporation of organic components into inorganic silicas expanded the scope of potential applications of these fabricated materials, particularly in biochemistry, organic chemistry and host-guest chemistry.67–69 OMOSs functionalized with organic groups usually are comprised of amorphous frameworks. In contrast, the frameworks of mesoporous organosilica materials with bridged aromatic groups are capable of creating crystal-like layered structures with molecular-scale periodicity. A remarkable development in OMOSs was the first preparation of an ordered mesoporous benzenesilica hybrid material featured with crystalline pore walls by Inagaki et al. in 2002.70 This material was synthesized through a surfactant-mediated strategy, exhibiting a hexagonal array of mesopores with a lattice constant of 5.25 nm and crystal-like pore walls with structural periodicity of a 0.76 nm spacing along the channel direction. The study further revealed that alternating hydrophilic and hydrophobic layers, originated from silica and benzene respectively, mainly induced the formation of this periodic pore structure. This investigation paved a new way to generating OMOSs with crystalline walls. Since crystallinity is a feature of significant importance for materials, a more controllable chemical modification is possible for this crystallined organosilicas compared with the disordered counterparts.

Fig. 5 Schematic illustration of the strategy for reinforced crystallization of mesoporous metal oxides: (A) back filling, (B) coating, (C) crystallization, and (D) removal of the reinforcement. Reprinting with permission from Kondo, J. N.; Domen, K. Crystallization of Mesoporous Metal Oxides. Chem. Mater. 2008, 20(3), 835–847. Copyright©2008, American Chemical Society.

46

Mesostructured materials

Fig. 6 Three pathways for the preparation of OMOSs with organic moieties in their frameworks. Adapted with permission from Park, S.S.; Santha Moorthy, M.; Ha, C.-S. Periodic Mesoporous Organosilicas for Advanced Applications. NPG Asia Mater. 2014, 6(4), 96. Copyright©2014, Springer Nature.

6.03.2.3.2

Mesoporous metal-organic frameworks (MOFs)

MOFs are crystalline porous materials with hybrid network structures built from the assembly of metal ions or metal clusters and organic linkers, and sustained by coordination bonds. An important feature of MOFs lies in the fact that their framework structure, pore dimension and functionality can be finely tuned by choosing different organic building blocks and metals. Most of MOFs have been reported to date with pores in the microporous range.71,72 The first MOFs with mesopore structures were prepared by Yaghi et al. in 2002 through using a linear linker, TPDC ([1,10 :40 ,100 -terphenyl]-4,400 -dicarboxylate) and Zn4O(CO2)6 as the secondary building unit.73 The as-obtained 3D IRMOF-16 exhibited channel sizes or fixed free diameter values of 2.88 nm, as shown in Fig. 7A. A series of frameworks with pore sizes in the mesoporous range were systematically designed and fabricated based on the same skeleton of MOF-5 (known as IRMOF-1). Moreover, the cubic topology remained as before even though the size and functionality of the pores were changed. The calculated low crystal density (0.21 g cm 3) and high free volume (91.1%) revealed that IRMOF-16 possessed an excellent mesoporosity, which consequently improved the gas uptake capacity. In addition, with the use of 4,40 ,400 -[benzene-1,3,5-triyl-tris(ethyne-2,1-diyl)]tribenzoate (BTE) and 4,40 ,400 -[benzene-1,3,5-triyl-tris(benzene-4,1-diyl)]tribenzoate (BBC) as the ligands, mesoporous MOFs featured with ultrahigh porosity, including MOF-180 and MOF-200, were synthesized by Yaghi’s group in 2010.74 The cage sizes of MOF-180 (Fig. 7B) and MOF-200 (Fig. 7C) were 1.5 by 2.3 nm and 1.8 by 2.8 nm respectively, which is on the border of micropores and mesopores. The bulk density for MOF-200 was as low as 0.22 g cm 3, which consequently afforded a void volume reaching 90% of the crystal volume. Additionally, joining Zn4O(CO2)6 with two kinds of organic link, BTE and biphenyl-4-40 -dicarboxylate (BPDC), gave rise to the highly porous MOF210 with larger internal pore diameter and surface area, showing up to 4.8 nm and 6240 m2 g 1 respectively. The synthesis of mesoporous MOFs can also be achieved via an alternative approach, which is involved using surfactants, copolymers or designed spacers as the structure directing agents.75–77 For instances, hierarchically porous MOF with tailored interconnecting micropores and mesopores were prepared via self-assembly of the framework building blocks in the presence of surfactant

Fig. 7 Crystal structures of MOFs: (A) IRMOF-16, (B) MOF-180, (C) MOF-200. Reprinting from Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O’Keeffe, M.; Yaghi, O.M. Systematic Design of Pore Size and Functionality in Isoreticular MOFs and Their Application in Methane Storage. Science 2002, 295(5554), 469–472, Furukawa, H.; Ko, N.; Go, Y.B.; Aratani, N.; Choi, S.B.; Choi, E.; Yazaydin, A.Ö.; Snurr, R.Q.; O’Keeffe, M.; Kim, J.; et al. Ultrahigh Porosity in Metal-Organic Frameworks. Science 2010, 329 (5990), 424–428.

Mesostructured materials

47

micelles by Qiu et al. in 2008.75 They employed cetyltrimethylammonium bromide (CTAB) surfactant as the structure directing agent, and chose Cu2þ and benzene-1,3,5-tricarboxylate ions as the framework building blocks. The obtained mesostructured MOFs possessed a mesopore system with diameters tunable in the range of 3.8–31.0 nm, relying on the synthetic conditions employed. In addition, Zhang and Han et al. reported the synthesis of Zn-containing MOF nanospheres in the presence of supercritical CO2 and an ionic liquid as well as N-ethyl perfluorooctylsulfonamide surfactant (Fig. 8).76 The study demonstrated that the synthesized MOF nanospheres incorporated well-ordered mesopores with a pore size distribution centered at around 3.6 nm, and the walls of mesopores were constructed by a microporous framework. For a more in-depth examination of the synthesis of mesoporous MOFs, readers are kindly directed to the recent excellent reviews, which give more systematic summaries of the synthetic strategies.78,79

6.03.2.4

Ordered mesoporous carbons (OMCs)

Activated carbon is the most widespread and commonly known carbon-based materials, showing a form of graphite with a random porous structure and containing pores over a wide range of pore sizes. It has long been used as adsorbents for purifying air and water. Compared with activated carbon, OMCs are a more advanced type of porous materials, possessing combined property merits of both mesoporous inorganic and organic polymer materials, such as high porosity, hydrophobicity, thermal stability in an inert atmosphere, unique chemical stability in acidic and basic conditions, and electrical/thermal conductivity, which make them particularly useful in various application fields.3,80 OMCs can be synthesized via two different approaches, which are hard-templated (nanocasting strategy) and soft-templated (direct synthesis) methods. Both methods involve the preparation of an ordered mesoporous polymer in the initial step, which is subsequently transformed into an ordered mesoporous carbon. A general overview of the different steps in both approaches is illustrated in Fig. 9. Ordered mesoporous carbon was synthesized for the first time by Ryoo et al. in 1999, using mesoporous silica MCM-48 as a hard template and sucrose as a carbon precursor.81 The synthesis procedures started with an impregnation of siliceous pores with sucrose. After carbonization of this carbon precursor via a mild process with the use of a sulfuric acid catalyst and a subsequent silica removal process in a strong alkaline solution, an ordered mesoporous carbon (CMK-1) with uniform pores of 3.0 nm in dimension was obtained. CMK-3, an analogous material, featuring pore sizes of around 4.5 nm was subsequently prepared through the same synthesis strategy by utilizing silica SBA-15 as the template.82 The OMCs, including CMK-1 and CMK-3, prepared through the hard templating route afforded high specific surface areas, showing 1380 and 1520 m2 g 1 respectively. Further studies revealed that the synthesis of interconnected mesoporous carbon nanopipes could be achieved in the case that the filling of the SBA-15 template pores with a suitable carbon precursor was incomplete.83

Fig. 8 Formation of the MOF in the surfactant/IL/CO2 system. (A) Formation of N-EtFOSA cylindrical micelles; (B) MOF with ordered mesopores and microporous structured walls. Reprinting with permission from Zhao, Y.; Zhang, J.; Han, B.; Song, J.; Li, J.; Wang, Q. Metal-Organic Framework Nanospheres With Well-Ordered Mesopores Synthesized in an Ionic Liquid/CO2/Surfactant System. Angew. Chem. Int. Ed. 2011, 50(3), 636–639. Copyright©2010, Wiley-VCH.

48

Mesostructured materials

Fig. 9 Two typical approaches for the preparation of OMC materials. Reprinting with permission from Ma, T.Y.; Liu, L.; Yuan, Z.Y. Direct Synthesis of Ordered Mesoporous Carbons. Chem. Soc. Rev. 2013, 42(9), 3977–4003. Copyright©2013, The Royal Society of Chemistry.

The synthesis of OMCs based on soft templating methods has been rapidly developed since 2005.84–86 The softtemplated synthesis methods are able to avoid the tedious preparation of silica hard templates and their etching after carbonization by using NaOH or HF solutions. For instance, Zhao et al. chose phenol/formaldehyde (resol) as the carbon precursor, which was transformed to three-connected covalently bonded frameworks after thermopolymerization.84 These rigid polymers were further transformed into well-ordered and ultrastable (> 1400
C) mesoporous carbons (C-FDU-15 and C-FDU-16), which exhibited hexagonal and cubic mesostructures respectively with large uniform mesopores, surface areas and pore volumes. A wide variety of OMCs can be prepared by varying carbon precursors, templates or synthetic conditions.87–92 The combination of hard templating method with soft templating method can also give rise to carbon materials with hierarchically porous structures. OMCs with amorphous structure are less attractive due to their deficiency in property, such as low electrical conductivity. Graphitization is a conventionally efficient way to improve the crystallinity of carbon, which can be achieved by thermotreatment in an inert atmosphere at a high temperature or by catalytic graphitization.93–96 For example, a highly graphitized OMC with spherical pores on the borderline between mesopores and macropores (40–100 nm) was synthesized by Jaroniec and Yu et al. through graphitization of the pitch-based carbon, obtained by using silica colloidal crystal as the template, at 2500
C under an argon atmosphere.96

6.03.2.5

Mesoporous zeolites

In order to generate mesoporous zeolite materials, an additional mesopore needs to be incorporated with an interconnectivity with the inherent micropore system dictated by the zeolitic framework structure. A broad variety of approaches have been developed in the last two decades to accomplish the synthesis of zeolites with additional mesoporosity. It can be generally categorized into bottom-up, top-down, and mixed zeolite synthetic strategies. The bottom-up zeolite synthetic approaches refer to the ones that build the mesopore structure either during the zeolite crystallization period or through an aggregation process of pre-synthesized crystals. These methods can be mainly classified into three routes, which are hard templating, soft templating, and non-templating. The former two routes are defined according to the physicochemical nature of the mesopore template, being of more or less rigid nature, while the latter one indicates the synthetic process without using any mesopore template. The top-down strategies, also known as demetallization methods, mainly include dealumination and desilication routes, in which the mesostructure is introduced through selective removal of aluminum or silicon atoms from the zeolite crystal framework via a post-treatment. The mixed synthetic method represents a combination of bottom-up and top-down methods, or in other words zeolite recrystallization approach, in which usually with the assistance of surfactants mesoporosity of zeolite is produced through dissolution and recrystallization of the preformed zeolite crystals.

Mesostructured materials 6.03.2.5.1

49

Bottom-up zeolite synthetic strategies

A number of preparation approaches to mesoporous zeolite denoting the bottom-up strategy have been developed so far. Besides conventional micropore templates or structure directing agents (SDA) employed in the synthesis of zeolite, the use of additional mesopore templates for the creation of mesopore structure in zeolite takes a major proportion of these synthetic approaches. The mesopore templates, including hard and soft ones, are usually encapsulated within the zeolite during the early synthetic process and leached afterwards by calcination to generate mesoporosity. The pioneering work of mesoporous zeolite synthesis with the aid of carbon nanoparticle was reported by the group of Jacobsen.97,98 They used commercial carbon black pearls (BP 2000 or BP700) as a matrix to assist the synthesis of nanozeolites with controlled particle size. To this end, zeolites were crystallized in the voids of the porous carbons or in the confined space between the carbon particles. When the amount of zeolite synthesis gel was equal to or lower than the total pore volume of the carbon matrix, zeolites were only confined to grow in the voids of the matrix and finally provided as nanozeolites with very small nanoparticle sizes (ZSM-5,  20 nm). In contrast, when an excess of the zeolite precursor gel was used in the impregnating process, carbon nanoparticles was encapsulated in zeolite crystals during the crystallization, which resulted in mesoporous zeolite single crystals with a wide mesopore size distribution in the range of 5–50 nm instead (Fig. 10A and F).99 In addition, carbon aerogels can also be used as hard templates to fabricate hierarchical zeolite monoliths with interconnected mesoporous channels. Mesoporous zeolite ZSM-5 (MFI), A (LTA), and Y (FAU) monoliths were successfully prepared by Tao et al. by using the carbon aerogels, which were made via CO2 supercritical drying and subsequent thermal pyrolysis of resorcinol-formaldehyde gels (Fig. 10B and G).100–102 The mesopore sizes from the porous aerogels can be tuned through controlling the reaction conditions. Christensen’s group developed a new method that applied in situ generation of the required carbon template via decomposition of sucrose directly onto the silica raw material employed for the preparation of mesoporous zeolites (Fig. 10C and H).103 The mesoporous zeolite crystals contain interconnected micropores and mesopores inside each individual single crystal with a disordered character. It should be noted that this approach allows fine control of the porosity of the mesoporous zeolite in a very simple manner through just tuning the molar carbon/silica ratio. In 2001 Jacobsen’s group first reported the use of commercial multiwall carbon nanotube templates, with the diameter of 12 nm and the length of several micrometers, for synthesis of mesoporous silicalite-1, giving rise to the formation of straight and uniform mesoporous channels penetrating the zeolite single crystals (Fig. 10D and I).104 Moreover, alternative colloidal mesoporous silica was used to imprint ordered mesoporous carbon for the preparation of desired mesoporous zeolites (Fig. 10E and J).105–107 Tsapatsis’ group realized a wide range of crystal morphologies through confined growth within three-dimensionally ordered mesoporous (3DOm) carbon, which synthesized by replication of colloidal crystals composed of size-tunable silica nanoparticles (10–40 nm).106 Confined crystal growth within these templates gave rise to uniformly shaped silicalite-1 nanocrystals with size tunability as well as 3DOm-imprinted single-crystal zeolite particles.

Fig. 10 Overview of hard templating strategies using different forms of carbon materials. Top (A-E): Schematic diagram showing mesoporous zeolite templated by different carbon. (A) carbon nanoparticles, (B) carbon aerogel, (C) carbon from sugar calcination, (D) carbon nanotubes, (E) 3D ordered mesoporous (3DOm) carbon; bottom (F–J): Electron micrographs of final mesoporous zeolite products of a-e respectively. Adapted with permission from Jacobsen, C.J.H.; Madsen, C.; Houzvicka, J.; Schmidt, I.; Carlsson, A. Mesoporous Zeolite Single Crystals. J. Am. Chem. Soc. 2000, 122(29), 7116–7117, Tao, Y.; Kanoh, H.; Kaneko, K. ZSM-5 Monolith of Uniform Mesoporous Channels. J. Am. Chem. Soc. 2003, 125(20), 6044– 6045, Kustova, M.; Egeblad, K.; Zhu, K.; Christensen, C.H. Versatile Route to Zeolite Single Crystals With Controlled Mesoporosity: In Situ Sugar Decomposition for Templating of Hierarchical Zeolites. Chem. Mater. 2007, 19(12), 2915–2917, Schmidt, I.; Boisen, A.; Gustavsson, E.; Ståhl, K.; Pehrson, S.; Dahl, S.; Carlsson, A.; Jacobsen, C.J.H. Carbon Nanotube Templated Growth of Mesoporous Zeolite Single Crystals. Chem. Mater. 2001, 13(12), 4416–4418, Chen, H.; Wydra, J.; Zhang, X.; Lee, P.S.; Wang, Z.; Fan, W.; Tsapatsis, M. Hydrothermal Synthesis of Zeolites with ThreeDimensionally Ordered Mesoporous-Imprinted Structure. J. Am. Chem. Soc. 2011, 133(32), 12390–12393. Copyright©2000, 2003, 2007, 2001, 2011, American Chemical Society.

50

Mesostructured materials

Besides hard templates, soft templates with relatively flexible species, such as surfactants and polymers, have also been employed as the mesopore templates to achieve better control of the mesopore size and pore connectivity. For instances, Xiao et al. firstly reported a strategy that combined small organic ammonium salts and cationic polymers as micropore and mesopore templates respectively for the synthesis of mesoporous zeolites in 2006.108 They used tetraethylammonium hydroxide and polydiallyldimethylammonium chloride (PDADMAC) templates and successfully obtained a Beta zeolite with a mesopore size distribution of 5– 40 nm. The cationic polymer PDADMAC was also found its use as a dual-function template for preparing single Beta crystals with mesoporosity in the range of 4–10 nm by simply adjusting the molecular weight of PDADMAC.109 Moreover, Xiao’s group also successfully synthesized b-axis-aligned mesoporous ZSM-5 with a size range of 10–50 nm by using the designed amphiphilic copolymer (C-PSt-co-P4VP) as the mesoporous soft template.110 C-PSt-co-P4VP was obtained from copolymer polystryrene-co-4polyvinylpyridine after positively charging by methyl iodide treatment. The particular orientation of the acquired mesopores was probably due to the copolymer template being energetically prone to occupy the (010) face in the self-assembly according to DFT calculations. Quaternary ammonium surfactants, including single and dual-function ones, were also designed for the preparation of mesoporous zeolites. A whole new idea was introduced by the group of Ryoo in 2009 for synthesizing ultrathin MFI zeolite nanosheets with the use of an elegantly designed dual-function surfactant C22H45-Nþ(CH3)2-C6H12-Nþ(CH3)2-C6H13(Br)2, which was abbreviated as C22–6-6.111 The study demonstrated that the hydrophobic long-chain C22 group caused the formation of mesoscale micellar structure and restrained the excessive growth of zeolite, in the meanwhile two quaternary ammonium sites on the surfactant C22–6-6 spaced by a C6 alkyl chain directed the microporous MFI structure generation, as shown in Fig. 11. Multilamellar and unilamellar nanosheets were obtained via this synthetic strategy. The multilamellar nanosheets exhibited 20–40 nm thickness, which consisted of alternate layers of MFI zeolite framework and surfactant micelles with 2.0 nm and 2.8 nm thickness respectively. A one-step approach for synthesis of mesoporous zeolites with the use of an amphiphilic organosilane as the mesopore template was also developed by Ryoo’s group.112 The required amphiphilic organosilanes were comprised of a hydrolysable methoxysilyl moiety, a long-chain hydrophobic group and a quaternary ammonium moiety functioning as the zeolite SDA. The synthesized mesoporous ZSM-5 exhibited globular morphologies with rugged surfaces and a uniform mesopore size, which was able to be finely tuned in the range of 2–8 nm by changing the chain length of the organosilane and the hydrothermal synthesis temperature. Additionally, the group of Pinnavaia used a silane-functionalized polyethylenimine as the mesoporogen for constructing intracrystalline mesopores within the ZSM-5 zeolite.113 The addition of the silylated polymer to the zeolite precursor gel allowed it to be grafted on

Fig. 11 Crystallization of MFI nanosheets. (A) Proposed structure model for the single MFI nanosheet. Surfactant molecules are aligned along the straight channel of MFI framework. Two quaternary ammonium groups (indicated as a red sphere) are located at the channel intersections; one is inside the framework, and the other is at the pore mouth of the external surface; (B) Many MFI nanosheets form either multilamellar stacking along the b-axis; (c) A random assembly of unilamellar structure. Reprinting with permission from Choi, M.; Na, K.; Kim, J.; Sakamoto, Y.; Terasaki, O.; Ryoo, R. Stable Single-Unit-Cell Nanosheets of Zeolite MFI as Active and Long-Lived Catalysts. Nature 2009, 461(7261), 246–249. Copyright©2009, Springer Nature.

Mesostructured materials

51

the zeolite surface through covalent Si–O–Si linkages in the nucleation stage. The grafted polymer became phase-segregated from the zeolite matrix as the zeolite crystal grew, constructing an intracrystal polymer network linked to the zeolite framework in the crystal. The ZSM-5 zeolite with uniform intracrystal mesopores in the range of 2–3 nm was obtained after calcination. In addition to the hard and soft templating, several non-templating synthetic methods for creating mesopore structure without using mesopore templates have also been explored.114 In this regard, the intergrowth of the zeolite nanocrystals without the participation of mesopore templates accounts for giving rise to mesoporosity (Fig. 12). Steam-assisted crystallization (SAC) was used to synthesize mesoporous zeolites without using any mesopore templates. In the presence of tetraethylammonium hydroxide (TEAOH), a common zeolite SDA, mesopore zeolite Beta was synthesized by Bein et al. through employing the SAC procedure.115 This was achieved by a dense-gel synthesis to induce a burst of nucleation. The final self-sustaining zeolite Beta aggregates were assembled from 20 nm crystalline domains, resulting in a mesoporous structure with dominant pore sizes of around 13 nm. A quasi-solid-state reaction was also employed to synthesize hierarchically micro-meso-macroporous zeolites by Yang and Su et al., affording TS-1 aggregates with pores at three length scales.116 Additionally, a new synthetic concept named “nanofusion” was raised by Möller and Bein et al. towards the preparation of mesopore zeolite Beta.117 A concentrated precursor gel containing TEAOH as the SDA was firstly hydrothermally converted to nanosized Beta crystals with sizes in the range of 20–40 nm, which was then instantly fused into stable zeolite aggregates after drying and calcination, showing tunable mesopore sizes from 15 to 35 nm. Another strategy named “Repetitive Branching” was also used during one-step hydrothermal crystal growth by Tsapatsis’s group to generate a new mesoporous MFI-type zeolite made of orthogonally connected microporous nanosheets.118 The house-of-cards arrangement of the nanosheets formed a permanent network of 2–7 nm mesopores, along with the high external surface area and reduced micropore diffusion length, accounting for high reaction rates for bulky molecules. Besides, seed-directed synthesis (SDS) strategy was recently adopted by Xiao’s group for the synthesis of mesoporous Al-rich MFI zeolite in the absence of organic template.119

6.03.2.5.2

Top-down synthetic strategy via demetallization

Demetallization, namely removing framework T atoms of the zeolite, constitutes a simple and direct approach to create mesoporosity in zeolite. Specifically, dealumination or desilication leaching procedures can be carried out to generate amorphous areas in the zeolite framework, which could be further transformed to the mesopores upon extraction of the amorphous debris. Steaming at high temperature and acid leaching are the most widespread techniques for dealumination. The extraction of aluminum atom is caused by the cleavage of the SieOeAl bonds in zeolite through hydrolysis under certain conditions, leading to the formation of voids and a partial collapse of the zeolite structure. Steaming, a hydrothermal treatment, is generally carried out at temperature above 500
C in the presence of steam. The fabrication of mesopores via steaming is widely considered to be highly dependent on the Al content and the stability of Al sites. Thus, zeolites with low pristine Si/Al, such as Y and Beta zeolite, are usually chosen as the substrates for steaming. To gain a deeper understanding of the dealumination process of Y zeolite via steaming, Lamberti and van Bokhoven et al. investigated it with in situ, time-dependent, synchrotron radiation XRPD and in situ Al K-edge XAS.120 It was found that considerable structural collapse caused by steaming did not occur at the highest temperature, however, defects formed at much lower temperature when water was able to enter the pores again, leading to significant migration of framework Al3þ to extraframework positions. Desilication has also become a versatile route to tailor zeolites with a mesoporous framework in the presence of a base. Matsukata et al. were the first authors who realized that mesopores in ZSM-5 zeolite can be acquired through alkaline treatments.121 It was found that the alkaline treatment dissolved roughly 40% of the ZSM-5 crystals and led to a distinct increasement of the mesopore

Fig. 12 Synthetic routes to prepare mesoporous zeolites through nanozeolite assembly. Adapted with permission from Möller, K.; Bein, T. MesoporositydA New Dimension for Zeolites. Chem. Soc. Rev. 2013, 42(9), 3689–3707. Copyright©2013, The Royal Society of Chemistry.

52

Mesostructured materials

volume in the zeolite, whereas the crystallinity of the zeolite was preserved well. In order to optimize the mesopore formation in MFI zeolites, a deep investigation of base treatment conditions was conducted by Groen and Pérez-Ramírez et al.122,123 It was found that a molar Si/Al ratio in the range of 25–50 was identified as the optimal window for the pristine zeolite, leading to an optimal mesoporosity centered around 10 nm and mesopore surface areas of up to 235 m2 g1 while preserving the intrinsic crystalline and acidic properties, as shown in Fig. 13. Organic additives, such as tetrapropyl ammonium and tetrabutyl ammonium, were also found their function acting as the external pore directing agents to regulate the intracrystalline mesoporosity during the base etching process. The specific interaction of the organic additives with the zeolite surface under alkaline conditions was studied by the group of Pérez-Ramírez.124 This interaction can provide an adjustable protection against zeolite dissolution and enable the creation of similar mesopore surface areas of the zeolite with a smaller mesopore size of around 5 nm and less damage to the original micropore volumes when compared to the one treated solely by an alkaline. Due to the repulsion between the negatively charged lattice and OHin ZSM-5 zeolite with lower Si/Al ratios, minor extraction of framework silicon and limited mesopore formation occurs, whereas at higher Si/Al non-selective and excessive extraction appears, affording substantial mesopore formation at the expense of a severe crystallinity loss. Boundaries, intergrowths, and defects within each zeolite particle are much more susceptible to dissolution or etching. Desilication unavoidably modulates the Si/Al ratio of the zeolite framework just like dealumination, whereas the desilication results in a decrease in Si/Al ratio. Owing to the existence of re-alumination it is most likely to observe some extra-framework aluminum species after the base leaching, introduction of an additional acid treatment or ion-exchange step thus seems necessary to get rid of these species for unblocking the micropores and mesopores.

6.03.2.5.3

Mixed synthetic strategy

In order to finely control mesoporosity and other influential properties in mesoporous zeolite preparation, several synthetic works involve the combination of both bottom-up and top-down methods. The most representative example is the method for mesoporous zeolite preparation following a dissolution-recrystallization route, in which dissolution achieved through a base treatment of zeolites is most often carried out in the presence of surfactants. Tsapatsis and Stein et al. synthesized ZSM-5 zeolite crystals with secondary mesopores by alkaline desilication and surfactantinduced reassembly of dissolved species originating from the pristine ZSM-5 crystals.125 The synthesized mesoporous ZSM-5 product exhibited a dual-mesopore size distribution, well-preserved crystallinity, and acidity as well as enhanced external surface areas. The textures of the products revealed that desilication and reassembly occurred simultaneously to generate a dual-pore system. The desilication process created large mesopores (10–30 nm) in the zeolite crystals, while reassembly of the dissolved species, consisting of silicates, aluminosilicates, and fragments of the ZSM-5 crystals, by micellization of a surfactant produced

Fig. 13 Simplified schematic representation of the influence of the Al content on the desilication treatment of MFI zeolites in NaOH solution and the associated mechanism of pore formation. Reprinting with permission from Groen, J.C.; Jansen, J.C.; Moulijn, J.A.; Pérez-Ramírez, J. Optimal Aluminum-Assisted Mesoporosity Development in MFI Zeolites by Desilication. J. Phys. Chem. B 2004, 108(35), 13062–13065. Copyright©2004, American Chemical Society.

Mesostructured materials

53

smaller mesopores ( 3 nm). In addition, mesoporous zeolite Y with uniformly distributed intracrystalline mesopores of around 4 nm was prepared by García-Martínez and Ying et al. through a hydrothermal treatment of the parent zeolite Y with a mixture of diluted NH4OH and cetyltrimethylammonium bromide surfactant at 150
C.126 The mesopores can be tuned in the range of 2.5–4.5 nm by varying the chain length of the surfactants. The introduction of controlled mesoporosity into crystalline zeolites proceeded presumably via a surfactant-assisted crystal rearrangement mechanism, as shown in Fig. 14. Some of the Si–O–Si linkages were first cleaved in basic reaction conditions to provide some flexibility in the crystalline structure, and resulted in the creation of negatively charged sites in the zeolite framework that attracted cationic surfactants. Electrostatic interactions between the negatively charged sites and the positively charged surfactants, and the formation of micelles within the zeolite crystals through self-assembly of surfactant cations led to the rearrangement of the crystal structure to generate mesopores around the micelles. This process allowed an almost complete recovery of the zeolite Y through effectively preventing the dissolution of the zeolite crystals.

6.03.3

Catalytic applications of mesoporous materials

Catalysts play a key role in chemical processes and are widely utilized to produce over 90% of chemical products in petroleum refinery and fine chemical industries. Recently mesoporous catalysts are of great scientific and technological interest in consideration of their enhanced performances in mass diffusion and transfer as well as their high surface area. These mesoporous catalysts are generally classified into four categories based on the types of materials, which include (i) mesoporous metal and metal oxides; (ii) representative ordered mesoporous materials (OMSs, OMOSs and OMCs); (iii) mesoporous MOFs; (iv) mesoporous zeolites. In this section, catalytic applications of mesoporous materials, mainly focusing on prominent recent examples in catalytic reactions, will be discussed.

6.03.3.1

Mesoporous metals and metal oxides for catalysis

Metals and metal oxides with mesoporous structures are of great interest due to their high catalytic activity as well as excellent thermal and chemical stability. Mesoporous metals can be used alone as catalysts, while mesoporous metal oxides can be used not only alone but also as catalyst supports to immobilize noble metals. In 2011, monodispersed polyhedral- and olive-shaped mesoporous Pt nanoparticles with uniform particle sizes, prepared through using KIT-6 and SBA-15 as the hard templates, by Yamauchi et al. was investigated as a catalytic electrode for methanol electrochemical oxidation reaction.34 The study indicated that small-sized meso-Pt (deposited for 10 min) showed a remarkably larger current density of the electrocatalytic methanol oxidation when compared with meso-Pt with large sizes (deposited for 1 and 12 h) and commercially available Pt black catalyst (Fig. 15). The onset potentials of meso-Pt were found to be clearly negatively shifted in comparison to commercially available Pt catalysts (Fig. 15B). Additionally, small-sized meso-Pt showed high durability in its catalytic performance, which was demonstrated by chronoamperometric curves recorded at 0.6 V (Fig. 15C). The acquired catalytic results revealed that small-sized meso-Pt provided more readily accessible catalytic active sites to methanol reactant through improving accessibility of the inner mesopores, therefore facilitating methanol’s diffusion in the catalyst. In addition, mesoporous Pt films with accessible pores and controlled mesopore sizes, synthesized using diluted surfactant solutions, were also found their accelerating effect on the methanol oxidation reaction 1 year later.37 A number of metal oxides have been extensively investigated for various catalytic reactions.127 For instances, mesoporous CeO2 synthesized using an ordered silica KIT-6 as the hard template, showed a superior catalytic activity in CO oxidation compared with CeO2 nanoparticles.128 The enhanced catalytic performance of the mesoporous catalyst was attributed to its large effective surface area. Additionally, the synthesis of Co3O4 hollow spheres with tunable hierarchical pores and high surface area was reported by

Fig. 14 Schematic of the speculated zeolite mesopore formation process: (A) Original zeolite Y; (B) SieOeSi bond opening/reconstruction in basic media; (C) Crystal rearrangement to accommodate the surfactant micelles; (D) Removal of the template to expose the mesoporosity introduced. Reprinting with permission from García-Martínez, J.; Johnson, M.; Valla, J.; Li, K.; Ying, J.Y. Mesostructured Zeolite YdHigh Hydrothermal Stability and Superior FCC Catalytic Performance. Cat. Sci. Technol. 2012, 2(5), 987–994. Copyright©2012, The Royal Society of Chemistry.

54

Mesostructured materials

Fig. 15 Demonstration of methanol oxidation activity of meso-Pt with different particle sizes in 0.5 M H2SO4 containing 0.5 M methanol. (A) Cyclic voltammograms for small-sized meso-Pt (deposited for 10 min) and Pt black. (B) Linear sweep voltammograms measured. (C) Chronoamperometric curves at 0.6 V. (D) Summary on the specific activities at 0.5 V and the onset potentials for all the samples. Reprinting with permission from Beck, J.S.; Vartuli, J.C.; Roth, W.J.; Leonowicz, M.E.; Kresge, C.T.; Schmitt, K.D.; Chu, C.T.W.; Olson, D.H.; Sheppard, E.W.; McCullen, S.B.; et al. A New Family of Mesoporous Molecular Sieves Prepared With Liquid Crystal Templates. J. Am. Chem. Soc. 1992, 114(27), 10834–10843. Copyright©2011, American Chemical Society.

Wang and An et al.129 By tuning the shell thickness and mesoporous structure via proper pre-treatment of the carbon spheres with alkali/acid, the obtained Co3O4 hollow spheres exhibited an outstanding catalytic performance and long lifetime in methane combustion due to their abundance of well interconnected mesopores and high accessible surface area. Promoting metal-support interaction by using mesoporous structures with a large surface area as a noble metal support is considered to be one of key factors defining catalyst performance. In 2013, Somorjai et al. utilized mesoporous Co3O4, NiO, MnO2, Fe2O3 and CeO2 as supports for a Pt catalyst and studied the effect of metal-support interaction on CO oxidation performance.130 In Fig. 16A–F, representative TEM images of KIT-6 silica template, mesoporous Co3O4, and Pt-loaded Co3O4 catalysts and a general nanocasting method for preparing ordered mesoporous oxides are presented. It was found that pure Co3O4 and NiO provided higher TOFs when compared to MnO2, Fe2O3 and CeO2 under both oxidizing and reducing conditions (Fig. 16G). The TOFs of MnO2, Fe2O3 and CeO2 under reducing conditions showed much higher than those under oxidizing conditions. Additionally, metal oxide-supported Pt exhibited a much higher activity than mesoporous SBA-15 silica supported Pt, implying a significant interaction between the Pt nanoparticles and the metal oxides. A comparison of various mesoporous metal oxide-supported Pt catalysts in CO oxidation was also conducted, as shown in Fig. 16H. The observed catalytic TOF of Pt/Co3O4 was significantly higher than those of other metal oxide-supported Pt catalysts. The ordered mesoporous architecture of the metal oxide supports rendered greater surface areas including internal and external ones for effective dispersion of Pt nanoparticles, leading to generate a larger oxide-metal interface.

Mesostructured materials

55

Fig. 16 TEM image of (A) the mesoporous silica template and (B) the resulting Co3O4 replica. TEM (C) and HR-TEM (E) image of Pt/Co3O4 catalysts and (D) their corresponding energy-dispersive spectroscopy (EDS) phase mapping. (F) Illustration of the hard-templating approach for the preparation of mesoporous-oxide-supported Pt nanoparticle catalysts. CO oxidation over (G) pure mesoporous oxides and (H) Pt-nanoparticle-loaded oxide catalysts. Adapted from with permission An, K.; Alayoglu, S.; Musselwhite, N.; Plamthottam, S.; Melaet, G.; Lindeman, A.E.; Somorjai, G.A. Enhanced CO Oxidation Rates at the Interface of Mesoporous Oxides and Pt Nanoparticles. J. Am. Chem. Soc. 2013, 135(44), 16689–16696. Copyright©2013, American Chemical Society.

Mesoporous Ta2O5 and Nb2O5 with sulfuric acid or phosphoric acid treatment were investigated for benzylation or alkylation reactions.131,132 For instance, by varying the carbon chain length of alkyl amine templates (C6, C12 and C18), ordered mesoporous Ta2O5 with different pore sizes was prepared by Antonelli et al.132 The obtained mesoporous Ta2O5 was further treated with sulfuric acid to afford the desired acidic Ta2O5 catalyst. Among all these acidic catalysts, the 1-dodecylamine templated Ta2O5 catalyst exhibited the highest selectivity for alkylation of benzene with dodecene, which was ascribed to the shape selectivity of the catalyst. In the reaction of 1-hexene isomerization to trans-2-hexene, sulfuric acid treated mesoporous Ta2O5 catalyst synthesized by using

56

Mesostructured materials

1-dodecylamine as the template afforded not only the highest catalytic activity but also the best selectivity, providing trans-2hexene/cis-2-hexene ratio of around 4 after reaction at 70
C for 6 h, which was much higher than those of acidic mesoporous Ta2O5 catalysts templated by n-hexylamine and n-octadecylamine (< 1).133

6.03.3.2

Catalytic applications of representative ordered mesoporous materials

6.03.3.2.1

Ordered mesoporous silicas for catalysis

6.03.3.2.2

Ordered mesoporous organosilicas for catalysis

The first efforts of utilizing OMSs in the field of heterogeneous catalysis are to exploit the large pore sizes of the catalysts to convert large substrates that are not able to enter the pores of microporous materials. Anchoring functional groups on the surface of silicas presents an effective way to prepare functionalized OMSs. It is worth noting that two separated functions can be introduced to the OMSs at the same time, which are not possible to coexist in the homogeneous phase due to the fact that the active groups would interact or react with each other and get defunctionalized in the end. For instance, bifunctional mesoporous silicas with both sulfonic acid and amino groups were prepared by Huang et al. through co-condensation of one of the two functional groups onto the internal channels and subsequent grafting of the second group onto the external surface (Fig. 17A).134 Only the external surface would be exposed to the second amino grafting reagent because of the pre-existence of template CTAB in the mesoporous channels. The bifunctional mesoporous silica with sulfonic acid on its internal surface and amino groups on its external surface was labeled as SAMSN-AP. Likewise, another bifunctional mesoporous silica with amino groups on its internal surface and sulfonic acid groups on its external surface was named as APMSN-SA. The bifunctional catalysts were investigated in a cascade reaction, which was comprised of two separate reactions including an acid-catalyzed hydrolysis of an acetal towards an aldehyde and a basecatalyzed Henry reaction (Fig. 17B). Both SAMSN-AP and APMSN-SA catalysts exhibited excellent catalytic performances, affording conversion of acetal reactant (17a) and yield of the desired nitroalkene product (17c), 100% and 98% respectively.

Metal or metal complex supported OMOSs are typical functionalized OMOS catalysts for various reactions. Metal sites can be incorporated in OMOS materials via an isomorphic substitution approach, which involves adding hydrolysable metal precursors to the synthetic gel during the synthesis process and results in the incorporation of isolated tetrahedral metal sites in the OMOSs. Many metals, including Sn, Ti, V, Cr, Al and Nb, have already been incorporated in OMOSs, particularly ethane bridged ones.57 The catalytic activity per metal center is found to be generally high owing to the hydrophobic microenvironment generated by the organic bridges. The positive effect on the selectivity is also obviously observed. For instance, hydrophobized Ti-containing OMOSs showed significantly higher catalytic activities in the epoxidation of 1-octene with hydroperoxides in comparison with Ti-SBA-15.135 Their

Fig. 17 (A) Schematic illustration of the synthesis of bifunctional mesoporous silica nanoparticles with sulfonic acid groups on the internal surface and organic amine groups on the external surface. (B) One-pot cascade reaction composed of acid-catalyzed hydrolysis and base-catalyzed Henry reaction. Adapted with permission from Huang, Y.; Xu, S.; Lin, V.S.Y. Bifunctionalized Mesoporous Materials With Site-Separated Brønsted Acids and Bases: Catalyst for a Two-Step Reaction Sequence. Angew. Chem. Int. Ed. 2011, 50(3), 661–664. Copyright©2011, Wiley-VCH.

Mesostructured materials

57

catalytic behavior was found to be correlated with the hydrophobicity of the catalyst, being benefited greatly from the hydrophobic environment. In addition, a strategy for heterogenization of NHC ligands was reported by Van Der Voort et al. recently.136 As shown in Fig. 18, Au-NHC complexes were anchored via a post-modification process after the synthesis of OMOSs support, which was an ethane bridged material bearing sulfonic acid groups. The sulfonic acid groups possessed a high affinity to Au and reacted with OH moiety of the [Au(OH)(NHC)] complex via an acid-base reaction. The resulting material acted as a boomerang-type catalyst; that is, it allowed the active species to detach from the catalyst surface to participate the catalysis process and recombined with the solid after consuming all starting material. The resulting catalyst exhibited excellent activity for the hydration of diphenylacetylene and could be readily recycled and reused in several catalytic runs without decomposition or degradation of the complex.

6.03.3.2.3

Ordered mesoporous carbons for catalysis

The OMC-based catalysts have been employed in several organic transformations such as selective oxidation, dehydrogenation and base-catalyzed reactions. Mesoporous carbon itself can act as a catalyst in some certain reactions, presenting a perfect alternative for current metal oxide-based catalysts in certain cases. For instance, carbon materials, particularly OMCs without metal particle functionalization, have exhibited excellent catalytic performances in both selectivity and activity for dehydrogenation reactions under milder reaction conditions than those for industrial catalysts.137 Surface basic oxygen groups formed during the reaction process were considered to function as the active sites for dehydrogenation of propane to propylene. In addition to oxygen, other elements, including N, B, S, P, Cl, I and Se, have also been doped into mesoporous carbon materials to afford a metal-free electrocatalyst for oxygen reduction reaction with enhanced performance compared with undoped carbon. For instance, mesoporous carbon nitride doped with boron and fluorine showed excellent performance in cyclohexane oxidation, providing superior cyclohexane conversion and selectivity for cyclohexanone product.138 In order to extend the application scope of OMCs, other functional groups were also introduced to their surface or framework by efficient methods, e.g., via metal loading or grafting through chemical or electrochemical reduction of aryl diazoniums, reductive alkylation and arylation. The presence of an adequate number of catalytically active sites exposed on mesoporous carbon materials makes them attractive catalysts in various reactions.139–142 For examples, OMCs bearing SO3H groups afforded remarkable catalytic performance in acid-catalyzed esterification of oleic acid with ethanol for biodiesel production.139 In addition, Zhao et al. prepared OMCs with uniform metal-containing nanoparticles via a chelate-assisted multicomponent co-assembly method and used them in Fischer-Tropsch (FT) synthesis (Fig. 19A).140 For the case of OMCs incorporated with Fe2O3 nanoparticles, the obtained nanocomposites possessed constant Fe2O3 contents of around 10 wt% and the size of Fe2O3 nanoparticles could be readily tuned from 8.3 to 22.1 nm. Interestingly, it was found that the carbon-based composite catalysts exhibited a unique semi-exposure structure, that is, the metal nanoparticles partially embedded in the carbon framework with the remaining part exposed in the mesoporous channels,

Fig. 18 Synthesis of the OMOSs-based gold(I)-NHC catalysts. Reprinting with permission from De Canck, E.; Nahra, F.; Bevernaege, K.; Vanden Broeck, S.; Ouwehand, J.; Maes, D.; Nolan, S.P.; Van Der Voort, P. PMO-Immobilized AuI–NHC Complexes: Heterogeneous Catalysts for Sustainable Processes. ChemPhysChem 2018, 19(4), 430–436. Copyright©2018, Wiley-VCH.

58

Mesostructured materials

Fig. 19 (A) Illustration of the chelate-assisted co-assembly route for the synthesis of OMC nanocomposites incorporated with metallic nanoparticles. (B) Catalytic performances of the iron-based nanoparticles incorporated OMC in FT synthesis; (a) plots of C5 þ selectivity and methane yield as a function of Fe2O3 particle size, (b) plots of CO conversion and olefin/paraffin ratio as a function of Fe2O3 particle size, (c) conversions of CO and H2 vs reaction time over a period of 100 h on stream, (d) TEM image of the used catalyst Fe-C-8. Adapted with permission from Sun, Z.; Sun, B.; Qiao, M.; Wei, J.; Yue, Q.; Wang, C.; Deng, Y.; Kaliaguine, S.; Zhao, D. A General Chelate-Assisted Co-Assembly to Metallic NanoparticlesIncorporated Ordered Mesoporous Carbon Catalysts for Fischer-Tropsch Synthesis. J. Am. Chem. Soc. 2012, 134(42), 17653–17660. Copyright©2012, American Chemical Society.

Mesostructured materials

59

which not only provided a confinement effect as well as adequate exposed surface for catalysis but also contributed to the trapping of nanoparticles and preventing it from aggregating during catalysis. The catalytic results of FT synthesis indicated that the catalytic performance of the mesoporous Fe-carbon nanocomposite catalysts significantly improved as the decrease of the size of Fe nanoparticles, giving up to 68% selectivity of C5 þ (Fig. 19B).

6.03.3.3

Mesoporous metal-organic frameworks for catalysis

MOFs generated through the strong coordination bonds between metal ions and organic linkers provide a feasibility to easily identify and analyze the numeral and spatial status of the active sites. The highly crystalline structure of MOFs ensures the stability and durability of the surface catalysts, as their highly porous volumes, especially for mesoporous MOFs, provide a large area to accommodate catalytic sites.78 High surface areas and tunability as well as large channels enable MOFs to exhibit excellent and unique properties in a variety of catalytic applications.77 Mesoporous MOFs usually outperform the microporous counterparts in certain catalytic reactions. For example, using mesoporous NU-1000 as the cobalt catalyst supporter could readily facilitate oxidative dehydrogenation of propane at low temperature.143 In addition to the conventional methods of using MOFs to load metal nanoparticles either through immersing metal into the mesoporous structure or encapsulating metal with MOF coating,144,145 multishelled hollow in MOFs was also ingeniously utilized to accommodate Pt nanoparticles as catalysts. Sandwich-like MIL-101@Pt@MIL-101 nanostructures encapsulating Pt were used as the catalysts for selective hydrogenation of a,b-unsaturated aldehydes to unsaturated alcohols, which was reported by Tang et al. recently.146 By growing Pt nanoparticles on the MIL-101 platform, a unique two-layer structure was first generated, which was further encapsulated with MIL-101 to afford the desired MIL-101@Pt@MIL-101 catalysts. Due to the special contact form between Pt nanoparticles and MIL-101 structures as well as the presence of large pores for efficient diffusion of reactant and product, these catalysts provided exceptionally high conversion efficiency and product selectivity in the conversion of a range of a,b-unsaturated aldehydes, including cinnamaldehyde, furfural, 3-methyl-2-butenal and acrolein, to unsaturated alcohols. Moreover, the catalysts also rendered extraordinary recuperability, maintaining the same high level of reactant conversion and product selectivity even after five successive catalytic runs in the case of using MIL-101(Cr)@Pt@MIL-101(Fe) catalyst. Mesoporous MOFs offer abundant voids for the deposition of various metal nanoparticles to generate composite catalysts, which exhibit outstanding catalytic performance due to their high surface area and adjustable pore sizes to ensure sufficiently exposed catalytic active sites and ignorable diffusion resistance for reactant and product to reach the surface of active sites. Astruc et al. synthesized non-noble metal nanoparticles using ZIF-8 as the MOF template and used them in the hydrolysis of ammonia borane for hydrogen production.147 Metal nanoparticles were first well incorporated into the cages of as-synthesized ZIF-8 via a deposition-precipitation process. The nucleation and growing of metal nanoparticles can be controlled due to the confinement and stabilization effect of the porous structure. Among all catalysts tested, Ni/ZIF-8 catalyst showed the highest catalytic activity, 1 , which outperformed even some noble metal-based catalysts. Mechanistic studies affording a TOF value of 85.7 molH2 molcat particularly using kinetic isotope effects revealed that the cleavage by oxidative addition of an OeH bond in H2O is the ratedetermining step in the hydrolysis reaction (Fig. 20A). Thus, an attractive and effective “on-off” control of hydrogen production by simply altering the addition of OH and Hþ was proposed (Fig. 20B). The astonishing high TOF achieved by Ni/ZIF-8 indicated a promising potential of this material for chemical storage of fuel-cell-based hydrogen. The research on MOFs decorated with metal nanoparticles so far mainly focuses on their synthetic design and catalytic performance, scarce studies have been reported on precise control of the distribution and diffusion of nanoparticles inside MOFs. Owing to the molecular sieving effects endowed by the porous framework of MOFs, MOFs with metal nanoparticles incorporated inside the pores demonstrate excellent catalytic properties. Yolk-shell structure composition based on mesoporous ZIF-8 presents a typical example of these materials. Nevertheless, the narrow channels in MOFs hinder the molecular diffusion and therefore give rise to a low catalytic efficiency. Lu et al. recently found that the catalytic activity increased as metal nanoparticles spatially distributed

Fig. 20 (A) Proposed mechanism for the hydrolysis of ammonia borane catalyzed by Ni/ZIF-8. (B) “On-off” control of H2 production in the ammonia borane hydrolysis in water. Adapted with permission from Wang, C.; Tuninetti, J.; Wang, Z.; Zhang, C.; Ciganda, R.; Salmon, L.; Moya, S.; Ruiz, J.; Astruc, D. Hydrolysis of Ammonia-Borane over Ni/ZIF-8 Nanocatalyst: High Efficiency, Mechanism, and Controlled Hydrogen Release. J. Am. Chem. Soc. 2017, 139(33), 11610–11615. Copyright©2017, American Chemical Society.

60

Mesostructured materials

as close as possible to the MOF surface.148 Moreover, the study indicated that spatial distribution of metal nanoparticles in MOFs can be controlled through varying the concentration of organic ligands employed and the reaction period. Metal oxides with nanoparticles supported on the surface were used as the sacrificial template in the synthesis. The strategy proposed in this study demonstrated how the MOF based metal catalysts can be well regulated by controlling the conditions for MOF nucleation, therefore significantly improving catalytic performance in selective hydrogenation of olefin mixtures.

6.03.3.4

Mesoporous zeolites for catalysis

The presence of additional mesopores in zeolites typically accounts for the often observed improvement in activity, stability and longer catalytic lifetime in catalytic reactions. Over the years, mesoporous zeolites have been employed in a variety of reactions, such as acid-catalyzed condensation, methanol conversion, Fischer-Tropsch (FT) reaction, isomerization, hydrogenation, and have exhibited superior catalytic performance. The synthesized mesoporous zeolites were initially investigated in the catalytic reactions involving bulky organic molecules. Taking the condensation of benzaldehyde with bulky alcohols as an example, Ryoo et al. demonstrated superior catalytic performances of mesoporous ZSM-5 nanosheets with both multilamellar and unilamellar architectures in these condensation reactions.111 These MFI nanosheet catalysts were synthesized via an appropriate design of bifunctional surfactants and their introduction effect on simultaneously forming zeolite structures on the mesoporous and microporous length scales, affording much higher activities than conventional ZSM-5 zeolite in the condensation of benzaldehyde with bulky alcohols (Table 1). The enhancement in catalytic activity was attributed to a large number of acid sites located at the mesopore surface of ZSM-5 nanosheets. Compared to multilamellar zeolite, the unilamellar ZSM-5 exhibited higher activity due to its larger external surface area. A further improvement in catalytic activity was subsequently achieved by Xiao et al. through employing b-axis-aligned mesoporous ZSM-5 for the condensation reaction.110 ZSM-5 single crystals with b-axis-aligned mesopores were prepared using a designed cationic amphiphilic copolymer as a mesoscale template, providing superior catalytic activities to conventional ZSM-5 and even ZSM-5 nanosheets. For instance, in the condensation of benzaldehyde with hydroxyacetophenone, conventional ZSM-5 and randomly oriented mesoporous ZSM-5 (ZSM-5-M) showed low (19%) and medium (46%) activities respectively. In contrast, ZSM-5 single crystals with b-axis-aligned mesopores (ZSM-5-OM) were highly active in this condensation, exhibiting as high as 91% conversion, which was even higher than the activities from multilamellar and unilamellar ZSM-5 nanosheets with conversion of 48% and 76% respectively (Table 1). In consideration of the similarities of all zeolites in terms of Si/Al ratio, aluminum distribution and acidic strength, the superior catalytic activities of ZSM-5-OM should be directly assigned to the contribution of b-axisaligned mesopores in zeolite. In order to take advantage of the special properties of mesoporous zeolites, such as high surface area, acidity and large pore channels for diffusion, attempts of using mesoporous zeolites for converting methanol to either gasoline, olefins or aromatics have been extensively explored.149 For instance, hierarchical SSZ-13 zeolite was prepared with the aid of a surfactant mesoporogen and studied Table 1

Catalytic conversion of bulky substrates over various ZSM-5 catalysts.

Conversion (%) Catalyst

a

ZSM-5 Multilamellar ZSM-5 nanosheet Unilamellar ZSM-5 nanosheet ZSM-5-M ZSM-5-OM a

The Si/Al ratios of all zeolite catalysts are in the range of 37–53. Protection of benzaldehyde with pentaerythritol (PBP). c Condensation of benzaldehyde with hydroxyacetophenone (CBH). b

PBP 13 86 86 50 88

b

CBH c

References

19 48 76 46 91

110 111 111 110 110

Mesostructured materials

61

by Hensen and coworkers in the methanol-to-olefins (MTO) reaction.150 The synthesized SSZ-13 zeolites possessed trimodal porosity consisting of native micropores of the CHA framework, larger micropores and mesopores, which attributed to a substantial improvement of mass transport inside the zeolite crystals. The mesoporous SSZ-13 zeolites exhibited a lower rate of deactivation in the MTO reaction when compared with the bulk SSZ-13 zeolite. Confocal fluorescence microscopy indicated that aromatic species that gave rise to deactivation were evenly deposited throughout the crystals of such mesoporous SSZ-13 zeolites, implying the fact that micropore space was well utilized. FT synthesis is considered to be a recognized route to produce liquid transportation fuels.151 A range of products can be produced during the FT reaction. So far mesoporous zeolites such as ZSM-5, Beta, and Y, supported Co or Ru have been considerably reported to be efficient FT catalysts.152–158 Ru/meso-ZSM-5 catalyst was prepared and first employed in the FT reaction by Wang and Zhang et al.152 This catalyst exhibited a highly selectivity for the production of C5-C11 isoparaffins from syngas. The selectivity reached around 80% with a ratio of isoparaffins to n-paraffins of about 2.7, which was significantly higher than the maximum expected from Anderson-Schulz-Flory distribution ( 45%). It was further revealed that the hierarchical porous structure and the relatively weak acidity, mainly Lewis acidity, contributed to the enhancement of selectivity to C5-C11 and the decrease in selectivity to light hydrocarbons (CH4 and C5-C11 alkanes). Later on the authors further extended the mesoporous zeolite support to Beta and Y zeolites, which led to Ru/meso-Beta and Co/Na-meso-Y catalyst respectively, affording excellent selectivity to the specific length of alkanes in FT reactions.153,159 Ryoo et al. used mesoporous MFI zeolite nanosponge supporting Co nanoparticles, Co/NS-MFI, as the FT catalyst in the syngas conversion.154 The mesoporous MFI nanosponge was synthesized by a seed-assisted hydrothermal preparation route with the use of C22H45-Nþ(CH3)2-C6H12-Nþ(CH3)2-C6H13 as the SDA, possessing disordered network of 2.5nm-thick MFI zeolite nanolayers with a narrow distribution of mesopore sizes of around 4 nm. The Co/NS-MFI zeolite presented high catalytic conversion of CO to hydrocarbons and long catalytic lifetime in the FT synthesis due to the excellent stability of the Co nanoparticles against sintering (Fig. 21A). When compared with conventional alumina-based catalysts, the Co/NS-MFI catalyst demonstrated higher selectivity for branched hydrocarbons in the gasoline range, achieving 74% gasoline selectivity at 82% CO conversion (Fig. 21B), which was ascribed to the short diffusion path lengths in the extremely thin zeolite frameworks for branched hydrocarbons during hydroisomerization reactions. Catalytic isomerization of hydrocarbon substrates into value-added chemicals is always highly desirable in the petrochemical industry. Zeolites with mesopore structure are the catalysts of choice for this transformation. For instance, Somorjai et al. investigated several types of aluminosilicate catalyst for the isomerization of n-hexane, both with and without Pt nanoparticles.160 The catalysts that were studied included MCF-17 mesoporous silica, an aluminum modified MCF-17 (Al-MCF-17/Pt), mesoporous MFI and BEA type zeolites. Al-MCF-17 catalyst was found to provide the best catalytic performance, showing the highest isomer production, which was due to the mild Lewis acid and Brønsted acid sites supplied by the aluminum modification. The study indicated that Al-MCF-17/Pt catalyst exhibited higher activity and selectivity in the reaction than the other zeolitic counterparts. The isomer selectivity for each catalyst with the same Si/Al ratio is shown in Fig. 22A and B. For the Al-MCF-17/Pt catalyst, the Pt loading played an important role in activity and isomer selectivity (Fig. 22C and D). If the Pt loading was too low, activity was too low to react the adsorbed reaction intermediates. If the Pt loading was too high, catalysis dominated the reaction process and selectivity came to harm. Al-MCF-17 with a Si/Al ratio of 8 as well as loaded with 0.5 wt% Pt was found to be the optimum catalyst for isomerization of n-hexane.

Fig. 21 (A) Schematic diagram presenting FT reaction over Co/NS-MFI. (B) Product selectivity of Co/NS-MFI, Co/B-MFI and Co/g-Al2O3 catalysts, which were averaged over 100 h of the FT reaction time. Reprinting with permission from Kim, J.C.; Lee, S.; Cho, K.; Na, K.; Lee, C.; Ryoo, R. Mesoporous MFI Zeolite Nanosponge Supporting Cobalt Nanoparticles as a Fischer-Tropsch Catalyst With High Yield of Branched Hydrocarbons in the Gasoline Range. ACS Catal. 2014, 4(11), 3919–3927. Copyright©2014, American Chemical Society.

62

Mesostructured materials

Fig. 22 (A) Isomer selectivities for BEA/Pt zeolite (black) and Al-MCF-17/Pt (blue), both catalysts shown contain a Si/Al ratio of 8:1; (B) Isomer selectivities for MFI/Pt (red) and Al-MCF-17/Pt (green), both catalysts shown contain a Si/Al ratio of 36:1; (C) Effect of Pt nanoparticle loading on the isomer selectivity and (D) the overall hexane conversion mass activity for the Al-MCF-17 (8) supported material. Adapted with permission from Musselwhite, N.; Na, K.; Sabyrov, K.; Alayoglu, S.; Somorjai, G.A. Mesoporous Aluminosilicate Catalysts for the Selective Isomerization of N-Hexane: The Roles of Surface Acidity and Platinum Metal. J. Am. Chem. Soc. 2015, 137(32), 10231–10237. Copyright©2015, American Chemical Society.

Mesoporous zeolites decorated with metal can function as efficient catalysts for hydrogenation reactions. For examples, Xiao and Tang et al. carried out catalytic hydrogenation studies on aromatics, including naphthalene, pyrene and phenanthrene, with the use of mesoporous zeolite-supported metal catalysts.161–163 Pd/Beta-H, a catalyst with Pd metal supported on the mesoporous Beta zeolite, was found to exhibit better sulfur tolerance in the hydrogenation of naphthalene and pyrene when compared to Pd/AlMCM-41 catalyst. The study further revealed that the differences in the catalytic performance from these two catalysts were attributed to the discrepancy of the acidity of the two zeolite supports. In addition, Shape-selective hydrogenation of nitroarenes was investigated by Chen and Li et al. using Pd@mnc-S1 catalyst.164 The catalyst was elaborately designed with Pd nanoparticles encapsulated inside mesoporous MFI zeolite nanocrystals. The catalyst presented high stability and general shape-selectivity in the hydrogenation of nitroarenes due to its unique nanostructure and porosity.

6.03.4

Summary and perspective

Much promise for the development of novel heterogeneous catalysts has been held by the proper pore size of mesostructured materials with extremely high surface area. Well-isolated function sites with uniform properties can be generated via substitution of elements in inorganic walls or impregnation of active species with a pre-determined structure. Introduction of new functions can result from varying wall components or treatment of surface with special organic compounds. The diffusion of reactants and products can be facilitated in the mesopore channels. All these facts highlight the huge potential of mesoporous materials. In this article, we made an attempt to illustrate the broad horizons encompassed in the world of mesoporous materials. However, it is impossible to conduct a comprehensive survey covering all aspects of these astonishing materials. The topics discussed provide only a brief overview of currently representative mesoporous materials with prominent examples in terms of their synthesis and catalytic application. These mesoporous materials, including metals, metal oxides, silicas, organosilicas, carbons, MOFs and zeolites, have been mainly synthesized via hard templating or soft templating route. Superiority of these mesoporous materials in catalysis benefited from their unique properties has been demonstrated in a vast number of cases. Nevertheless, industrial applications of these materials are very limited so far, which is to a large extent related to their tedious synthesis procedures and high costs for preparation. The synthesis cost could not be offset by the enhanced catalytic performance they offer. The challenge to overcome this limitation will stimulate the researchers to develop innovative strategies for the preparation of mesoporous materials at low economic and environmental cost. Idea materials featuring an ordered system of mesopores as well as crystalline microporous walls are preferred, since they could encompass the advantages offered by the ordered array of regular mesopores (e.g., efficient mass diffusion) and by the crystalline structure (e.g., high thermal stability, easy characterization and control of active sites). We truly believe that plentiful novel research into the synthesis and application, especially in catalysis, of such materials, highlighted above as examples, will cast new light on rational developments for eventually practical applications.

References 1. Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck, J. S. Ordered Mesoporous Molecular Sieves Synthesized by a Liquid-Crystal Template Mechanism. Nature 1992, 359 (6397), 710–712. 2. Xu, R.; Pang, W.; Yu, J.; Huo, Q.; Chen, J. Chemistry of Zeolites and Related Porous Materials, John Wiley & Sons, Ltd: Chichester, UK, 2007.

Mesostructured materials

63

3. Zhao, D.; Wan, Y.; Zhou, W. Ordered Mesoporous Materials, Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2013. 4. Linares, N.; Silvestre-Albero, A. M.; Serrano, E.; Silvestre-Albero, J.; García-Martínez, J. Mesoporous Materials for Clean Energy Technologies. Chem. Soc. Rev. 2014, 43 (22), 7681–7717. 5. Thommes, M.; Kaneko, K.; Neimark, A. V.; Olivier, J. P.; Rodriguez-Reinoso, F.; Rouquerol, J.; Sing, K. S. W. Physisorption of Gases, With Special Reference to the Evaluation of Surface Area and Pore Size Distribution (IUPAC Technical Report). Pure Appl. Chem. 2015, 87 (9–10), 1051–1069. 6. Davis, M. E. Ordered Porous Materials for Emerging Applications. Nature 2002, 417 (6891), 813–821. 7. Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Fredrickson, G. H.; Chmelka, B. F.; Stucky, G. D. Triblock Copolymer Syntheses of Mesoporous Silica With Periodic 50 to 300 Angstrom Pores. Science 1998, 279 (5350), 548–552. 8. Che, S.; Garcia-Bennett, A. E.; Yokoi, T.; Sakamoto, K.; Kunieda, H.; Terasaki, O.; Tatsumi, T. A Novel Anionic Surfactant Templating Route for Synthesizing Mesoporous Silica With Unique Structure. Nat. Mater. 2003, 2 (12), 801–805. 9. Sakamoto, Y.; Kaneda, M.; Terasaki, O.; Zhao, D. Y.; Kim, J. M.; Stucky, G.; Shin, H. J.; Ryoo, R. Direct Imaging of the Pores and Cages of Three-Dimensional Mesoporous Materials. Nature 2000, 408 (6811), 449–453. 10. Xiao, C.; Fujita, N.; Miyasaka, K.; Sakamoto, Y.; Terasaki, O. Dodecagonal Tiling in Mesoporous Silica. Nature 2012, 487 (7407), 349–353. 11. Qiu, H.; Che, S. Chiral Mesoporous Silica: Chiral Construction and Imprinting Via Cooperative Self-Assembly of Amphiphiles and Silica Precursors. Chem. Soc. Rev. 2011, 40 (3), 1259–1268. 12. Innocenzi, P.; Malfatti, L.; Soler-Illia, G. J. A. A. Hierarchical Mesoporous Films: From Self-Assembly to Porosity With Different Length Scales. Chem. Mater. 2011, 23 (10), 2501–2509. 13. Wei, J.; Zhou, D.; Sun, Z.; Deng, Y.; Xia, Y.; Zhao, D. A Controllable Synthesis of Rich Nitrogen-Doped Ordered Mesoporous Carbon for CO2 Capture and Supercapacitors. Adv. Funct. Mater. 2013, 23 (18), 2322–2328. 14. Parlett, C. M. A.; Wilson, K.; Lee, A. F. Hierarchical Porous Materials: Catalytic Applications. Chem. Soc. Rev. 2013, 42 (9), 3876–3893. 15. Wagner, T.; Haffer, S.; Weinberger, C.; Klaus, D.; Tiemann, M. Mesoporous Materials as Gas Sensors. Chem. Soc. Rev. 2013, 42 (9), 4036–4053. 16. Li, Z.; Barnes, J. C.; Bosoy, A.; Stoddart, J. F.; Zink, J. I. Mesoporous Silica Nanoparticles in Biomedical Applications. Chem. Soc. Rev. 2012, 41 (7), 2590–2605. 17. Lin, T.; Chen, I. W.; Liu, F.; Yang, C.; Bi, H.; Xu, F.; Huang, F. Nitrogen-Doped Mesoporous Carbon of Extraordinary Capacitance for Electrochemical Energy Storage. Science 2015, 350 (6267), 1508–1513. 18. Beck, J. S.; Vartuli, J. C.; Roth, W. J.; Leonowicz, M. E.; Kresge, C. T.; Schmitt, K. D.; Chu, C. T. W.; Olson, D. H.; Sheppard, E. W.; McCullen, S. B.; et al. A New Family of Mesoporous Molecular Sieves Prepared With Liquid Crystal Templates. J. Am. Chem. Soc. 1992, 114 (27), 10834–10843. 19. Tanev, P. T.; Pinnavaia, T. J. A Neutral Templating Route to Mesoporous Molecular Sieves. Science 1995, 267 (5199), 865–867. 20. Gérardin, C.; Reboul, J.; Bonne, M.; Lebeau, B. Ecodesign of Ordered Mesoporous Silica Materials. Chem. Soc. Rev. 2013, 42 (9), 4217. 21. Kruk, M.; Jaroniec, M.; Ko, C. H.; Ryoo, R. Characterization of the Porous Structure of SBA-15. Chem. Mater. 2000, 12 (7), 1961–1968. 22. Ryoo, R.; Ko, C. H.; Kruk, M.; Antochshuk, V.; Jaroniec, M. Block-Copolymer-Templated Ordered Mesoporous Silica: Array of Uniform Mesopores or Mesopore-Micropore Network? J. Phys. Chem. B 2000, 104 (48), 11465–11471. 23. Matos, J. R.; Kruk, M.; Mercuri, L. P.; Jaroniec, M.; Zhao, L.; Kamiyama, T.; Terasaki, O.; Pinnavaia, T. J.; Liu, Y. Ordered Mesoporous Silica With Large Cage-Like Pores: Structural Identification and Pore Connectivity Design by Controlling the Synthesis Temperature and Time. J. Am. Chem. Soc. 2003, 125 (3), 821–829. 24. Zhou, W.; Hunter, H. M. A.; Wright, P. A.; Ge, Q.; Thomas, J. M. Imaging the Pore Structure and Polytypic Intergrowths in Mesoporous Silica. J. Phys. Chem. B 1998, 102 (36), 6933–6936. 25. Zhao, D.; Huo, Q.; Feng, J.; Chmelka, B. F.; Stucky, G. D. Nonionic Triblock and Star Diblock Copolymer and Oligomeric Surfactant Syntheses of Highly Ordered, Hydrothermally Stable, Mesoporous Silica Structures. J. Am. Chem. Soc. 1998, 120 (24), 6024–6036. 26. Kim, J. M.; Stucky, G. D. Synthesis of Highly Ordered Mesoporous Silica Materials Using Sodium Silicate and Amphiphilic Block Copolymers. Chem. Commun. 2000, (13), 1159–1160. 27. Tanev, P. T.; Pinnavaia, T. J. Mesoporous Silica Molecular Sieves Prepared by Ionic and Neutral Surfactant Templating: A Comparison of Physical Properties. Chem. Mater. 1996, 8 (8), 2068–2079. 28. Celer, E. B.; Jaroniec, M. Temperature-Programmed Microwave-Assisted Synthesis of SBA-15 Ordered Mesoporous Silica. J. Am. Chem. Soc. 2006, 128 (44), 14408– 14414. 29. Attard, G. S.; Glyde, J. C.; Göltner, C. G. Liquid-Crystalline Phases as Templates for the Synthesis of Mesoporous Silica. Nature 1995, 378 (6555), 366–368. 30. Sayari, A.; Liu, P.; Kruk, M.; Jaroniec, M. Characterization of Large-Pore MCM-41 Molecular Sieves Obtained Via Hydrothermal Restructuring. Chem. Mater. 1997, 9 (11), 2499–2506. 31. Zhao, D.; Sun, J.; Li, Q.; Stucky, G. D. Morphological Control of Highly Ordered Mesoporous Silica SBA-15. Chem. Mater. 2000, 12 (2), 275–279. 32. Che, S.; Liu, Z.; Ohsuna, T.; Sakamoto, K.; Terasaki, O.; Tatsumi, T. Synthesis and Characterization of Chiral Mesoporous Silica. Nature 2004, 429 (6989), 281–284. 33. Wan, Y.; Zhao, D. On the Controllable Soft-Templating Approach to Mesoporous Silicates. Chem. Rev. 2007, 107 (7), 2821–2860. 34. Wang, H.; Jeong, H. Y.; Imura, M.; Wang, L.; Radhakrishnan, L.; Fujita, N.; Castle, T.; Terasaki, O.; Yamauchi, Y. Shape- and Size-Controlled Synthesis in Hard Templates: Sophisticated Chemical Reduction for Mesoporous Monocrystalline Platinum Nanoparticles. J. Am. Chem. Soc. 2011, 133 (37), 14526–14529. 35. Attard, G. S.; Göltner, C. G.; Corker, J. M.; Henke, S.; Templer, R. H. Liquid-Crystal Templates for Nanostructured Metals. Angew. Chem. Int. Ed. Engl. 1997, 36 (12), 1315–1317. 36. Attard, G. S. Mesoporous Platinum Films From Lyotropic Liquid Crystalline Phases. Science 1997, 278 (5339), 838–840. 37. Wang, H.; Wang, L.; Sato, T.; Sakamoto, Y.; Tominaka, S.; Miyasaka, K.; Miyamoto, N.; Nemoto, Y.; Terasaki, O.; Yamauchi, Y. Synthesis of Mesoporous Pt Films With Tunable Pore Sizes From Aqueous Surfactant Solutions. Chem. Mater. 2012, 24 (9), 1591–1598. 38. Li, C.; Iqbal, M.; Jiang, B.; Wang, Z.; Kim, J.; Nanjundan, A. K.; Whitten, A. E.; Wood, K.; Yamauchi, Y. Pore-Tuning to Boost the Electrocatalytic Activity of Polymeric MicelleTemplated Mesoporous Pd Nanoparticles. Chem. Sci. 2019, 10 (14), 4054–4061. 39. Antonelli, D. M.; Ying, J. Y. Synthesis of Hexagonally Packed Mesoporous TiO2 by a Modified Sol–Gel Method. Angew. Chem. Int. Ed. Engl. 1995, 34 (18), 2014–2017. 40. Vaudry, F.; Khodabandeh, S.; Davis, M. E. Synthesis of Pure Alumina Mesoporous Materials. Chem. Mater. 1996, 8 (7), 1451–1464. 41. Yang, P.; Zhao, D.; Margolese, D. I.; Chmelka, B. F.; Stucky, G. D. Generalized Syntheses of Large-Pore Mesoporous Metal Oxides With Semicrystalline Frameworks. Nature 1998, 396 (6707), 152–155. 42. Yang, P.; Deng, T.; Zhao, D.; Feng, P.; Pine, D.; Chmelka, B. F.; Whitesides, G. M.; Stucky, G. D. Hierarchically Ordered Oxides. Science 1998, 282 (5397), 2244–2246. 43. Jiao, F.; Harrison, A.; Jumas, J. C.; Chadwick, A. V.; Kockelmann, W.; Bruce, P. G. Ordered Mesoporous Fe2O3 With Crystalline Walls. J. Am. Chem. Soc. 2006, 128 (16), 5468–5474. 44. Ren, Y.; Ma, Z.; Bruce, P. G. Ordered Mesoporous Metal Oxides: Synthesis and Applications. Chem. Soc. Rev. 2012, 41 (14), 4909–4927. 45. Lee, J.; Christopher Orilall, M.; Warren, S. C.; Kamperman, M.; Disalvo, F. J.; Wiesner, U. Direct Access to Thermally Stable and Highly Crystalline Mesoporous TransitionMetal Oxides With Uniform Pores. Nat. Mater. 2008, 7 (3), 222–228. 46. Marszewski, M.; Marszewska, J.; Pylypenko, S.; Jaroniec, M. Synthesis of Porous Crystalline Doped Titania Photocatalysts Using Modified Precursor Strategy. Chem. Mater. 2016, 28 (21), 7878–7888. 47. Kondo, J. N.; Domen, K. Crystallization of Mesoporous Metal Oxides. Chem. Mater. 2008, 20 (3), 835–847. 48. Noda, Y.; Lee, B.; Domen, K.; Kondo, J. N. Synthesis of Crystallized Mesoporous Tantalum Oxide and Its Photocatalytic Activity for Overall Water Splitting under Ultraviolet Light Irradiation. Chem. Mater. 2008, 20 (16), 5361–5367.

64

Mesostructured materials

49. Jiang, X.; Suzuki, N.; Bastakoti, B. P.; Wu, K. C. W.; Yamauchi, Y. Synthesis of Continuous Mesoporous Alumina Films With Large-Sized Cage-Type Mesopores by Using Diblock Copolymers. Chem. Asian J. 2012, 7 (7), 1713–1718. 50. Szeifert, J. M.; Fattakhova-Rohlfing, D.; Georgiadou, D.; Kalousek, V.; Rathouský, J.; Kuang, D.; Wenger, S.; Zakeeraddin, S. M.; Grätzel, M.; Bein, T. Brick and Mortar Strategy for the Formation of Highly Crystalline Mesoporous Titania Films From Nanocrystalline Building Blocks. Chem. Mater. 2009, 21 (7), 1260–1265. 51. Fattakhova-Rohlfing, D.; Szeifert, J. M.; Yu, Q.; Kalousek, V.; Rathousky, J.; Bein, T. Low-Temperature Synthesis of Mesoporous Titania -Silica Films With Pre-Formed Anatase Nanocrystals. Chem. Mater. 2009, 21 (12), 2410–2417. 52. Szeifert, J. M.; Fattakhova-Rohlfing, D.; Rathouský, J.; Bein, T. Multilayered High Surface Area “Brick and Mortar” Mesoporous Titania Films as Efficient Anodes in DyeSensitized Solar Cells. Chem. Mater. 2012, 24 (4), 659–663. 53. Ba, J.; Polleux, J.; Antonietti, M.; Niederberger, M. Non-Aqueous Synthesis of Tin Oxide Nanocrystals and Their Assembly Into Ordered Porous Mesostructures. Adv. Mater. 2005, 17 (20), 2509–2512. 54. Deshpande, A. S.; Pinna, N.; Smarsly, B.; Antonietti, M.; Niederberger, M. Controlled Assembly of Preformed Ceria Nanocrystals Into Highly Ordered 30 Nanostructures. Small 2005, 1 (3), 313–316. 55. Chaikittisilp, W.; Kim, H. J.; Jones, C. W. Mesoporous Alumina-Supported Amines as Potential Steam-Stable Adsorbents for Capturing CO2 from Simulated Flue Gas and Ambient Air. Energy Fuel 2011, 25 (11), 5528–5537. 56. Gonçalves, A. A. S.; Costa, M. J. F.; Zhang, L.; Ciesielczyk, F.; Jaroniec, M. One-Pot Synthesis of MeAl2O4 (Me ¼ Ni, Co, or Cu) Supported on g-Al2O3 With Ultralarge Mesopores: Enhancing Interfacial Defects in g-Al2O3 to Facilitate the Formation of Spinel Structures at Lower Temperatures. Chem. Mater. 2018, 30 (2), 436–446. 57. Van Der Voort, P.; Leus, K.; De Canck, E. Introduction to Porous Materials, John Wiley & Sons Ltd.: Hoboken, NJ, 2019. 58. Antochshuk, V.; Jaroniec, M. Functionalized Mesoporous Materials Obtained Via Interfacial Reactions in Self-Assembled Silica-Surfactant Systems. Chem. Mater. 2000, 12 (8), 2496–2501. 59. Hatton, B.; Landskron, K.; Whitnall, W.; Perovic, D.; Ozin, G. A. Past, Present, and Future of Periodic Mesoporous OrganosilicasdThe PMOs. Acc. Chem. Res. 2005, 38 (4), 305–312. 60. Inagaki, S.; Guan, S.; Fukushima, Y.; Ohsuna, T.; Terasaki, O. Novel Mesoporous Materials With a Uniform Distribution of Organic Groups and Inorganic Oxide in Their Frameworks. J. Am. Chem. Soc. 1999, 121 (41), 9611–9614. 61. Melde, B. J.; Holland, B. T.; Blanford, C. F.; Stein, A. Mesoporous Sieves With Unified Hybrid Inorganic/Organic Frameworks. Chem. Mater. 1999, 11 (11), 3302–3308. 62. Asefa, T.; MacLachlan, M. J.; Coombs, N.; Ozin, G. A. Periodic Mesoporous Organosilicas With Organic Groups Inside the Channel Walls. Nature 1999, 402 (6764), 867–871. 63. Hoffmann, F.; Fröba, M. Vitalising Porous Inorganic Silica Networks With Organic FunctionsdPMOs and Related Hybrid Materials. Chem. Soc. Rev. 2011, 40 (2), 608–620. 64. Sayari, A.; Hamoudi, S. Periodic Mesoporous Silica-Based Organic-Inorganic Nanocomposite Materials. Chem. Mater. 2001, 13 (10), 3151–3168. 65. Matos, J. R.; Kruk, M.; Mercuri, L. P.; Jaroniec, M.; Asefa, T.; Coombs, N.; Ozin, G. A.; Kamiyama, T.; Terasaki, O. Periodic Mesoporous Organosilica With Large Cagelike Pores. Chem. Mater. 2002, 14 (5), 1903–1905. 66. Olkhovyk, O.; Jaroniec, M. Periodic Mesoporous Organosilica With Large Heterocyclic Bridging Groups. J. Am. Chem. Soc. 2005, 127 (1), 60–61. 67. Park, S. S.; Santha Moorthy, M.; Ha, C.-S. Periodic Mesoporous Organosilicas for Advanced Applications. NPG Asia Mater. 2014, 6 (4), 96. 68. Gunathilake, C.; Górka, J.; Dai, S.; Jaroniec, M. Amidoxime-Modified Mesoporous Silica for Uranium Adsorption Under Seawater Conditions. J. Mater. Chem. A 2015, 3 (21), 11650–11659. 69. Bourda, L.; Jena, H. S.; Van Deun, R.; Kaczmarek, A. M.; Van Der Voort, P. Functionalized Periodic Mesoporous Organosilicas: From Metal Free Catalysis to Sensing. J. Mater. Chem. A 2019, 7 (23), 14060–14069. 70. Inagaki, S.; Guan, S.; Ohsuna, T.; Terasaki, O. An Ordered Mesoporous Organosilica Hybrid Material With a Crystal-Like Wall Structure. Nature 2002, 416 (6878), 304–307. 71. Li, H.; Eddaoudi, M.; Groy, T. L.; Yaghi, O. M. Establishing Microporosity in Open Metal-Organic Frameworks: Gas Sorption Isotherms for Zn(BDC) (BDC ¼ 1,4Benzenedicarboxylate) [28]. J. Am. Chem. Soc. 1998, 120 (33), 8571–8572. 72. Li, H.; Davis, C. E.; Groy, T. L.; Kelley, D. G.; Yaghi, O. M. Coordinatively Unsaturated Metal Centers in the Extended Porous Framework of Zn3(BDC)3$6Ch3OH (BDC ¼ 1,4Benzenedicarboxylate) [8]. J. Am. Chem. Soc. 1998, 120 (9), 2186–2187. 73. Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O’Keeffe, M.; Yaghi, O. M. Systematic Design of Pore Size and Functionality in Isoreticular MOFs and Their Application in Methane Storage. Science 2002, 295 (5554), 469–472. 74. Furukawa, H.; Ko, N.; Go, Y. B.; Aratani, N.; Choi, S. B.; Choi, E.; Yazaydin, A.Ö.; Snurr, R. Q.; O’Keeffe, M.; Kim, J.; et al. Ultrahigh Porosity in Metal-Organic Frameworks. Science 2010, 329 (5990), 424–428. 75. Qiu, L. G.; Xu, T.; Li, Z. Q.; Wang, W.; Wu, Y.; Jiang, X.; Tian, X. Y.; Zhang, L. D. Hierarchically Micro- and Mesoporous Metal-Organic Frameworks With Tunable Porosity. Angew. Chem. Int. Ed. 2008, 47 (49), 9487–9491. 76. Zhao, Y.; Zhang, J.; Han, B.; Song, J.; Li, J.; Wang, Q. Metal-Organic Framework Nanospheres With Well-Ordered Mesopores Synthesized in an Ionic Liquid/CO2/Surfactant System. Angew. Chem. Int. Ed. 2011, 50 (3), 636–639. 77. Song, L.; Zhang, J.; Sun, L.; Xu, F.; Li, F.; Zhang, H.; Si, X.; Jiao, C.; Li, Z.; Liu, S.; et al. Mesoporous Metal-Organic Frameworks: Design and Applications. Energ. Environ. Sci. 2012, 5 (6), 7508–7520. 78. Xuan, W.; Zhu, C.; Liu, Y.; Cui, Y. Mesoporous Metal–Organic Framework Materials. Chem. Soc. Rev. 2012, 41 (5), 1677–1695. 79. Liu, D.; Zou, D.; Zhu, H.; Zhang, J. Mesoporous Metal-Organic Frameworks: Synthetic Strategies and Emerging Applications. Small 2018, 14 (37), 1801454. 80. Ma, T. Y.; Liu, L.; Yuan, Z. Y. Direct Synthesis of Ordered Mesoporous Carbons. Chem. Soc. Rev. 2013, 42 (9), 3977–4003. 81. Ryoo, R.; Joo, S. H.; Jun, S. Synthesis of Highly Ordered Carbon Molecular Sieves Via Template-Mediated Structural Transformation. J. Phys. Chem. B 1999, 103 (37), 7743–7746. 82. Jun, S.; Joo, S. H.; Ryoo, R.; Kruk, M.; Jaroniec, M.; Liu, Z.; Ohsuna, T.; Terasaki, O. Synthesis of New, Nanoporous Carbon With Hexagonally Ordered Mesostructure [5]. J. Am. Chem. Soc. 2000, 122 (43), 10712–10713. 83. Kruk, M.; Jaroniec, M.; Kim, T. W.; Ryoo, R. Synthesis and Characterization of Hexagonally Ordered Carbon Nanopipes. Chem. Mater. 2003, 15 (14), 2815–2823. 84. Meng, Y.; Gu, D.; Zhang, F.; Shi, Y.; Yang, H.; Li, Z.; Yu, C.; Tu, B.; Zhao, D. Ordered Mesoporous Polymers and Homologous Carbon Frameworks: Amphiphilic Surfactant Templating and Direct Transformation. Angew. Chem. Int. Ed. 2005, 44 (43), 7053–7059. 85. Liang, C.; Dai, S. Synthesis of Mesoporous Carbon Materials Via Enhanced Hydrogen-Bonding Interaction. J. Am. Chem. Soc. 2006, 128 (16), 5316--5317. 86. Meng, Y.; Gu, D.; Zhang, F.; Shi, Y.; Cheng, L.; Feng, D.; Wu, Z.; Chen, Z.; Wan, Y.; Stein, A.; et al. A Family of Highly Ordered Mesoporous Polymer Resin and Carbon Structures from Organic-Organic Self-Assembly. Chem. Mater. 2006, 18 (18), 4447–4464. 87. Ryoo, R.; Joo, S. H.; Kruk, M.; Jaroniec, M. Ordered Mesoporous Carbons. Adv. Mater. 2001, 13 (9), 677–681. 88. Joo, S. H.; Choi, S. J.; Oh, I.; Kwak, J. Ordered Nanoporous Arrays of Carbon.Pdf. Nature 2001, 412 (July), 169–172. 89. Lee, J. S.; Joo, S. H.; Ryoo, R. Synthesis of Mesoporous Silicas of Controlled Pore Wall Thickness and Their Replication to Ordered Nanoporous Carbons With Various Pore Diameters. J. Am. Chem. Soc. 2002, 124 (7), 1156–1157. 90. Zhong, M.; Kim, E. K.; McGann, J. P.; Chun, S. E.; Whitacre, J. F.; Jaroniec, M.; Matyjaszewski, K.; Kowalewski, T. Electrochemically Active Nitrogen-Enriched Nanocarbons With Well-Defined Morphology Synthesized by Pyrolysis of Self-Assembled Block Copolymer. J. Am. Chem. Soc. 2012, 134 (36), 14846–14857. 91. Liu, J.; Wickramaratne, N. P.; Qiao, S. Z.; Jaroniec, M. Molecular-Based Design and Emerging Applications of Nanoporous Carbon Spheres. Nat. Mater. 2015, 14 (8), 763–774. 92. Zhang, P.; Wang, L.; Yang, S.; Schott, J. A.; Liu, X.; Mahurin, S. M.; Huang, C.; Zhang, Y.; Fulvio, P. F.; Chisholm, M. F.; et al. Solid-State Synthesis of Ordered Mesoporous Carbon Catalysts Via a Mechanochemical Assembly Through Coordination Cross-Linking. Nat. Commun. 2017, 8 (1), 15020.

Mesostructured materials

65

93. Kim, T. W.; Park, I. S.; Ryoo, R. A Synthetic Route to Ordered Mesoporous Carbon Materials With Graphitic Pore Walls. Angew. Chem. Int. Ed. 2003, 42 (36), 4375–4379. 94. Xia, Y.; Mokaya, R. Synthesis of Ordered Mesoporous Carbon and Nitrogen-Doped Carbon Materials With Graphitic Pore Walls Via a Simple Chemical Vapor Deposition Method. Adv. Mater. 2004, 16 (17), 1553–1558. 95. Su, F.; Zeng, J.; Bao, X.; Yu, Y.; Lee, J. Y.; Zhao, X. S. Preparation and Characterization of Highly Ordered Graphitic Mesoporous Carbon as a Pt Catalyst Support for Direct Methanol Fuel Cells. Chem. Mater. 2005, 17 (15), 3960–3967. 96. Yoon, S. B.; Chai, G. S.; Kang, S. K.; Yu, J. S.; Gierszal, K. P.; Jaroniec, M. Graphitized Pitch-Based Carbons With Ordered Nanopores Synthesized by Using Colloidal Crystals as Templates. J. Am. Chem. Soc. 2005, 127 (12), 4188–4189. 97. Madsen, C.; Jacobsen, C. J. H. Nanosized Zeolite CrystalsdConvenient Control of Crystal Size Distribution by Confined Space Synthesis. Chem. Commun. 1999, (8), 673–674. 98. Schmidt, I.; Madsen, C.; Jacobsen, C. J. H. Confined Space Synthesis. A Novel Route to Nanosized Zeolites. Inorg. Chem. 2000, 39 (11), 2279–2283. 99. Jacobsen, C. J. H.; Madsen, C.; Houzvicka, J.; Schmidt, I.; Carlsson, A. Mesoporous Zeolite Single Crystals. J. Am. Chem. Soc. 2000, 122 (29), 7116–7117. 100. Tao, Y.; Kanoh, H.; Kaneko, K. ZSM-5 Monolith of Uniform Mesoporous Channels. J. Am. Chem. Soc. 2003, 125 (20), 6044–6045. 101. Tao, Y.; Kanoh, H.; Kaneko, K. Uniform Mesopore-Donated Zeolite Y Using Carbon Aerogel Templating. J. Phys. Chem. B 2003, 107 (40), 10974–10976. 102. Tao, Y.; Kanoh, H.; Kaneko, K. Synthesis of Mesoporous Zeolite a by Resorcinol-Formaldehyde Aerogel Templating. Langmuir 2005, 21 (2), 504–507. 103. Kustova, M.; Egeblad, K.; Zhu, K.; Christensen, C. H. Versatile Route to Zeolite Single Crystals With Controlled Mesoporosity: In Situ Sugar Decomposition for Templating of Hierarchical Zeolites. Chem. Mater. 2007, 19 (12), 2915–2917. 104. Schmidt, I.; Boisen, A.; Gustavsson, E.; Ståhl, K.; Pehrson, S.; Dahl, S.; Carlsson, A.; Jacobsen, C. J. H. Carbon Nanotube Templated Growth of Mesoporous Zeolite Single Crystals. Chem. Mater. 2001, 13 (12), 4416–4418. 105. Kim, S. S.; Shah, J.; Pinnavaia, T. J. Colloid-Imprinted Carbons as Templates for the Nanocasting Synthesis of Mesoporous ZSM-5 Zeolite. Chem. Mater. 2003, 15 (8), 1664–1668. 106. Fan, W.; Snyder, M. A.; Kumar, S.; Lee, P. S.; Yoo, W. C.; McCormick, A. V.; Lee Penn, R.; Stein, A.; Tsapatsis, M. Hierarchical Nanofabrication of Microporous Crystals With Ordered Mesoporosity. Nat. Mater. 2008, 7 (12), 984–991. 107. Chen, H.; Wydra, J.; Zhang, X.; Lee, P. S.; Wang, Z.; Fan, W.; Tsapatsis, M. Hydrothermal Synthesis of Zeolites With Three-Dimensionally Ordered Mesoporous-Imprinted Structure. J. Am. Chem. Soc. 2011, 133 (32), 12390–12393. 108. Xiao, F.-S.; Wang, L.; Yin, C.; Lin, K.; Di, Y.; Li, J.; Xu, R.; Su, D. S.; Schlögl, R.; Yokoi, T.; et al. Catalytic Properties of Hierarchical Mesoporous Zeolites Templated With a Mixture of Small Organic Ammonium Salts and Mesoscale Cationic Polymers. Angew. Chem. Int. Ed. 2006, 45 (19), 3090–3093. 109. Zhu, J.; Zhu, Y.; Zhu, L.; Rigutto, M.; Van Der Made, A.; Yang, C.; Pan, S.; Wang, L.; Zhu, L.; Jin, Y.; et al. Highly Mesoporous Single-Crystalline Zeolite Beta Synthesized Using a Nonsurfactant Cationic Polymer as a Dual-Function Template. J. Am. Chem. Soc. 2014, 136 (6), 2503–2510. 110. Liu, F.; Willhammar, T.; Wang, L.; Zhu, L.; Sun, Q.; Meng, X.; Carrillo-Cabrera, W.; Zou, X.; Xiao, F. S. ZSM-5 Zeolite Single Crystals With b-Axis-Aligned Mesoporous Channels as an Efficient Catalyst for Conversion of Bulky Organic Molecules. J. Am. Chem. Soc. 2012, 134 (10), 4557–4560. 111. Choi, M.; Na, K.; Kim, J.; Sakamoto, Y.; Terasaki, O.; Ryoo, R. Stable Single-Unit-Cell Nanosheets of Zeolite MFI as Active and Long-Lived Catalysts. Nature 2009, 461 (7261), 246–249. 112. Choi, M.; Cho, H. S.; Srivastava, R.; Venkatesan, C.; Choi, D. H.; Ryoo, R. Amphiphilic Organosilane-Directed Synthesis of Crystalline Zeolite With Tunable Mesoporosity. Nat. Mater. 2006, 5 (9), 718–723. 113. Wang, H.; Pinnavaia, T. J. MFI Zeolite With Small and Uniform Intracrystal Mesopores. Angew. Chem. Int. Ed. 2006, 45 (45), 7603–7606. 114. Möller, K.; Bein, T. MesoporositydA New Dimension for Zeolites. Chem. Soc. Rev. 2013, 42 (9), 3689–3707. 115. Möller, K.; Yilmaz, B.; Jacubinas, R. M.; Müller, U.; Bein, T. One-Step Synthesis of Hierarchical Zeolite Beta Via Network Formation of Uniform Nanocrystals. J. Am. Chem. Soc. 2011, 133 (14), 5284–5295. 116. Chen, L. H.; Li, X. Y.; Tian, G.; Li, Y.; Rooke, J. C.; Zhu, G. S.; Qiu, S. L.; Yang, X. Y.; Su, B. L. Highly Stable and Reusable Multimodal Zeolite TS-1 Based Catalysts With Hierarchically Interconnected Three-Level Micro-Meso-Macroporous Structure. Angew. Chem. Int. Ed. 2011, 50 (47), 11156–11161. 117. Mcller, K.; Yilmaz, B.; Müller, U.; Bein, T. Nanofusion: Mesoporous Zeolites Made Easy. Chem. Eur. J. 2012, 18 (25), 7671–7674. 118. Zhang, X.; Liu, D.; Xu, D.; Asahina, S.; Cychosz, K. A.; Agrawal, K. V.; Al Wahedi, Y.; Bhan, A.; Al Hashimi, S.; Terasaki, O.; et al. Synthesis of Self-Pillared Zeolite Nanosheets by Repetitive Branching. Science 2012, 336 (6089), 1684–1687. 119. Zhang, H.; Wang, L.; Zhang, D.; Meng, X.; Xiao, F. S. Mesoporous and Al-Rich MFI Crystals Assembled With Aligned Nanorods in the Absence of Organic Templates. Microporous Mesoporous Mater. 2016, 233, 133–139. 120. Agostini, G.; Lamberti, C.; Palin, L.; Milanesio, M.; Danilina, N.; Xu, B.; Janousch, M.; Van Bokhoven, J. A. In Situ XAS and XRPD Parametric Rietveld Refinement to Understand Dealumination of Y Zeolite Catalyst. J. Am. Chem. Soc. 2010, 132 (2), 667–678. 121. Ogura, M.; Shinomiya, S. Y.; Tateno, J.; Nara, Y.; Kikuchi, E.; Matsukata, M. Formation of Uniform Mesopores in ZSM-5 Zeolite Through Treatment in Alkaline Solution. Chem. Lett. 2000, (8), 882–883. 122. Groen, J. C.; Jansen, J. C.; Moulijn, J. A.; Pérez-Ramírez, J. Optimal Aluminum-Assisted Mesoporosity Development in MFI Zeolites by Desilication. J. Phys. Chem. B 2004, 108 (35), 13062–13065. 123. Groen, J. C.; Moulijn, J. A.; Pérez-Ramírez, J. Desilication: On the Controlled Generation of Mesoporosity in MFI Zeolites. J. Mater. Chem. 2006, 16 (22), 2121–2131. 124. Peréz-Ramírez, J.; Verboekend, D.; Bonilla, A.; Abelló, S. Zeolite Catalysts With Tunable Hierarchy Factor by Pore-Growth Moderators. Adv. Funct. Mater. 2009, 19 (24), 3972–3979. 125. Yoo, W. C.; Zhang, X.; Tsapatsis, M.; Stein, A. Synthesis of Mesoporous ZSM-5 Zeolites Through Desilication and Re-Assembly Processes. Microporous Mesoporous Mater. 2012, 149 (1), 147–157. 126. García-Martínez, J.; Johnson, M.; Valla, J.; Li, K.; Ying, J. Y. Mesostructured Zeolite YdHigh Hydrothermal Stability and Superior FCC Catalytic Performance. Cat. Sci. Technol. 2012, 2 (5), 987–994. 127. Wang, Y.; Arandiyan, H.; Scott, J.; Bagheri, A.; Dai, H.; Amal, R. Recent Advances in Ordered Meso/Macroporous Metal Oxides for Heterogeneous Catalysis: A Review. J. Mater. Chem. A 2017, 5 (19), 8825–8846. 128. Shen, W.; Dong, X.; Zhu, Y.; Chen, H.; Shi, J. Mesoporous CeO2 and CuO-Loaded Mesoporous CeO2: Synthesis, Characterization, and CO Catalytic Oxidation Property. Microporous Mesoporous Mater. 2005, 85 (1–2), 157–162. 129. Wang, C. A.; Li, S.; An, L. Hierarchically Porous Co3O4 Hollow Spheres With Tunable Pore Structure and Enhanced Catalytic Activity. Chem. Commun. 2013, 49 (67), 7427–7429. 130. An, K.; Alayoglu, S.; Musselwhite, N.; Plamthottam, S.; Melaet, G.; Lindeman, A. E.; Somorjai, G. A. Enhanced CO Oxidation Rates at the Interface of Mesoporous Oxides and Pt Nanoparticles. J. Am. Chem. Soc. 2013, 135 (44), 16689–16696. 131. Rao, Y.; Trudeau, M.; Antonelli, D. Sulfated and Phosphated Mesoporous Nb Oxide in the Benzylation of Anisole and Toluene by Benzyl Alcohol. J. Am. Chem. Soc. 2006, 128 (43), 13996–13997. 132. Kang, J.; Rao, Y.; Trudeau, M.; Antonelli, D. Sulfated Mesoporous Tantalum Oxides in the Shape Selective Synthesis of Linear Alkyl Benzene. Angew. Chem. Int. Ed. 2008, 47 (26), 4896–4899. 133. Rao, Y.; Kang, J.; Antonelli, D. 1-Hexene Isomerization over Sulfated Mesoporous Ta Oxide: The Effects of Active Site and Confinement. J. Am. Chem. Soc. 2008, 130 (2), 394–395.

66

Mesostructured materials

134. Huang, Y.; Xu, S.; Lin, V. S. Y. Bifunctionalized Mesoporous Materials With Site-Separated Brønsted Acids and Bases: Catalyst for a Two-Step Reaction Sequence. Angew. Chem. Int. Ed. 2011, 50 (3), 661–664. 135. Melero, J. A.; Iglesias, J.; Arsuaga, J. M.; Sainz-Pardo, J.; De Frutos, P.; Blazquez, S. Synthesis and Catalytic Activity of Organic-Inorganic Hybrid Ti-SBA-15 Materials. J. Mater. Chem. 2007, 17 (4), 377–385. 136. De Canck, E.; Nahra, F.; Bevernaege, K.; Vanden Broeck, S.; Ouwehand, J.; Maes, D.; Nolan, S. P.; Van Der Voort, P. PMO-Immobilized AuI–NHC Complexes: Heterogeneous Catalysts for Sustainable Processes. ChemPhysChem 2018, 19 (4), 430–436. 137. Liu, L.; Deng, Q. F.; Agula, B.; Zhao, X.; Ren, T. Z.; Yuan, Z. Y. Ordered Mesoporous Carbon Catalyst for Dehydrogenation of Propane to Propylene. Chem. Commun. 2011, 47 (29), 8334–8336. 138. Wang, Y.; Zhang, J.; Wang, X.; Antonietti, M.; Li, H. Boron- and Fluorine-Containing Mesoporous Carbon Nitride Polymers: Metal-Free Catalysts for Cyclohexane Oxidation. Angew. Chem. Int. Ed. 2010, 49 (19), 3356–3359. 139. Liu, R.; Wang, X.; Zhao, X.; Feng, P. Sulfonated Ordered Mesoporous Carbon for Catalytic Preparation of Biodiesel. Carbon 2008, 46 (13), 1664–1669. 140. Sun, Z.; Sun, B.; Qiao, M.; Wei, J.; Yue, Q.; Wang, C.; Deng, Y.; Kaliaguine, S.; Zhao, D. A General Chelate-Assisted Co-Assembly to Metallic Nanoparticles- Incorporated Ordered Mesoporous Carbon Catalysts for Fischer-Tropsch Synthesis. J. Am. Chem. Soc. 2012, 134 (42), 17653–17660. 141. Wang, S.; Zhao, Q.; Wei, H.; Wang, J. Q.; Cho, M.; Cho, H. S.; Terasaki, O.; Wan, Y. Aggregation-Free Gold Nanoparticles in Ordered Mesoporous Carbons: Toward Highly Active and Stable Heterogeneous Catalysts. J. Am. Chem. Soc. 2013, 135 (32), 11849–11860. 142. Wang, G. H.; Cao, Z.; Gu, D.; Pfänder, N.; Swertz, A. C.; Spliethoff, B.; Bongard, H. J.; Weidenthaler, C.; Schmidt, W.; Rinaldi, R.; et al. Nitrogen-Doped Ordered Mesoporous Carbon Supported Bimetallic PtCo Nanoparticles for Upgrading of Biophenolics. Angew. Chem. Int. Ed. 2016, 55 (31), 8850–8855. 143. Li, Z.; Peters, A. W.; Bernales, V.; Ortuño, M. A.; Schweitzer, N. M.; Destefano, M. R.; Gallington, L. C.; Platero-Prats, A. E.; Chapman, K. W.; Cramer, C. J.; et al. MetalOrganic Framework Supported Cobalt Catalysts for the Oxidative Dehydrogenation of Propane at Low Temperature. ACS Cent. Sci. 2017, 3 (1), 31–38. 144. Park, Y. K.; Sang, B. C.; Kim, H.; Kim, K.; Won, B. H.; Choi, K.; Choi, J. S.; Ahn, W. S.; Won, N.; Kim, S.; et al. Crystal Structure and Guest Uptake of a Mesoporous MetalOrganic Framework Containing Cages of 3.9 and 4.7 Nm in Diameter. Angew. Chem. Int. Ed. 2007, 46 (43), 8230–8233. 145. Li, P.; Moon, S. Y.; Guelta, M. A.; Harvey, S. P.; Hupp, J. T.; Farha, O. K. Encapsulation of a Nerve Agent Detoxifying Enzyme by a Mesoporous Zirconium Metal-Organic Framework Engenders Thermal and Long-Term Stability. J. Am. Chem. Soc. 2016, 138 (26), 8052–8055. 146. Zhao, M.; Yuan, K.; Wang, Y.; Li, G.; Guo, J.; Gu, L.; Hu, W.; Zhao, H.; Tang, Z. Metal-Organic Frameworks as Selectivity Regulators for Hydrogenation Reactions. Nature 2016, 539 (7627), 76–80. 147. Wang, C.; Tuninetti, J.; Wang, Z.; Zhang, C.; Ciganda, R.; Salmon, L.; Moya, S.; Ruiz, J.; Astruc, D. Hydrolysis of Ammonia-Borane over Ni/ZIF-8 Nanocatalyst: High Efficiency, Mechanism, and Controlled Hydrogen Release. J. Am. Chem. Soc. 2017, 139 (33), 11610–11615. 148. Yang, Q.; Liu, W.; Wang, B.; Zhang, W.; Zeng, X.; Zhang, C.; Qin, Y.; Sun, X.; Wu, T.; Liu, J.; et al. Regulating the Spatial Distribution of Metal Nanoparticles Within MetalOrganic Frameworks to Enhance Catalytic Efficiency. Nat. Commun. 2017, 8 (1), 14429. 149. Hartmann, M.; Machoke, A. G.; Schwieger, W. Catalytic Test Reactions for the Evaluation of Hierarchical Zeolites. Chem. Soc. Rev. 2016, 45 (12), 3313–3330. 150. Zhu, X.; Hofmann, J. P.; Mezari, B.; Kosinov, N.; Wu, L.; Qian, Q.; Weckhuysen, B. M.; Asahina, S.; Ruiz-Martínez, J.; Hensen, E. J. M. Trimodal Porous Hierarchical SSZ-13 Zeolite With Improved Catalytic Performance in the Methanol-to-Olefins Reaction. ACS Catal. 2016, 6 (4), 2163–2177. 151. Khodakov, A. Y.; Chu, W.; Fongarland, P. Advances in the Development of Novel Cobalt Fischer-Tropsch Catalysts for Synthesis of Long-Chain Hydrocarbons and Clean Fuels. Chem. Rev. 2007, 107 (5), 1692–1744. 152. Kang, J.; Cheng, K.; Zhang, L.; Zhang, Q.; Ding, J.; Hua, W.; Lou, Y.; Zhai, Q.; Wang, Y. Mesoporous Zeolite-Supported Ruthenium Nanoparticles as Highly Selective FischerTropsch Catalysts for the Production of C5-C11 Isoparaffins. Angew. Chem. Int. Ed. 2011, 50 (22), 5200–5203. 153. Cheng, K.; Kang, J.; Huang, S.; You, Z.; Zhang, Q.; Ding, J.; Hua, W.; Lou, Y.; Deng, W.; Wang, Y. Mesoporous Beta Zeolite-Supported Ruthenium Nanoparticles for Selective Conversion of Synthesis Gas to C 5-C 11 Isoparaffins. ACS Catal. 2012, 2 (3), 441–449. 154. Kim, J. C.; Lee, S.; Cho, K.; Na, K.; Lee, C.; Ryoo, R. Mesoporous MFI Zeolite Nanosponge Supporting Cobalt Nanoparticles as a Fischer-Tropsch Catalyst With High Yield of Branched Hydrocarbons in the Gasoline Range. ACS Catal. 2014, 4 (11), 3919–3927. 155. Lin, Q.; Yang, G.; Chen, Q.; Fan, R.; Yoneyama, Y.; Wan, H.; Tsubaki, N. Design of a Hierarchical Meso/Macroporous Zeolite-Supported Cobalt Catalyst for the Enhanced Direct Synthesis of Isoparaffins From Syngas. ChemCatChem 2015, 7 (4), 682–689. 156. Xing, C.; Yang, G.; Wu, M.; Yang, R.; Tan, L.; Zhu, P.; Wei, Q.; Li, J.; Mao, J.; Yoneyama, Y.; et al. Hierarchical Zeolite Y Supported Cobalt Bifunctional Catalyst for Facilely Tuning the Product Distribution of Fischer-Tropsch Synthesis. Fuel 2015, 148, 48–57. 157. Mazonde, B.; Cheng, S.; Zhang, G.; Javed, M.; Gao, W.; Zhang, Y.; Tao, M.; Lu, C.; Xing, C. A Solvent-Free: In Situ Synthesis of a Hierarchical Co-Based Zeolite Catalyst and Its Application to Tuning Fischer-Tropsch Product Selectivity. Cat. Sci. Technol. 2018, 8 (11), 2802–2808. 158. Sartipi, S.; Alberts, M.; Meijerink, M. J.; Keller, T. C.; Pérez-Ramírez, J.; Gascon, J.; Kapteijn, F. Towards Liquid Fuels From Biosyngas: Effect of Zeolite Structure in Hierarchical-Zeolite-Supported Cobalt Catalysts. ChemSusChem 2013, 6 (9), 1646–1650. 159. Kang, J.; Wang, X.; Peng, X.; Yang, Y.; Cheng, K.; Zhang, Q.; Wang, Y. Mesoporous Zeolite Y-Supported co Nanoparticles as Efficient Fischer-Tropsch Catalysts for Selective Synthesis of Diesel Fuel. Ind. Eng. Chem. Res. 2016, 55 (51), 13008–13019. 160. Musselwhite, N.; Na, K.; Sabyrov, K.; Alayoglu, S.; Somorjai, G. A. Mesoporous Aluminosilicate Catalysts for the Selective Isomerization of N-Hexane: The Roles of Surface Acidity and Platinum Metal. J. Am. Chem. Soc. 2015, 137 (32), 10231–10237. 161. Tang, T.; Yin, C.; Wang, L.; Ji, Y.; Xiao, F. S. Good Sulfur Tolerance of a Mesoporous Beta Zeolite-Supported Palladium Catalyst in the Deep Hydrogenation of Aromatics. J. Catal. 2008, 257 (1), 125–133. 162. Fu, W.; Zhang, L.; Wu, D.; Xiang, M.; Zhuo, Q.; Huang, K.; Tao, Z.; Tang, T. Mesoporous Zeolite-Supported Metal Sulfide Catalysts With High Activities in the Deep Hydrogenation of Phenanthrene. J. Catal. 2015, 330, 423–433. 163. Tang, T.; Yin, C.; Wang, L.; Ji, Y.; Xiao, F. S. Superior Performance in Deep Saturation of Bulky Aromatic Pyrene over Acidic Mesoporous Beta Zeolite-Supported Palladium Catalyst. J. Catal. 2007, 249 (1), 111–115. 164. Cui, T. L.; Ke, W. Y.; Zhang, W. B.; Wang, H. H.; Li, X. H.; Chen, J. S. Encapsulating Palladium Nanoparticles Inside Mesoporous MFI Zeolite Nanocrystals for Shape-Selective Catalysis. Angew. Chem. Int. Ed. 2016, 55 (32), 9178–9182.

6.04 Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts ski, Department of Chemistry and Applied Biosciences, ETH Zürich, Zürich, Christophe Cope´ret and Maciej Damian Korzyn Switzerland © 2023 Elsevier Ltd. All rights reserved.

6.04.1 6.04.2 6.04.3 6.04.4 6.04.5 6.04.6 6.04.7 6.04.8 References

Complexity of heterogeneous catalysis Single-site catalyst concept The core principles of the surface organometallic chemistry approach Surface of the supporting material as a ligand Tailored molecular precursors for well-defined surface species Characterization techniques utilized in the SOMC approach Selected success stories in SOMC single site catalysis Conclusion

67 68 68 69 74 76 78 81 81

Abstract Surface Organometallic Chemistry (SOMC), an approach to generate molecularly defined surface sites on solids, has opened possibilities to understand heterogeneous catalysis with atomic-level precision. Treating the surface of traditional supports as defined (albeit complex) ligands and applying the rules of molecular inorganic chemistry allowed for formation of countless examples of well-defined surface moieties, some of which were shown to display remarkable catalytic performances for variety of chemical transformations, surpassing their homogeneous analogs, and behaving as single-site catalysts. Additionally, the ability to precisely design the coordination environment of the metal center brought us closer to elucidation of the nature of active sites in more complex industrial catalysts. The aim of this article is to familiarize the reader with the logic behind and the techniques involved in the SOMC approach. We will begin by outlining the complexity of heterogeneous catalysis and how SOMC enters its realm, followed by description of the methodology and its defining features. Finally, we will discuss some of the success stories in using SOMC approach in single-site heterogeneous catalysis.

6.04.1

Complexity of heterogeneous catalysis

The science of catalysis is one of the key components of “The 12 Principles of Green Chemistry,” and thus a cornerstone of a sustainable and atom-efficient chemical industry.1 It allows us to realize chemical transformations, which otherwise would not be possible due to factors such as slow overall reaction rate or unfavorable selectivity for the desired product in a complex reactant space.2 This purely kinetic phenomenon involves steering the reaction through alternative step(s) with lower overall activation barrier(s) resulting in a faster and ultimately more selective transformation. The presence of an additional component of the reaction mixture, a catalyst, brings about those new pathways by interacting with the reactant(s).3 At the end of catalytic cycle the chemical constitution of the catalyst is restored to its initial state, resulting in no net change to catalyst amount over the course of the reaction (in the absence of deactivation). Catalysts and accordingly the respective catalytic processes are traditionally classified as either homogeneous or heterogeneous.4 In the case of the former, the catalyst and other reactants are dispersed in the same phase (typically all of the reaction mixture components are in the liquid phase), while in the latter the catalyst is in a different phase than the rest of the reaction mixture constituents (typically solid catalyst with the other reactants in the liquid or gas phase). The reason for this rather pragmatic classification becomes apparent when one factors in the goal of the catalytic process development, which is process intensification and its ultimate industrial implementation.5 At some stage of the process the catalyst must be separated from the rest of the reaction mixture so that the pure product can be obtained and the catalytically active species (often containing expensive and rare metals in its formulation) can be used for subsequent transformations. Heterogeneous catalysts have a huge advantage over their homogeneous counterparts in this regard as they can be readily removed from the reaction mixture (e.g., by filtration), minimizing the need for more energy demanding and costly separation process. In the ideal scenario the heterogeneous catalyst allows for a continuous operation (e.g., flow reactor) with the reactants being passed over the catalytically active stationary phase, rendering the catalyst recovery a non-existent issue.6 Thus, it is rather unsurprising that presently over 80% of all industrial catalytic processes use heterogeneous catalysts.7 With such a clear pre-eminence of heterogeneous catalysis in industry one would expect that by now we have reached a good understanding of the precise nature of the catalytically active sites and structure-activity relationships in solid-state catalysts. On the

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00003-0

67

68

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

contrary, the nature of catalytically active species is typically highly debated, even in the case of processes that routinely yield bulk chemicals on a million tonnes scale per annum.8 This seemingly paradoxical situation stems from the inherent complexity of solids resulting in challenges associated with their characterization. Moreover, the typical heterogeneous catalyst formulations consist of the active phase, which is dispersed onto a high-surface supporting material (e.g., silica, alumina, titania).9 This approach, termed supported catalysis, can boost the mass and thermal transport while providing dispersion of the actives species and preventing their sintering.10 At the same time the structural uncertainties related to the supporting material propagate onto the active centers and compound the challenges associated with understanding of their structure. This is most evident when amorphous supports are utilized.11 The broad distribution of local environments associated with amorphous nature results often in a small subset of sites which are suited for the desired transformation. The contribution of the remainder of the potentially active sites can span from being completely inactive (resulting in a noneconomical utilization of the often-expensive material) to showing a different selectivity than the one desired (ultimately compromising the overall catalyst selectivity). Additionally, the presence of various environments obfuscates the spectroscopic signatures belonging to sites relevant in catalysis. While leveraging the bulk ordering associated with crystalline catalyst supports (or catalysts themselves) could potentially ameliorate some of the issues described, the exposure of various crystalline facets, presence of defects and other factors can lead to more than one unique catalytically-relevant environment.12 Overall, all these factors (multitude of coordination environments, low amount of the active site of interest, overshadowing of the information from the surface by the bulk of the material) make characterization and understanding of the active sites in heterogeneous catalysts a daunting task.

6.04.2

Single-site catalyst concept

The issues described in the previous section are much less prevalent in homogeneous catalysis.13 The active species is usually a welldefined molecular entity, commonly a metal complex, consisting of the desired active metal center surrounded by moieties, ligands, which tailor the reactivity of the metal center in terms of electronics and sterics. One can optimize catalyst performance by introducing incremental changes to the complex (ligands) and observing their influence on the activity. This is only possible because of the molecular precision with which these complexes are assembled, the diversity of ligands and the availability of various characterization methods. Uncovering the intimate structure of the catalyst is facilitated by the single-crystal diffraction methods, which can provide a snapshot of the molecule of interest. Additionally, analysis of spectroscopic data is often much more straightforward in liquid phase due to the averaging effect of the rapid molecular motion.14 The unambiguous and uniform nature of the active sites in homogeneous catalysis led to usage of the term single-site catalyst. One of the more prominent exemplars of this concept are metallocene catalysts used in polymerization of olefins, which due to their bespoke coordination environment, can afford polyolefins with rigorous control over the polymer chain stereospecifity and narrow distribution of molecular weights.15,16 Such feats could not be achieved with their predecessors, the heterogeneous Ziegler-Natta systems. The concept of a single-site catalyst has also permeated into the realm of heterogeneous systems. These single-site heterogeneous catalysts are spatially separated ensembles of active sites (one or more atoms), each with the exact same chemical environment and interaction energy with substrates.17 This set of requirements ensures that every active center will react with the substrate in the same manner and that the individual sites cannot interact with each other. Accessing such sites has been an ongoing challenge for many years and a wide variety of synthetic methodologies have been explored in this context. Herein, we will discuss the Surface Organometallic Chemistry (SOMC) approachdthe union of the molecular inorganic and organometallic chemistry with surface science and materials chemistry.

6.04.3

The core principles of the surface organometallic chemistry approach

The naissance of the SOMC approach is a result of the desire to develop a molecular approach to generation of single-sites akin to those present in the homogeneous catalysis, but on the surface of solids.18–23 One crucial realization in the early days of SOMC research that made it possible was that, despite its complexity, the surface of the support can be viewed as a ligand, similar to tailored moieties used in the assembly of molecular complexes. Knowing the properties of such ligands and guided by the principles of synthetic molecular chemistry, one can pick a well-defined molecular precursor that undergoes predictable reaction with the surface moieties, leading to the desired grafted molecular architecture. The last component that completes the SOMC approach is the thorough characterization of the resulting surface species. The development of the surface organometallic chemistry was always closely tied to the advances in characterization of solid-state materials. The molecular parentage and the uniform character of the surface moieties permits comprehensive characterization using the combination of complementary spectroscopic, diffraction and microscopy-based techniques leading to an in-depth understanding of the heterogeneous catalyst. We will now discuss its defining features (Fig. 1) in more detail with the aim of highlighting the parallels and differences between SOMC and homogeneous chemistry.

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

Fig. 1

6.04.4

69

Defining features of SOMC approach.

Surface of the supporting material as a ligand

The key to the successful support functionalization in SOMC is the thorough understanding of the interfacial region. The hallmark of the SOMC approach is the effort placed in preparation of the support prior to functionalization in order to elicit ligand-like controlled behavior. Most of the research in the field has focused on oxidic supports, a testament to their widespread usage in industrial catalysis. Despite their deceptively simple MOx formulas, the surface structure of these materials is quite complicated. It is worth pointing out that while inherited from the bulk structure, the surfaces of solid particles expose specific sites not found in the bulk.24 In other words, one cannot expect to have similar connectivity and environments present on the surface as in the bulk of the material. In case of oxides this commonly manifests itself by the presence of low-coordinate metal sites and OH groups on the surface of the material. These surface hydroxyl moieties are the most common targets for functionalization using the SOMC approach as they can easily engage in the acid-base chemistry with the precursor used in the functionalization (Fig. 2). However, low coordinate metal sites are intrinsically Lewis acidic and can also interact with molecular precursors (vide infra). To perform such reactions in a controlled fashion with well-defined stoichiometry, it is necessary to know the amount and the distribution of the reactive OH moieties present on the surface. Additionally, it is prudent to remove any adventitious water present on the surface (due to extensive hydrogen bonding with the hydroxyl groups) that could engage in deleterious side reactions. Thus, in the case of almost any oxidic material, the first step in support preparation is thermal treatment under (high) vacuum or inert gas.25 The evolution of the material generally occurs in the following sequence of steps: initially, the physisorbed and hydrogenbonded water molecules are removed from the surface leaving behind only the hydroxyl moieties. As the temperature increases, some of the neighboring OH groups start undergoing condensation (releasing additional water molecules and generating bridging oxo species), decreasing the number of hydroxyls present on the surface. Eventually one obtains a material with spatially isolated OH groups, primed for anchoring of metal species. Due to their isolation on the surface and lack of mobility, surface OH groups engender the formation of monomeric and well-defined metal species, acting similarly to bulky ligands in the classical coordination chemistry. The species resulting from such monopodal coordination with surface are referred to as monografted complex, while the reaction with more OH groups per one precursor molecule would give rise to multigrafted (e.g., bisgrafted) species. Before we discuss some of the typical supports used in the SOMC methodology we will introduce a framework for describing properties of these materials.26,27 As one of the objectives of this article is to underline the molecular lineage of the SOMC approach we will refer to one of the classical concepts used in coordination chemistry. This system, commonly known as Covalent Bond Classification and introduced by late Malcolm Green, differentiates three types of functionalities denoted L, X and Z.28 X-type ligands are typically 1 electron donors in the covalent electron counting model with the typical examples including anionic species such as hydrides, halides, simple hydrocarbyls, alkoxides (in the absence of p donation), amides and others. The L-type ligands typically donate two electrons, often a lone pair, to the metal center. Here, the most prominent examples involve water, ethers, carbon monoxide, tertiary amines and phosphines etc. Finally, the Z-type ligand features an empty orbital, which can accept an electron pair. This behavior is mostly associated with Lewis acids. Some ligands used in modern coordination chemistry show much more complex properties and to fully describe their properties, LlXxZz notation can be used, where l, x and z subscripts denote the number of the respective functionalities. Using this convention, a benzene molecule coordinated in the h6 fashion is an L3 ligand, h3-allyl ligand can be described as LX-type species and the cyclopentadienyl group acts as a L2X fragment. It is also important to note that one ligand can behave in a diverse fashion, depending on metal center involved in the reaction. An example, which is highly relevant to heterogeneous catalysis, is a terminal oxo ligand. By the virtue of its dianionic character, this ligand can be

Fig. 2

Grafting of molecular precursor on surface hydroxyl groups.

70

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

classified as X2 species. However, if bound to an oxophilic metal center it can change its hybridization from sp2 to sp. This allows for an additional donation of one of the oxygen lone pairs to the metal center turning the oxo moiety into an LX2 ligand. Finally, using this analysis, one can identify potential analogies between ligands. One particularly relevant and not necessarily obvious resemblance is that between alkoxide (and siloxide) moieties and cyclopentadienyl anion. In theory, all of these can act as six electron donors with effective L2X character. Although this approach has to be used cautiously, it can provide some initial guidance in finding suitable targets for SOMC functionalization. Equipped with this theoretical framework, we can discuss some of the common supports used in the SOMC methodology. Silica (SiO2) can refer to a variety of polymeric materials, which are primarily assembled from corner sharing sp3-hybridized SiO4 units, that can be either amorphous or crystalline.29 Crystalline silicas are not as popular in catalysis as the amorphous materials due to their low specific surface areas.30 The most common form of SiO2 used in the SOMC studies is the pyrogenic (fumed) silica, which is produced in the flame hydrolysis process using the volatile silicon precursor, typically SiCl4, to afford a particulate material with well-controlled particle size. What makes this form of silica particularly attractive for SOMC applications is the high specific surface area, which can span from 50 to 400 m2/g (200 m2/g is most commonly used).31 The surface of silica has been extensively studied and we presently have a good understanding of its behavior during the thermal treatment based on the Zhuravlev model;32 we will briefly summarize it here. Before any pre-treatment the surface of the silica is fully hydroxylated (approximately 5 silanol groups per nm2)33 and covered with hydrogen-bonded and physisorbed water. At approximately 190
C the adventitious water present originally in the material is removed, leaving behind only the silanol moieties (maximum hydroxylation). Only from this point onwards the surface of SiO2 becomes of interest from SOMC perspective. Utilizing the classification introduced above, the surface is now exclusively covered with X-type moieties, and thus is best described as Xx ligand in the Green notation. Upon increasing of the pre-treatment temperature above 200
C the condensation of the silanol groups begins with the formation of siloxane bridges. The Si–O–Si bridges can be classified as L-type functionality. From this point onwards it is possible to distinguish three kinds of silanols on the surface: isolated, vicinal and geminal. The isolated silanols as the name suggests do not interact with neighboring moieties and have a specific IR signature at 3747 cm 1.34 Vicinal Si–OH groups are engaged in hydrogen bonding interaction with each other and appear as a broad IR features at lower wavenumbers. Finally, geminal sites correspond to a single Si center bound to two OH, that do not form OH bonds and have therefore an IR signature similar to isolated OH groups.35 With increasing temperature, the relative ratio between these types of sites changes with the simultaneous decrease in the total amount of Si-OH groups. A common pre-treatment temperature is 700
C, leading to a material typically coined silica-700 or also SiO2-700. This temperature results in a material with a high surface area, no longer containing vicinal silanol groups (which cease to be present on the surface at 400
C), and with the density of isolated Si–OH moieties falling below 1 OH per nm2. With that low concentration of surface silanols in SiO2-700 one can expect a preferential formation of monografted species compared to silica pre-treated at lower temperature. This has been demonstrated by grafting of tris(neopentyl)neopentylidenetantalum(V) on the surface of silicas pre-treated at temperatures ranging from 300 to 700
C.36 While reaction with SiO2-700 leads to primarily monopodal species, the bisgrafted sites predominate on silica-300 (Fig. 3A). Such behavior has been also observed for other metals.37,38 Predictably, reaction of the precursor with the support pre-treated at intermediate temperature, SiO2-500, leads to mixture of both species.39 Silica-700 can act as X (e.g., (^SiO)Ta(CH2tBu)4, Fig. 3B left)40 or LX (e.g., (^SiO)Re(^CtBu)(]CHtBu)(CH2tBu)(L), L ¼ siloxane bridge, Fig. 3B right)41 ligand depending on the precursor used. It must be noted that the minor geminal sites persist in the material prepared at this temperature and have been shown to engage in grafting.42 At around 900
C these geminal silanols are completely removed and at 1200
C the surface of silica contains only siloxane bridges turning silica into formally an L-type ligand. However, heating silicas (particularly those with high surface areas) to such high temperatures is associated with the formation of highly strained and reactive siloxane bridges43 along with a significant loss of surface area and the collapse of porous structure when present;30,44 this collapse can also occur at significantly lower temperature (500
C) depending on the conditions (vacuum, synthetic air, steam.). To date, no well-characterized example of surface species bound purely in L-type fashion through siloxane bridges has been reported. However, these Si–O–Si moieties (particularly the strained ones) can still engage in a reaction with molecular precursor, during which they undergo opening (Fig. 3C). This Si–O–Si opening can occur either during the grafting process or following the heat treatment applied to already grafted surface moieties (with or without other reactants).45–52 To conclude our discussion of silica, thermal treatment is not the only way to decrease the number of OH groups present on the surface. A broad array of chemical approaches exists for passivation of surface silanols using organosilicon reagents such as hexamethyldisilazane.53,54 Such chemical approach to decreasing of Si-OH content is unique as it does not lead to generation of strained siloxanes on the surface. Aluminas (Al2O3)55,56 are another broad class of oxidic materials, which found use as supports and as catalysts themselves. These materials can be obtained in dehydration of aluminum(III) hydroxides; the thermodynamic product of such processes is corundum (a-Al2O3), which forms hexagonal crystals. This phase finds use in specialized catalytic applications requiring high stability and inert character.30 The big disadvantage of this material is its low specific surface area and research efforts in formation of high surface area particles are ongoing.57 Typically employed in catalysis are the alumina phases formed at intermediate temperatures, the so-called transition(al) aluminas,58 as they have wide thermal stability windows, exhibit large specific surface areas and expose highly reactive Lewis acid sites in specific facets (vide infra). SOMC research has mostly focused on employing g-Al2O3 and related phases (d) as a support. Compared to silica, understanding of these materials proved much more challenging. Even though g-alumina is crystalline, which should facilitate the characterization, uncertainties regarding its structure persist. Traditionally, the galumina structure has been described as that of a cubic defective spinel, AB2O4.59 Classical spinel structure contains þ II and þ III cations, while g-Al2O3 is assembled only from trivalent cations. Therefore, in order to maintain the charge neutrality, the presence of

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

71

Fig. 3 Silica as SOMC support; (A) difference between mono- and bisgrafting; (B) examples of binding modes; (C) example of reactivity involving opening of siloxane bridge.

vacancies is necessary. These can arise from missing cations either in tetrahedral or octahedral environment (both geometries are present in the spinel structure), but the latter seems more plausible based on spectroscopic evidence.60 As some of the experimental observations could not be easily explained by the spinel structure, an alternative theoretical model was proposed based on the topotactic transformation of boehmite, one crystalline form of AlO(OH).61 This non-spinel model has been used extensively to rationalize the experimental observations regarding the interfacial region of g-Al2O3. Recently, the model has been extended to include edges between (110) and (100) planes.62 The complex structure of bulk g-alumina and associated experimental uncertainties have direct influence on our understanding of its surface. The first thing that needs to be pointed out is that the crystalline particles expose various facets depending on their morphology. A crystalline facet can be considered as a slice through the crystalline bulk along various crystallographic planes, which results in the material particles presenting variety of well-defined surface sites. Transmission electron microscopy studies revealed that for g-alumina the commonly exposed facets are the (110), (100) and (111).63 The theoretical approach that has been used to understand the surface is based on sequential rehydration of fully dehydroxylated surface of these facets using the periodic nonspinel DFT model and comparison of the resulting energetics.64 Using this methodology, one would anticipate the presence of pentacoordinate, tetracoordinate and tricoordinate aluminum(III) sites on g-alumina at zero water/hydroxyl coverage (qOH ¼ 0). Undercoordinated Al(III) sites exhibit strong Lewis acidic character, which is the most significant difference when comparing gAl2O3 to silica in its properties as a support. Although, such complete dehydroxylation of alumina surface is not feasible experimentally due to the limited thermal stability of material (e.g., g-Al2O3-500 and g-Al2O3-700 exhibit the density of chemically accessible OH groups of 2 and 0.7 per nm2, respectively)30 there is an evidence for the existence of strongly Lewis acidic sites on partially dehydroxylated g-Al2O3. Possibly the most intriguing of these sites is the tricoordinate Al(III), which has been shown to coordinate dinitrogen, a molecule notorious for its low propensity for acting as a ligand in molecular coordination chemistry.65 A lot of theoretical work has been devoted to rationalization of existence of such highly undercoordinated sites in g-alumina pre-treated at relatively mild temperatures. The emergence of this very active site begins above 400
C and reaches maximum concentration during

72

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

pre-treatment at 700
C (0.03 sites per nm2 based on methane dissociation).66 The 100 facet is fully dehydroxylated already above 350
C, which means that the first molecule of water introduced in the theoretical rehydration process (leading to qOH ¼ 3 per nm2, slightly higher than the experimental value for g-Al2O3-500) will interact preferentially with the 110 surface.67 The outcome of such interaction, which is thermodynamically most feasible, involves heterolytic splitting of the water molecule across the threecoordinate Al-O bond, inevitably quenching the reactive tricoordinate aluminum(III) site. However, it has also been proposed based on computations that metastable structure in which the tri-coordinate site remains accessible and water molecule interacts with the two tetra-coordinate Al(III) environments (forming a bridging OH) is not prohibitively uphill in terms of overall energetics.68 This can rationalize the persistence of these defect sites under common pre-treatment conditions. One additional and unexpected observations resulting from the described theoretical studies is the apparent increase of the Lewis acidity of one of the tetracoordinate Al(III) sites after the first molecule of water is introduced. This effect is most pronounced if the water molecule quenches the tricoordinate Al(III) site. The studies outlined above highlight some important facts regarding the properties of the partially dehydroxylated g-Al2O3 as a ligand. Firstly, as apparent from the theoretical hydration studies described above there is a variety of OH groups that can be present on the surface of the material. While the X-type behavior is predominant (e.g., surface (^AlO)W(^CtBu)(CH2tBu)2, Fig. 4A left),69 the presence of Al-O-Al can also lead to LX-type binding on the surface of g-alumina (e.g., (^AlO)Au(CH3)2(L), Fig. 4A right).70 In addition to monografted species there is also possibility of formation of bisgrafted surface moieties, which may undergo subsequent reactivity (e.g., grafting of tetrakis(neopentyl)zirconium and hafnium on the surface of g-Al2O3-500, Fig. 4B).69,71 Like in the silica case, the pure L-type behavior has not been documented in a well-defined surface species. The presence of Lewis acidic sites opens up a new grafting pathway, in which transfer of the anionic ligand from the metal precursor to Al(III) site can be observed as a primary grafting mechanism. An example of such behavior can be observed when (Cp*)Ta(CH3)4 (Cp* ¼ pentamethylcyclopentadienyl) is deposited on highly dehydroxylated alumina (in addition to typical grafting product, Fig. 4C).72 Finally, the proximity of the highly Lewis acidic aluminum(III) sites to the Lewis basic oxo linkages offers some unique reactivity with small molecules such as heterolytic splitting of dihydrogen or methane (Fig. 4D, top).73 Similar behavior has been observed when certain metal precursors such as methyltrioxorhenium (CH3ReO3) are grafted on the surface of g-Al2O3 (Fig. 4D, bottom).74 While the chemistry of the two supports described above provides a broad overview of the ligand properties expected from metal oxides as supports in SOMC research, some differences can be expected on a case by case basis. A good example of a support which shows a distinct surface character is magnesium oxide (MgO). This crystalline oxide adopts a simple cubic structure of NaCl with the singly coordinatively unsaturated (five coordinate) Mg2þ and O2 surface sites on the primary 100 termination.75 Hydration studies showed that for this facet the dissociative water adsorption occurs partially upon approaching a monolayer coverage, but is not favored thermodynamically for isolated water molecule.76–79 The surface of MgO contains also a variety of defects,80 which also undergo hydration with formation of hydroxyl moieties.81,82 The unique feature of this support is that it has a pronounced basic character arising from the presence of both O2 and OH groups.83,84 The discussion of other less common supports can be found in a recent review.85 Significant research efforts have been devoted to developing and utilization of high-surface area porous materials in catalysis as they could maximize the loading of the catalytically active species, its dispersion and impose confinement effects on the reaction outcome.86 Consequently, these materials are also of immense interest as supports in SOMC approach. With the success of silica as a platform for surface organometallic chemistry it was a logical next step to explore the surface chemistry of periodic mesoporous silicas (PMSs).87 These materials exhibit an ordered network of pores with the walls of the material remaining amorphous. In terms of their chemical behavior periodic mesoporous silicas closely resemble amorphous silicas described above. A related class of materials, periodic mesoporous organosilicas (PMOs), were also targets for SOMC functionalization.88,89 The main difference between the PMS and PMO is that the latter is assembled with organosilicon linkers, and thus are hybrid inorganic-organic materials.90 The additional organic component allows for extended tunability of the framework’s physical and chemical properties in addition to the presence of Si-OH moieties available for grafting.91 Another class of porous materials, which are explored in the context of SOMC research are zeolites.92 These materials found a widespread use in industrial applications due to their structural variety, shape selective properties, tuneable chemical character and good stability under harsh conditions.93,94 Zeolites are crystalline aluminosilicates, which are assembled from tetrahedral Si(–O–)4 and Al(–O–)4 units.95 While the former building block is neutral, the latter is negatively charged, and the presence of charge balancing cations is necessary within the framework. Introduction of protons into the material leads to Brønsted acidic character (X-type), but Lewis acidic character has to also be considered in some cases.96 The well-defined structure associated with crystalline order on the atomic scale holds promise for very narrow distribution of surface sites, and consequently grafted catalytically active moieties.97 So far, the majority of examples of SOMC in zeolites involves few late transition metal complexes as well as gallium98 and tin99 alkyl species. Despite these demonstrations of SOMC within zeolites, the significant drawback of these materials is their primarily microporous character.100 This largely limits the pool of precursors that can be used to functionalize the internal surfaces of the material. This issue could be potentially ameliorated by using hierarchical101 and two-dimensional zeolites,102 but such studies remain rare. Recently, metal-organic frameworks (MOFs)103,104 made its debut as supports in SOMC applications. MOFs are hybrid materials, which are assembled from the organic linkers (most commonly a multitopic carboxylates) and inorganic nodes (metal ions or clusters). The combination of these components gives rise to discrete crystalline compounds, which often exhibit recordsetting surface areas105,106 stemming from the presence of mesopores. Thus, these materials can accommodate a larger variety of

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

73

Fig. 4 Alumina as SOMC support; (A) examples of binding modes; (B) variety of surface species formed in grafting of tetrakis(neopentyl)zirconium and hafnium; (C) surface species resulting from grafting of (Cp*)Ta(CH3)4; (D) examples of reactivity involving Al–O fragment.

precursors used in SOMC approach without suffering from the diffusion limitations. There are many ways to functionalize MOFs,107–109 but the approach closest to the oxide parentage of the SOMC approach is the post-synthetic modification of inorganic nodes. As an example, we will look at zirconium-based MOFs as their node chemistry and catalytic applications have been studied extensively.110,111 The archetypal group of isoreticular Zr(IV) MOFs, the UiO family,112,113 feature the [Zr6(m3-O)4(m3-OH)4]12þ node, which can be viewed as small and crystallographically well-defined metal oxide cluster (charge neutrality in the material results from the negative charges of carboxylate linkers coordinated to the node). With the presence of the well-defined OH moiety one could anticipate an X-type behavior akin to that of the oxides mentioned above. Indeed, it has been shown that simple organometallics react with the bridging hydroxyl moieties resulting in the formation of surface-bound species.114,115 This result clearly proves the accessibility of these groups and viability of the material as a SOMC support. As the UiO MOFs possess relatively small

74

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

pore apertures (particularly the ones with shorter linkers), a lot of attention has focused on mesoporous Zr(IV) MOFs. Probably one of the most prominent examples of these large pore materials is NU-1000,116 a MOF composed of the pyrene-based linker and [Zr6(m3-O)4(m3-OH)4(OH)4(H2O)4]8þ node. The largest pores in NU-1000 have a diameter of approximately 30 Å making it (and its Hf(IV) analog) an excellent candidate for grafting of single site-catalysts and studying their activity. It is worth pointing out, that the inorganic building block of this MOF (as well as those from UiO family) can be dehydroxylated at high temperature (under vacuum) leading to [Zr6(m3-O)8]8þ node (with preservation of the material’s crystallinity).117 The undercoordinated Zr(IV) sites generated in such manner have strong Lewis acid character. Despite their desirable characteristics there are some limitations to utilization of MOFs in catalysis. The presence of the organic components within these materials significantly limits their thermal stability, especially when compared with the purely inorganic supports. Several examples of controlled functionalization of this material with well-defined organometallic species have been published to date.118–123 A recent SOMC example by Thiam et al. looked at grafting of W(^CtBu)(CH2tBu)3 on the node of NU-1000.124 The combination of hydroxy and aqua ligands led to more complex transformation of the precursor than in the case of a simple grafting on highly dehydroxylated oxide supports and resulted in the surface (^ZrO)W(]O)(CH2tBu)3. These sites evolve under the olefin atmosphere to form species active in propylene metathesis.

6.04.5

Tailored molecular precursors for well-defined surface species

Aside from choosing the metal center suitable for the desired transformation, the selection of an appropriate molecular precursor entails three important points of consideration: the leaving group, the remainder of the coordination sphere that will persist bound to metal center and its nuclearity. First, we will discuss the selection of the ligand(s) that will be removed during grafting as this aspect builds up on our discussion of the surface as a ligand. Most of the functionalization in the SOMC fashion occurs via reaction with surface hydroxyl groups. Thus, the immobilization of the desired species occurs mainly via acid-base chemistry, where the basic ligand present in the molecular precursor undergoes protonation and is replaced with the X-type surface functionality. In that regard, hydrocarbyl moieties125,126 (e.g., R ¼ alkyl, aryl, allyl as well as cyclopentadienyl ligands and others) gained a lot of attention as they are often very basic, and thus highly reactive towards surface hydroxyls (particularly when bound to early transition metals). In addition, the protonation leads often to simple hydrocarbons (most ideally gaseous product such as methane, Fig. 5A), which do not adsorb on the surface and further drive the reaction to completion i.e., quantitative grafting on all accessible isolated hydroxyl moieties. However, preparation of these molecular precursors might be challenging, and it is not uncommon for them to decompose at low temperatures and/or exposure to light. Metal amides127 are a viable alternative to strongly basic hydrocarbyl ligands. While still highly reactive towards surface OH groups, these NR2 moieties often pose less synthetic and handling challenges. Their protonation leads to formation of amines (Fig. 5B), which can also be removed with relative ease after grafting. While metal amides with smaller R groups tend to aggregate, the bulkier substituents often lead to mononuclear complexes. In particular, the bis(trimethylsilyl)amide derivatives are commonly employed as they are widely available for majority of metals in various oxidation states. Grafting of these metal derivatives releases hexamethyldisilazane, which can subsequently passivate some of the available OH groups leading to lower than theoretical loading of the desired metal.128,129 Metal alkoxides130 are a family of precursors which bear close resemblance to the surface O-anchorages as both are X-type oxygen-based ligands. RO moieties are in general less basic than amides and contain only one substituent leading to less steric bulk, and consequently higher chance for agglomeration. Their grafting results in release of alcohols as by-products (Fig. 5C). A related class of ligands are metal siloxides,131 which show similar electronic behavior (albeit they have weaker donor abilities) to alkoxides, and consequently similar grafting behavior (Fig. 5D). Another commonly employed oxygen-based ligands are acetylacetonates (acac),132 which can be easily obtained commercially for the majority of metals (Fig. 5E). This class of complexes have been used in functionalization of MgO, where upon grafting the acac moiety can partially transfer to Mg2þ sites.133 Finally, highly reactive metal chlorides134–136 and oxychlorides46,137,138 have also been used in the SOMC approach. Grafting of theses precursors (Fig. 5F) results in the release of HCl, which may lead to adventitious side reactivity. In addition, the characterization of surface halides is often more complex as there are hardly any specific spectroscopic handles. As outlined above, most of the grafting reactions occur via the acid-base chemistry. However, the chemical character of the metal precursor itself may also enable an unusual grafting mechanism as exemplified by the reaction of Ir pincer (POCOP ligand with tBu substituents on phosphorus atoms) complexes active for catalytic hydrogenation of alkenes with the mesoporous silica SBA-15.139 While the (POCOP)IrIII(H)2 complex undergoes an expected grafting with the evolution of dihydrogen resulting in ^SieOeIrIII(H)(POCOP), the related (POCOP)IrI(alkene) complexes graft via oxidative addition into OeH bond (and concomitant release of the olefin) resulting in the same surface species. One of the unique features of SOMC functionalization, which stems from the molecular-like chemistry involved, is that we can anticipate (based on the known reactivity arguments) which ligands are going to depart metal coordination sphere and which are going to stay bound to the metal center following the grafting process. This knowledge has important implications for generation of catalytically active single sites on the surface. In this context, one of the most important benefits of SOMC approach is the possibility of choosing a precursor, which already contains the proven (or putative) catalytically active moiety. This allows us to either directly access the catalytic cycle or test hypotheses regarding the nature of active sites in yet poorly understood systems.140 The classical methods of supported catalyst preparation commonly involve impregnation of the support with a simple salt of the desired metal, followed by the calcination of the material, which often leads to ill-defined species present on the surface.10 These are often only

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

Fig. 5

75

Grafting of the most common precursors on the surface of silica.

pre-catalysts, which still must undergo a chemical transformation leading to generation of the catalytically relevant moieties (e.g., metal alkylidenes for olefin metathesis).141 Those transformations often happen on stream and understanding of the catalyst structure under operando conditions is often challenging. Fortunately, in many catalytic transformations we can anticipate which moiety is necessary to bring about the desired reactivity. This understanding is often a result of complimentary studies on homogeneous catalytic systems, which are often less challenging to characterize and have a wealth of well-defined reactivity. Additional reactive groups can be accessed on the surface via targeted reactions, even if the precursors containing these functionalities are not readily available. This is the case for many early transition metal surface hydrides, which can be conveniently obtained by hydrogenolysis of grafted hydrocarbyl derivatives (Fig. 6A).142 The same principle can also be utilized when the precursor containing the desired functionality is highly unstable. As an example, homoleptic high-valent tungsten peralkyls acting as precursors of surface bound WRn moiety can explosively decompose and their handling is not straightforward.143 The alternative and less dangerous route involves grafting of tungsten(VI) chloride first to form surface bound WClx species, which can then be alkylated leading to the desired moieties (Fig. 6B).136 It is important to note that the post grafting alkylation is not trivial as adventitious alkylation of the oxide surface may also take place. One of the important ways to obtain the undercoordinated and highly active metal sites on the surface is to thermally decompose the remaining ligands coordinated to the metal center under vacuum. As an example, one can decompose surface M(NR2)n moieties resulting from the grafting of metal amides. However, in some cases the nitrogen migrates to the surface forming Si– N–M connectivity in addition to oxygen linkages present initially on the surface of the oxide.144 A much more elegant approach, termed Thermolytic Molecular Precursor (TMP),145 proved a powerful method of formation of single sites in combination with SOMC approach. The idea behind the TMP methodology involves preparation of well-defined complexes featuring the desired metal center and a ligand, which under thermal treatment decomposes with traceless removal of organic component leading to

76

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

Fig. 6

Examples of post-grafting reactivity; (A) formation of surface metal hydrides; (B) surface alkylation; (C) and (D) TMP approach.

purely inorganic materials. For example, the thermal decomposition of metal siloxides yields silicas with high loading and homogeneous distribution of desired metal. This approach has been combined with the SOMC principles by first grafting the TMP precursor on the support followed by thermal treatment to yield “naked” surface metal sites.146 If grafting of the siloxide precursor is done on silica, then the remaining ligands are eventually transformed into part of the support (Fig. 6C). Alternatively, one can graft siloxide precursors on other oxides such as alumina, where the thermal treatment leads to mixed oxide environment in the vicinity of the catalytically active metal center (Fig. 6D). It is also worth pointing out that the thermal decomposition process leads to reappearance of OH groups in the material, which allows for subsequent functionalization and formation of well-defined multimetallic systems. The last important point of consideration in picking the molecular precursor for functionalization of the surface is its nuclearity. As described above some of the precursors are not monomeric and form well-defined polynuclear species (dimers, trimers, tetramers etc.), which can be present in solid-state and persist in liquid phase. These complexes can graft with and/or without preservation of their original nuclearity. Finally, it is also possible to form polynuclear species starting from the mononuclear precursor. Such flexibility in altering the nuclearity of surface species is very useful in studying its influence on catalytic activity. For example grafting of the dimeric rhodium precursor with bridging methoxy ligands, Rh2(m-OCH3)2(COD)2 (COD ¼ cyclooctadiene), on magnesium oxide proceeds with preservation of the dimeric structure.147 Interestingly, the dimeric structure also persists upon exposure to H2 at elevated temperature. Comparison of the activity of this material for catalytic H-D exchange with the related, but initially monomeric olefin complex Rh(C2H4)2 on MgO shows higher activity for the former signifying the importance of the neighboring metal site. The classical methods used for heterogeneous catalyst preparation often do not allow for such precise control of the resulting sites.

6.04.6

Characterization techniques utilized in the SOMC approach

One of the critical aspects of the SOMC approach is the rigorous characterization of the resulting surface species, which ideally should parallel the precision with which we can understand molecular compounds. This includes quantitative identification of all the system components, understanding their electronic state, their spatial distribution as well as any bonding between the individual constituents. Thus, the development of the SOMC approach has always been closely tied to advances in the techniques and instrumentation used for the structural elucidation in solid-state materials. Nowadays, the access to sophisticated characterization methods provides us with intimate details about the surface sites with atomic precisionda condition necessary for rational design and preparation of single-site catalysts. We will briefly summarize the state-of-the-art characterization toolkit for SOMC-derived single sites and relate it to approaches used in the molecular coordination chemistry. It is crucial to first determine the quantity of surface hydroxyl groups available before grafting of the molecular precursor on the surface is attempted. This knowledge enables control over metal precursor:surface OH group stoichiometry. Additionally, it allows

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

77

for comparison between the theoretical and experimental metal content on the surface following the grafting. While various methods can be applied to determine the amount of OH groups on the surface,148 typically chemical titration with organometallic reagents followed by quantification of the released protonated alkyl/aryl group is utilized. This approach is most relevant in the context of SOMC as the quantification itself is a grafting process. It is important to note that this method cannot be applied to reducible oxides or materials with other moieties reactive towards organometallic reagent as the results may not be reliable. With the information regarding the number of surface OH groups in hand, the reaction with the molecular precursor can be performed. The initial question is whether the surface functionalization occurred at all. One of the basic tools used to probe that is Fouriertransform infrared spectroscopy (FTIR),149,150 which clearly shows perturbation and consumption of the bands associated with the surface hydroxyl groups. Additionally, one can potentially observe unique IR features associated with the ligands still bound to metal center after grafting and washing of the material. In that regard Raman spectroscopy151,152 (the resonant variant of this technique is particularly suited for the characterization of dilute surface sites) often can provide complementary information to that obtained in FTIR measurements (presence of oxo species). The analyses of the vibrational manifold can be aided by usage of probe molecules,153 which experience shifts of their characteristic bands depending on the nature of interaction with the surface site of interest. Then, the elemental analysis provides quantitative insights about the metal loading as well as the content of other elements such as carbon, hydrogen and nitrogen, which can facilitate closing of the mass balance and validate the postulated grafting mechanism and stoichiometry. It is not uncommon that the metal ion can change its oxidation state during grafting process or under the catalytic reaction conditions. Several spectroscopic tools can be utilized to analyze the oxidation state of the metal center such as electron paramagnetic resonance (EPR) spectroscopy154 and X-ray photoelectron spectroscopy (XPS).155 EPR is a particularly valuable technique in identification of the species with unpaired electrons (paramagnetism) located on the introduced metal sites, ligands bound to it or the support itself. XPS provides information about the oxidation state based on the energy of electrons, which are emitted from the material upon irradiation with high energy photons. It can also provide information about the abundance of the elements in the sample. The increasing accessibility of near-ambient pressure XPS (NAP-XPS)156 provides an opportunity to investigate the catalyst in the presence of catalytically-relevant substrates. Finally, UV-Vis spectroscopy157 can also be used to infer the oxidation state of the metal site, the presence of chromophores on the surface with characteristic absorption bands as well as some information on particle sizes. All the techniques mentioned above are routinely used in the molecular coordination chemistry. Although the same applies to nuclear magnetic resonance (NMR) spectroscopy,158 the measurements in the solid-state come with a unique set of challenges and opportunities.159 NMR is an extremely versatile technique, which is nucleus-specific and can be applied to a variety of main group elements as well as metals in the d block, thus giving possibility of characterization of the support, grafted metal as well as the ancillary ligands bound to the metal. The signals with different chemical shifts arise from nuclei in differing environments within the sample and can be used for their identification. Additionally, the couplings between the neighboring nuclei (either through bonds or space) can be correlated with each other to elucidate the intimate structure of the site of interest. Their low density on the surface can be challenging when signal is to be obtained within a reasonable measurement time, but with the variety of cross polarization sequences a significant sensitivity enhancement can be achieved. Recently, the developments in dynamic nuclear polarization (DNP), a technique in which the transfer of polarization occurs from external organic radical to the desired nucleus (typically via the proton bath) with the assistance of microwave radiation, allowed for significant enhancement of signals associated with surface single-sites and opened a new vista in characterization of heterogeneous catalysts prepared via the SOMC approach.160,161 It should be noted that due to the lack of averaging present in the liquid phase the solid-state spectra typically exhibit more complicated line shapes. Single crystal X-ray diffractometry is a technique which is routinely utilized in characterization of molecular complexes and which provides a snapshot of the investigated molecule.162 The wealth of the information that can be extracted from a refined crystal structure allows for identification of extensive structure-function relationships. Powder X-ray diffraction163 is useful when the crystallite size is unsuitable for single crystal measurements. Application of those techniques has limitations in characterization of the supported species prepared using SOMC approach as most commonly the supports themselves are amorphous. Even in the case of supports, which are crystalline, it is not a given that the introduced surface moieties are not disordered within the crystalline material, and thus undetectable using these X-ray diffraction techniques. The development and increased accessibility of synchrotron radiation led to widespread utilization of techniques such as X-ray absorption spectroscopy (XAS),164 which can provide local structural information akin to that obtained from a single crystal XRD measurements. This element specific spectroscopic technique consists of two main spectral regionsdX-ray absorption near edge structure (NEXAFS) and extended X-ray absorption fine structure (EXAFS). When the incident beam energy matches the binding energy of a core electron for the desired element, an absorption edge is observed. After the edge energy is reached the ejected core photoelectron can interact with the immediate coordination spheres leading to oscillations in the EXAFS region of the spectrum. The analysis of these oscillations allows us to identify the nature and quantity of scattering atoms as a function of distance from the absorbing atom. With the aid of computations, a reliable model(s) of the surface site can be derived and validated against other techniques such as pair distribution function (PDF). Aside from the structural information encoded in the EXAFS spectrum, the pre-edge features in the NEXAFS region can also provide information about the symmetry around the absorbing nucleus. In addition, careful analysis of the pre-edge features and the edge itself may allow for determination of the oxidation state as well as the electronic structure. Some of the characterization techniques are unique to solid-state materials and are not commonly used in molecular inorganic chemistry. The measurement of adsorption isotherms165 of the material using various probe gasses and temperatures can provide important insights into the porosity of the support. As mentioned earlier, the high specific surface area is a must for a support to be

78

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

of use in the SOMC approach. The measurement of the gas uptake into pores of the support followed to fitting to appropriate models (e.g., Brunauer-Emmett-Teller theory) can provide the specific surface area of the material. In addition, the pore size distribution can also be determined in many cases. Finally, the shape of the isotherm (both at the adsorption and desorption stage) can hint the nature of the material (micro-, meso- or macroporous). This knowledge is essential as grafting of a moiety within support’s porous structure can significantly alter the surface area measured after grafting. This information can be used to determine the mechanism and/or location of the moieties within the support. One additional aspect, which is not an issue in the molecular chemistry is the spatial distribution of sites within the material. The inherent kinetics of the process, diffusion limitations or nuclearity changes during grafting can affect distribution of sites on the support. The remarkable progress in the development of the microscopic techniques allows now for collection of surface images with atomic resolution.166,167 Two main techniques used to visualize solid-state materials are scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The images provided by SEM result from the impact of focused electron beam on the surface of the material which leads to ejection of electrons from the material within a certain depth. Those ejected electrons are then used for visualization of the material surface. SEM can be integrated with the energy-dispersive X-ray spectrometer (EDX), which can assign the energy of ejected electrons to the individual elements and establish ratio between them. Based on that data SEM-EDX can form maps of elemental distribution in the sample. While SEM probes mainly the surface, TEM can provide information about the bulk of the sample and offers higher resolution. The thin sample is placed in an electron beam and the analysis of the transmitted electrons gives rise to the sample image. The advantages of two techniques can be combined in scanning electron tunneling microscopy (STEM), which can provide images of large sample areas with the resolution below 1 Å. This allows for visualization of single atom on the surface and coupled with electron energy-loss spectroscopy (EELS)168 allows for very precise electron mapping and extraction of information akin to that obtained in XAS measurements. Finally, one should not overlook the importance of the computational approaches169 used in understanding of the surface species. The predictive nature of the molecule-like transformations central to SOMC concept allows for prediction of potential surface site structure present on the material. The geometry of these sites can be optimized using density functional theory (DFT) methods using the molecular, cluster or periodic models.170 The optimized geometries of the surface species can be compared in terms of their relative energies to establish the feasibility of models/grafting pathways when compared to each other. The optimized geometry models can be used to predict the spectral data for majority of spectroscopic characterization techniques. All these predictions combined can be tested against the experimental data to identify the most credible model of the surface site. In the case of more complex systems and/or species under catalytically relevant conditions a more sophisticated computational approaches need to be utilized in order to realistically model the evolution of surface sites.171,172

6.04.7

Selected success stories in SOMC single site catalysis

In the last section of this article we will look at the selected examples of the successful utilization of SOMC approach in the formation of single-site catalysts. This list is not exhaustive; the aim was to identify examples, which best illustrate the principles described in previous sections. As it was stated in the introductory part, one of the archetypal examples of single-site catalysis is the formation of polyolefins using early transition metal complexes. Thus, it is unsurprising that heterogeneous olefin polymerization173 became a transformation of interest in early SOMC research. The initial studies by groups of Ballard18,174 and Yermakov175,176 targeted supported early transition metal alkyls and hydrides as catalysts for this transformation. Basset and colleagues looked at surface hafnium hydrides obtained from hydrogenolysis of the (^SiO)Hf(CH2tBu)3 species at various temperatures.177 The mixture of hydrides obtained in such manner was active in polymerization of ethylene, propene and isobutene at room temperature. Another work from this group explored grafting of the metallocene complexes themselves on the surface of silica partially dehydroxylated at 800
C.178 Reaction of Cp*ZrMe3 with the surface of SiO2-800 leads to two kinds of species. The majority of the precursor reacts with silanols to form (^SiO)Zr(Cp*)(Me)2(L) sites (where L represents the adjacent siloxane bridge) resulting from the simple pronation of the methyl group. The minor species results from the opening of the siloxane bridge with the formation of (^SiO) Zr(Cp*)(Me)2 and transfer of the methyl group to the proximal Si site. It has been demonstrated in the homogeneous olefin polymerization catalysis literature that the presence of activator is necessary for efficient catalytic behavior of metallocene-based systems. The role of the activator, particularly Lewis acidic one, is to generate coordinatively unsaturated and electron deficient cationic species.179 Treatment of the surface organozirconium species mentioned above with B(C6F5)3 led to mixture of three cationic species, activating the material for the polymerization of ethylene. Approaches to generate catalysts active for this transformation without the need for post-grafting modification with activator has been described by the Marks group.180 One way to address formation of the active sites relies on taking advantage of the highly Lewis acidic sites present on the surface of the support. In this context highly dehydroxylated alumina (only 0.1 OH/nm2) was treated with organozirconium complexes.181 During grafting of the Cp2ZrMe2 on the surface of this support a transfer of the methyl group to Al(III) site occurs with concomitant formation of the cationic and surface-bound [Cp2ZrMe]þ species as evidenced by the solid-state NMR experiments (Fig. 7A). The material showed moderate activity towards formation of polyethylene, with only a fraction of sites active in this transformation. To account for this observation DFT calculations were performed using g-alumina model, which revealed different interaction strengths between the available surface oxo anchorages and the cationic organozirconium moieties.182 An extension of the studies in the Marks group involved moving to sulphated oxides as supports for metallocene species. The reason for exploration of this class of materials is related to the coordination strength of the available surface anchoring

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

Fig. 7

79

Selected examples of single-site catalysts accessed via SOMC approach.

points. During grafting on e.g., silica the deprotonation of the mildly acidic silanols generates a relatively strong conjugate base, the siloxide moiety, which forms strong bonding interaction with the organometallic moiety i.e., m-oxo linkage. Such strong interaction is detrimental to the activity of metallocene olefin polymerization catalysts. Conversely, the unique character of the sulphated oxide allows it to act more in line with weakly coordinating anion183,184 chemistry rather than the covalently bound ligand. As an example, we will look at grafting of organozirconium species on the surface of the sulphated alumina described by Nicholas et al.185 When this support is treated with (Cp*)2ZrMe2 an evolution of methane could be observed as in the conventional grafting on surface OH groups. However, the spectroscopic signature of the resulting surface sites was consistent with the cationic [(Cp*)2ZrMe]þ. With only minor presence of the methyl transfer to the Al(III) sites compared with the highly dehydroxylated alumina case, the authors surmised that grafting occurs primarily through protonation of the methyl group. This was confirmed by lack of spectroscopic signatures from the species bound via strong m-oxo linkages. The highest activity for ethylene polymerization was obtained with a modified precursor, (Cp*)ZrMe3. The activity of the resulting grafted species on sulphated alumina was 10 times higher than on the dehydroxylated alumina. Most notably, almost all surface sites were active in the desired transformation, a significant improvement over the pure alumina systems and an elegant demonstration of the single-site concept. One of the iconic applications of the SOMC is the exploration of heterogeneous olefin metathesis catalysis. This important reaction is realized industrially with ill-defined molybdenum and tungsten oxo species dispersed on the oxidic supports.186 A great deal of insights on the mechanism of the reaction and the nature of active species resulted from the investigation of well-defined molecular complexes.187 It is commonly accepted that the active species in this transformation are metal alkylidenes, which are the entry point into the so-called Chauvin mechanism.188 The alkylidene moiety is also presumed to form from the initially present oxo species in the case of the industrially-relevant catalysts.189 However, only a fraction of metal sites in these formulations are thought to transform into metathesis-competent species. The well-defined molecular alkylidene complexes are prone to bimolecular decomposition processes,190 which are facilitated by the translational freedom associated with the solution-based homogeneous chemistry. Thus, there is an ongoing interest in immobilization of these high-performing complexes on the surface to minimize the intermolecular interactions compromising the activity of the catalyst. The early studies involved grafting of Mo(^N)(CH2tBu)3 on the surface of silica,191,192 which undergoes selective grafting via the protonation of the nitrido moiety as judged by FTIR measurements. In addition, during grafting neopentane is evolved, likely a product of intramolecular a-abstraction process (removal of a proton from one alkyl substituent by adjacent alkyl group). Thus, the putative surface species were described as (^SiO)Mo(]NH)(]CHtBu)(CH2tBu). The surface species are active in ring-opening metathesis polymerization (ROMP) of norbornene already at 25
C, which is in line with the proposed presence of the alkylidene moiety. Under similar conditions, the parent molecular precursor showed no activity. With the introduction of the M(]NAr)(]CHtBu)(CH2tBu)2 (M ¼ Mo, W; Ar ¼ 2,6-di(isopropyl)phenyl) precursors it became possible to explore grafting of complexes with alkylidene moiety readily present.193,194 Reaction of these complexes with SiO2-700 leads to consumption of the majority of surface silanol groups with the formation of (^SiO)M(^NAr)(]CHtBu)(CH2tBu) species. In contrast to the previously mentioned studies, the spectroscopic signature of the alkylidene moiety could be observed in solid-state 1H and 13C NMR spectra. The well-defined character of (^SiO)Mo(^NAr)(]CHtBu)(CH2tBu) is underscored by the almost quantitative formation of the olefin resulting from the reaction of the substrate with the parent neopentylidene fragment. Finally, this surface-bound complex exhibited slower decomposition behavior compared to its molecular analog. Further synthetic elaboration of the molecular precursors allowed for preparation of the analogous (^SiO)Mo(]NAr)(]CHtBu)(NR2)195 and

80

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

(^SiO)Mo(]NAr)(]CHtBu)(OR)196 species. Over the years a variety of homogeneous molybdenum and tungsten olefin metathesis catalysts with different ligands have been grafted on silica. Most important modifications involved switching from the imido ligand196,197 to oxo moiety.198,199 Additionally, instead of the X-type alkoxide group the L-type N-heterocyclic carbene ligand has been introduced, which resulted in cationic variants of both imido200,201 and oxo202 derivatives. With the siloxide and alkylidene moieties kept invariant, it was possible to systematically investigate the influence of the changes on the activity of these heterogeneous catalysts. These studies identified the surface [(^SiO)W(]O)(]CHCMe2Ph)(IMes)þ][B(ArF)4] species (Fig. 7B) as one of the most stable and active sites towards both internal and terminal olefins.202 The comprehensive and rigorous comparison of the various surface sites available is a testament to well-defined and controlled molecular-like chemistry offered by the SOMC approach. A lot of studies focused on utilizing the SOMC approach to understand and improve early transition metal-based catalysts for epoxidation of olefins;203–205 in particular the formation of Ti(IV) single sites has been pursued extensively. One of the earliest examples of SOMC approach to formation of Ti single sites was that described by Maschmeyer et al., where titanocene dichloride was used to introduce titanium sites into the channels of MCM-41.206 Interestingly, the EXAFS spectra indicated that initially, the Cp2TiCl2 coordination environment remains mainly unperturbed within the porous structure. Only after addition of triethylamine, changes could be observed consistent with formation of the tripodal Ti(IV) sites capped with one cyclopentadienyl ring. Following the calcination, the material proved to be active in the oxidation of cyclohexene and pinene using tert-butyl hydroperoxide. The Tilley group initially used the TMP approach to thermally decompose (under variety of conditions) the Ti(IV) siloxide precursor, Ti(OSi(OtBu)3)4, to form silica with the high weight percentage of titanium sites within the material.207 Ti-containing silicas obtained in such manner proved to be active in the epoxidation of cyclohexene using tert-butyl and cumene hydroperoxides as oxidants. The catalytic performance was highly dependent on the surface area of the material indicating different accessibility of the Ti(IV) sites. Moreover, octahedrally coordinated titanium was observed in addition to tetrahedral sites, which are proposed to be the most active species for the desired transformation. In order to prepare a truly single-site catalysts the precursor was grafted on the surface of the silica leading to isolated Ti(IV) sites, which surpassed the commercial Shell catalyst in cyclohexene epoxidation activity. These initial and promising results were extended to an array of precursors grafted on the high surface area materialsdMCM-41 and SBA-15.208 The precursors useddTi[OSi(OtBu)3]4, (iPrO)Ti[OSi(OtBu)3]3, and (tBuO)3TiOSi(OtBu)3ddiffer in their size which directly translates to titanium loadings that can be achieved on mesoporous silicas. The comparison of activity between the surface-bound species identified (iPrO)Ti[OSi(OtBu)3]3 deposited on SBA-15 as the most active in the epoxidation of cyclohexene with cumene hydroperoxide. It is worth noting that in grafting of this material no iPrOH release could be observed indicating high selectivity in the ligand leaving the coordination sphere. Moreover, the UV-vis spectra suggest primarily isolated and tetrahedral, and thus single-site character of the resulting species. While the systems developed by Tilley group showed good activities when organic hydroperoxides were used, their performance with aqueous hydrogen peroxide as an oxidant was much poorer. The hydrophilic character of the catalyst surface resulting in strong interaction with water molecules was posited as the cause for the diminished activity of these catalysts. This issue could be partially alleviated by surface silylation of the calcined Ti containing mesoporous silica.209,210 An alternative approach to formation of the titanium single-site catalysts has been proposed by Notestein et al., who utilized bulky calixarene complexes of Ti(IV) as molecular precursors.211 The calixarene–Ti(IV)–Cl complex reacts with the surface of silica partially dehydroxylated at 500
C to yield a surface bound ^SiOeTiecalixarene species (Fig. 7C). The Ti-functionalized material showed activity in epoxidation of olefins using the organic hydroperoxides described above. The kinetic analysis performed indicated that the grafted species are the resting state of the catalyst. In contrast to the previous studies, the alcohol formed from the organic hydroperoxide during catalysis did not seem to inhibit the activity of the material. Under the reaction conditions the surface species did not seem to evolve, unless water was present in the mixture. In that case a loss of activity could be observed concomitant with dissociation of the calixarene ligands from the Ti sites. In a continuation of this work a systematic study of the influence of the support and the electron density on the calixarene ligand has been performed.212 While changing of the electron donating/withdrawing properties of the calixarene substituent had a minor influence on the catalytic performance, the surface acidity had a profound effect on the reaction rates. The analogous Ti–calixarene species grafted on supports with lower pKa of the surface hydroxyl moieties (Al2O3 and TiO2) showed 50 times lower initial epoxidation rates. The authors surmised that the intermediate strength of the Bronsted acid sites in the vicinity of the Ti site is crucial in stabilization of the hydrogen-bonded intermediates on the surface. The SOMC methodology has also been applied to generate single-site catalysts for non-oxidative alkane dehydrogenation.213,214 This reaction has gained a lot of attention in the context of propane valorisation to form propene. Precursors for metals such as vanadium,215,216 chromium,217,218 iron,219 cobalt,220 gallium221–223 and indium224 have been utilized to form surface sites active for this transformation in SOMC fashion. We will look closer on the case of Ga(III) single-site formation as they showcase the different implementations of the SOMC approach and signify the differences between supports themselves. Searles et al. approached the formation of the gallium single sites on SiO2-700 using Ga(III) siloxide precursor, [Ga(OSi(OtBu)3)3(THF)] with the aim of utilization of the TMP approach.222 In contrast to other siloxide-based precursor there was no evidence of HOSi(OtBu)3 being released. Instead, the products of siloxide decomposition could be detected already at the grafting stage. Analysis of these sites following drying under high vacuum suggested the presence of tripodal Ga(III) sites on the surface of silica capped with one equivalent of ROH (R ¼ Si(OtBu)3 or tBu). Following, the thermal treatment under high vacuum the residual ROH molecules could be removed to yield tetracoordinated Ga(III) sites with pronounced Lewis acidity as judged from pyridine adsorption probe experiments (Fig. 7D). To confirm the single-site character of the surface moieties and rule out the formation of dimeric species the wavelet transform analysis, a method to increase the resolution of the EXAFS data, has been performed. Indeed, only the presence

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

81

of monomeric sites could be detected. The material proved to be active in propane dehydrogenation with good selectivity for propene and slow deactivation kinetics. The initial turnover frequency surpassed that of the previously reported systems, which is proposed to be related to larger number of isolated sites in the SOMC-derived material. The analysis of the spent catalyst revealed that the isolated character of the sites is maintained during catalysis, but some reduction of Ga(III) to Ga(I) occurred consistent with fractional loss of activity over time. A more traditional SOMC approach to formation of gallium(III) based single-site catalyst have been explored by Szeto et al.223 Tris(isobutyl)gallium(III), Ga(iBu)3, was used as a precursor and grafted on both SiO2-700 and g-Al2O3-500. The analysis of the grafting processes revealed significant differences between supports. In the case of g-alumina the reaction of the precursor leads to monomeric [(^AlO)Ga(iBu)2(L)] species (L ¼ surface oxygen) as judged from the extensive spectroscopic and mass balance studies. During the grafting an almost stoichiometric amount of isobutane has been released with a fraction of isobutyl groups transferred to Al(III) sites. Conversely, an analogous reaction using silica as a support results in formation of dimeric sites with surface siloxides acting as both bridging and terminal ligands. Again, some alkylation of the silica occurred during grafting. The difference in the nuclearity of the surface species grafted on both supports could be clearly observed in the wavelet transform analysis, where GaGa contributions were evident in the case of functionalized silica. While both materials were active in the dehydrogenation of propane under flow conditions with good selectivity towards propene, the alumina-supported sites showed much higher initial conversion. Interestingly, this behavior has been attributed not to the different nuclearity, but to the beneficial influence of alumina as a support. Similar to the previous study, a slow deactivation of both catalysts could be observed over time. A wealth of research on heterogeneous catalysts based on late transition metals introduced via the controlled grafting of organometallic species was published by the Gates group. These studies concerned largely grafting of Group 9 complexes on crystalline supports such as zeolites and MgO.97 Reaction of the Ir(C2H4)2(acac) precursor with the surface of the dealuminated zeolite Y zeolite results in a selective loss of acac ligand and formation of the well-defined Ir(C2H4)2 sites bound to the support in a bipodal fashion.225 In addition, the lack of significant Ir–Ir contributions in EXAFS spectrum suggested that the mononuclear sites form selectively, with no multinuclear clusters present. The uniform nature of these Ir sites was evidenced by extremely narrow FTIR features arising from the CO probe molecule used to replace the bound ethylene ligands. The surface Ir(C2H4)2 sites are active in the hydrogenation of ethylene at room temperature to form ethane. The uniformity of the sites persists during catalysis under these conditions proving their single-site character. It is worth pointing out that the presence of the olefin ligands already on the metal center indicates that these species might already be part of the catalytic cycle for the hydrogenation of ethylene, exemplifying the principles of the SOMC approach. Grafting of the same precursor on MgO leads to analogous surface species (Fig. 7E), but with less degree of uniformity as judged from the CO probe FTIR experiments.226 Additionally, the study revealed that the support selection affects the electronic properties of the metal with more basic MgO resulting in more electron-rich Ir centers when compared with the zeolite-based catalyst. This directly influences catalysis with the MgO-supported species being less active in the ethylene hydrogenation reaction. It was shown that the structure of the site evolves when exposed to pure H2 (particularly at elevated temperatures) to transform the mononuclear sites to uniform sites with connectivity suggesting Ir4 stoichiometry (and presence of bridging ethylene and ethylidene ligands).227 Curiously, this agglomeration could be reversed when changing the gas flow to C2H4 (or controlled by the ethylene:hydrogen ratio).228 Those uniform clusters were showed to be more active for the catalytic transformation than the mononuclear sites on each support, but the improvement was much more pronounced in the MgO case.229 Due to the large difference between the atomic number of Ir and atoms making up the support a good contrast could be observed in STEM images allowing for visualization of the single sites before and after exposure to dihydrogen.230

6.04.8

Conclusion

As it was demonstrated in this bird’s-eye view summary, surface organometallic chemistry has established itself as a versatile approach to generating heterogeneous catalysts based on well-defined isolated metal sitesdso-called single-sitesdfor a variety of industrially relevant transformations. Rather than just a suite of techniques, SOMC should be considered as a way of looking at formation of catalytically relevant sites on the surfaces through the lens of molecular coordination chemistry. This is enabled by ligand-like behavior of the rigorously pre-treated surface of the support, utilization of tailored molecular precursors and leveraging the state-of-the-art characterization techniques. We hope that this primer will encourage utilization of SOMC approach in generation of more efficient catalysts and/or understanding the more complex systems generated with less controlled synthetic methods.

References 1. 2. 3. 4. 5. 6. 7.

Delidovich, I.; Palkovits, R. Green Chem. 2016, 18 (3), 590–593. Chorkendorff, I.; Niemantsverdriet, J. W. Concepts of Modern Catalysis and Kinetics, Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2003. Deutschmann, O.; Knözinger, H.; Kochloefl, K.; Turek, T. Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2011. van Santen, R. In Catalysis: From Principles to Applications; Beller, M., Renken, A., van Santen, R., Eds., Wiley-VCH Verlag GmbH, 2012; pp 3–19. Cole-Hamilton, D. J. Science 2003, 299 (5613), 1702–1706. Hagen, J. Industrial Catalysis: A Practical Approach, Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2015. van Santen, R. A. Modern Heterogeneous Catalysis, Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2017.

82 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67.

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts Groppo, E.; Martino, G. A.; Piovano, A.; Barzan, C. ACS Catal. 2018, 8 (11), 10846–10863. Munnik, P.; de Jongh, P. E.; de Jong, K. P. Chem. Rev. 2015, 115 (14), 6687–6718. Wegener, S. L.; Marks, T. J.; Stair, P. C. Acc. Chem. Res. 2012, 45 (2), 206–214. Goldsmith, B. R.; Peters, B.; Johnson, J. K.; Gates, B. C.; Scott, S. L. ACS Catal. 2017, 7 (11), 7543–7557. Nørskov, J. K.; Bligaard, T.; Hvolbæk, B.; Abild-Pedersen, F.; Chorkendorff, I.; Christensen, C. H. Chem. Soc. Rev. 2008, 37 (10), 2163–2171. Bhaduri, S.; Mukesh, D. Homogeneous Catalysis: Mechanisms and Industrial Applications, John Wiley & Sons, Inc., 2014. Axenrod, T. In Physical Methods on Biological Membranes and Their Model Systems; Conti, F., Blumberg, W. E., de Gier, J., Pocchiari, F., Eds.; NATO Advanced Science Institutes Series (Series A: Life Sciences)vol. 71; Springer: Boston, MA, 1985; pp 5–25. Kaminsky, W. J. Polym. Sci. Part A: Polym. Chem. 2004, 42 (16), 3911–3921. Kim, S. H.; Somorjai, G. A. Proc. Natl. Acad. Sci. U. S. A. 2006, 103 (42), 15289–15294. Thomas, J. M.; Raja, R.; Lewis, D. W. Angew. Chem. Int. Ed. 2005, 44 (40), 6456–6482. Ballard, D. G. H. J. Polym. Sci. Polym. Chem. Ed. 1975, 13 (10), 2191–2212. Yermakov, Y. I., Kuznetsov, B. N., Zakharov, V. A., Eds.; Studies in Surface Science and Catalysis; vol. 8; Elsevier, 1981. Basset, J. M.; Choplin, A. J. Mol. Catal. 1983, 21 (1–3), 95–108. Iwasawa, Y. In Advances in Catalysis, vol. 35, Eley, D. D., Pines, H., Weisz, P. B., Eds.; vol. 35; Academic Press, 1987; pp 187–264. Gates, B. C.; Lamb, H. H. J. Mol. Catal. 1989, 52 (1), 1–18. Basset, J. M.; Ugo, R. In Modern Surface Organometallic Chemistry; Basset, J.-M., Psaro, R., Roberto, D., Ugo, R., Eds., Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, 2009; pp 1–21. Sterrer, M.; Freund, H.-J. In Surface and Interface Science: Properties of Composite Surfaces: Alloys, Compounds, Semiconductors; Wandelt, K., Ed., Wiley-VCH Verlag GmbH & Co. KGaA, 2013; pp 229–278. Copéret, C.; Comas-Vives, A.; Conley, M. P.; Estes, D. P.; Fedorov, A.; Mougel, V.; Nagae, H.; Núñez-Zarur, F.; Zhizhko, P. A. Chem. Rev. 2016, 116 (2), 323–421. Candy, J.-P.; Copéret, C.; Basset, J.-M. In Surface and Interfacial Organometallic Chemistry and Catalysis. Topic in Organometallic Chemistry; Copéret, C., Chaudret, B., Eds., Springer: Berlin Heidelberg, 2005; pp 151–210. Samantaray, M. K.; D’Elia, V.; Pump, E.; Falivene, L.; Harb, M.; Ould Chikh, S.; Cavallo, L.; Basset, J.-M. Chem. Rev. 2020, 120 (2), 734–813. Green, M. L. H. J. Organomet. Chem. 1995, 500 (1–2), 127–148. Rimola, A.; Costa, D.; Sodupe, M.; Lambert, J.-F.; Ugliengo, P. Chem. Rev. 2013, 113 (6), 4216–4313. Rascón, F.; Wischert, R.; Copéret, C. Chem. Sci. 2011, 2 (8), 1449–1456. Flörke, O. W.; Graetsch, H. A.; Brunk, F.; Benda, L.; Paschen, S.; Bergna, H. E.; Roberts, W. O.; Welsh, W. A.; Libanati, C.; Ettlinger, M.; Kerner, D.; Maier, M.; Meon, W.; Schmoll, R.; Gies, H.; Schiffmann, D. Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2008. Zhuravlev, L. T. Colloids Surf. A 2000, 173, 1–38. Zhuravlev, L. T. Langmuir 1987, 3 (3), 316–318. Bordiga, S.; Roggero, I.; Ugliengo, P.; Zecchina, A.; Bolis, V.; Artioli, G.; Buzzoni, R.; Marra, G.; Rivetti, F.; Spanò, G.; Lamberti, C. J. Chem. Soc., Dalton Trans. 2000, 21, 3921–3929. Hoffmann, P.; Knözinger, E. Surf. Sci. 1987, 188 (1–2), 181–198. Lefort, L.; Chabanas, M.; Maury, O.; Meunier, D.; Copéret, C.; Thivolle-Cazat, J.; Basset, J.-M. J. Organomet. Chem. 2000, 593–594, 96–100. Amor Nait Ajjou, J.; Scott, S. L. Organometallics 1997, 16 (1), 86–92. Tosin, G.; Santini, C. C.; Taoufik, M.; De Mallmann, A.; Basset, J.-M. Organometallics 2006, 25 (14), 3324–3335. Dufaud, V.; Niccolai, G. P.; Thivolle-Cazat, J.; Basset, J.-M. J. Am. Chem. Soc. 1995, 117 (15), 4288–4294. Le Roux, E.; Chabanas, M.; Baudouin, A.; De Mallmann, A.; Copéret, C.; Quadrelli, E. A.; Thivolle-Cazat, J.; Basset, J.-M.; Lukens, W.; Lesage, A.; Emsley, L.; Sunley, G. J. J. Am. Chem. Soc. 2004, 126 (41), 13391–13399. Chabanas, M.; Baudouin, A.; Copéret, C.; Basset, J.-M.; Lukens, W.; Lesage, A.; Hediger, S.; Emsley, L. J. Am. Chem. Soc. 2003, 125 (2), 492–504. Conley, M. P.; Rossini, A. J.; Comas-Vives, A.; Valla, M.; Casano, G.; Ouari, O.; Tordo, P.; Lesage, A.; Emsley, L.; Copéret, C. Phys. Chem. Chem. Phys. 2014, 16 (33), 17822–17827. Chiang, C.-M.; Zegarski, B. R.; Dubois, L. H. J. Phys. Chem. 1993, 97 (27), 6948–6950. Sot, P.; Newton, M. A.; Baabe, D.; Walter, M. D.; van Bavel, A. P.; Horton, A. D.; Copéret, C.; van Bokhoven, J. A. Chem. Eur. J. 2020, 26 (36), 8012–8016. Scott, S. L.; Basset, J.-M. J. Am. Chem. Soc. 1994, 116 (26), 12069–12070. Demmelmaier, C. A.; White, R. E.; van Bokhoven, J. A.; Scott, S. L. J. Phys. Chem. C 2008, 112 (16), 6439–6449. Fleischman, S. D.; Scott, S. L. J. Am. Chem. Soc. 2011, 133 (13), 4847–4855. Mania, P.; Conrad, S.; Verel, R.; Hammond, C.; Hermans, I. Dalton Trans. 2013, 42 (35), 12725–12732. Maity, N.; Barman, S.; Callens, E.; Samantaray, M. K.; Abou-Hamad, E.; Minenkov, Y.; D’Elia, V.; Hoffman, A. S.; Widdifield, C. M.; Cavallo, L.; Gates, B. C.; Basset, J.-M. Chem. Sci. 2016, 7 (2), 1558–1568. Love, A. M.; Carrero, C. A.; Chieregato, A.; Grant, J. T.; Conrad, S.; Verel, R.; Hermans, I. Chem. Mater. 2016, 28 (15), 5495–5504. Barman, S.; Maity, N.; Bhatte, K.; Ould-Chikh, S.; Dachwald, O.; Haeßner, C.; Saih, Y.; Abou-Hamad, E.; Llorens, I.; Hazemann, J.-L.; Köhler, K.; D’Elia, V.; Basset, J.-M. ACS Catal. 2016, 6 (9), 5908–5921. Högerl, M. P.; Serena Goh, L. M.; Abou-Hamad, E.; Barman, S.; Dachwald, O.; Pasha, F. A.; Pelletier, J.; Köhler, K.; D’Elia, V.; Cavallo, L.; Basset, J.-M. RSC Adv. 2018, 8 (37), 20801–20808. Park, J.-W.; Park, Y. J.; Jun, C.-H. Chem. Commun. 2011, 47 (17), 4860–4871. Hara, K.; Akahane, S.; Wiench, J. W.; Burgin, B. R.; Ishito, N.; Lin, V. S.-Y.; Fukuoka, A.; Pruski, M. J. Phys. Chem. C 2012, 116 (12), 7083–7090. Hudson, L. K.; Misra, C.; Perrotta, A. J.; Wefers, K.; Williams, F. S. Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2000. Euzen, P.; Raybaud, P.; Krokidis, X.; Toulhoat, H.; Le Loarer, J.-L.; Jolivet, J.-P.; Froidefond, C. In Handbook of Porous Solids; Schüth, F., Sing, K. S. W., Weitkamp, J., Eds., Wiley-VCH Verlag GmbH: Weinheim, Germany, 2002; pp 1591–1677. Amrute, A. P.; Łodziana, Z.; Schreyer, H.; Weidenthaler, C.; Schüth, F. Science 2019, 366 (6464), 485–489. Busca, G. Catal. Today 2014, 226, 2–13. Wolverton, C.; Hass, K. C. Phys. Rev. B 2001, 63 (2), 024102. Prins, R. Angew. Chem. Int. Ed. 2019, 58 (43), 15548–15552. Krokidis, X.; Raybaud, P.; Gobichon, A.-E.; Rebours, B.; Euzen, P.; Toulhoat, H. J. Phys. Chem. B 2001, 105 (22), 5121–5130. Batista, A. T. F.; Wisser, D.; Pigeon, T.; Gajan, D.; Diehl, F.; Rivallan, M.; Catita, L.; Gay, A.-S.; Lesage, A.; Chizallet, C.; Raybaud, P. J. Catal. 2019, 378, 140–143. Nortier, P.; Fourre, P.; Saad, A. B. M.; Saur, O.; Lavalley, J. C. Appl. Catal. 1990, 61 (1), 141–160. Digne, M.; Sautet, P.; Raybaud, P.; Euzen, P.; Toulhoat, H. J. Catal. 2004, 226 (1), 54–68. Wischert, R.; Copéret, C.; Delbecq, F.; Sautet, P. Chem. Commun. 2011, 47 (17), 4890–4892. Wischert, R.; Copéret, C.; Delbecq, F.; Sautet, P. Angew. Chem. Int. Ed. 2011, 50 (14), 3202–3205. Digne, M.; Sautet, P.; Raybaud, P.; Euzen, P.; Toulhoat, H. J. Catal. 2002, 211 (1), 1–5.

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

83

68. Wischert, R.; Laurent, P.; Copéret, C.; Delbecq, F.; Sautet, P. J. Am. Chem. Soc. 2012, 134 (35), 14430–14449. 69. Joubert, J.; Delbecq, F.; Sautet, P.; Le Roux, E.; Taoufik, M.; Thieuleux, C.; Blanc, F.; Copéret, C.; Thivolle-Cazat, J.; Basset, J.-M. J. Am. Chem. Soc. 2006, 128 (28), 9157–9169. 70. Guzman, J.; Gates, B. C. Langmuir 2003, 19 (9), 3897–3903. 71. Delgado, M.; Santini, C. C.; Delbecq, F.; Wischert, R.; Le Guennic, B.; Tosin, G.; Spitz, R.; Basset, J.-M.; Sautet, P. J. Phys. Chem. C 2010, 114 (43), 18516–18528. 72. Ahn, H.; Marks, T. J. J. Am. Chem. Soc. 2002, 124 (24), 7103–7110. 73. Joubert, J.; Salameh, A.; Krakoviack, V.; Delbecq, F.; Sautet, P.; Copéret, C.; Basset, J.-M. J. Phys. Chem. B 2006, 110 (47), 23944–23950. 74. Salameh, A.; Joubert, J.; Baudouin, A.; Lukens, W.; Delbecq, F.; Sautet, P.; Basset, J. M.; Copéret, C. Angew. Chem. Int. Ed. 2007, 46 (21), 3870–3873. 75. Parker, S. C.; de Leeuw, N. H.; Redfern, S. E. Faraday Discuss. 1999, 114, 381–393. 76. Giordano, L.; Goniakowski, J.; Suzanne, J. Phys. Rev. Lett. 1998, 81 (6), 1271–1273. 77. Odelius, M. Phys. Rev. Lett. 1999, 82 (19), 3919–3922. 78. Kim, Y. D.; Stultz, J.; Goodman, D. W. J. Phys. Chem. B 2002, 106 (7), 1515–1517. 79. Kim, Y. D.; Lynden-Bell, R. M.; Alavi, A.; Stulz, J.; Goodman, D. W. Chem. Phys. Lett. 2002, 352 (5–6), 318–322. 80. Che, M.; Tench, A. J. In Advances in Catalysis, vol. 31, Eley, D. D., Pines, H., Weisz, P. B., Eds.; vol. 31; Academic Press, 1982; pp 77–133. 81. Chizallet, C.; Costentin, G.; Che, M.; Delbecq, F.; Sautet, P. J. Phys. Chem. B 2006, 110 (32), 15878–15886. 82. Chizallet, C.; Costentin, G.; Che, M.; Delbecq, F.; Sautet, P. J. Am. Chem. Soc. 2007, 129 (20), 6442–6452. 83. Bailly, M.-L.; Chizallet, C.; Costentin, G.; Krafft, J.-M.; Lauron-Pernot, H.; Che, M. J. Catal. 2005, 235 (2), 413–422. 84. Chizallet, C.; Costentin, G.; Lauron-Pernot, H.; Krafft, J. M.; Bazin, P.; Saussey, J.; Delbecq, F.; Sautet, P.; Che, M. Oil Gas Sci. Technol. - Rev. l’IFP 2006, 61 (4), 479–488. 85. Witzke, R. J.; Chapovetsky, A.; Conley, M. P.; Kaphan, D. M.; Delferro, M. ACS Catal. 2020, 11822–11840. 86. Taguchi, A.; Schüth, F. Microporous Mesoporous Mater. 2005, 77 (1), 1–45. 87. Anwander, R. Chem. Mater. 2001, 13 (12), 4419–4438. 88. Liang, Y.; Anwander, R. Dalton Trans. 2006, 60 (15), 1909–1918. 89. Liang, Y.; Erichsen, E. S.; Song, C.; Anwander, R. Eur. J. Inorg. Chem. 2012, 5741–5751. 90. Croissant, J. G.; Cattoën, X.; Wong Chi Man, M.; Durand, J.-O.; Khashab, N. M. Nanoscale 2015, 7 (48), 20318–20334. 91. Park, S. S.; Santha Moorthy, M.; Ha, C.-S. NPG Asia Mater. 2014, 6 (4), e96. 92. Flanigen, E. M.; Broach, R. W.; Wilson, S. T. In Zeolites in Industrial Separation and Catalysis; Kulprathipanja, S., Ed., Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2010; pp 1–26. 93. Yilmaz, B.; Müller, U. Top. Catal. 2009, 52 (6–7), 888–895. 94. Kosinov, N.; Liu, C.; Hensen, E. J. M.; Pidko, E. A. Chem. Mater. 2018, 30 (10), 3177–3198. 95. Li, G.; Pidko, E. A. ChemCatChem 2019, 11 (1), 134–156. 96. Derouane, E. G.; Védrine, J. C.; Ramos Pinto, R.; Borges, P. M.; Costa, L.; Lemos, M. A. N. D. A.; Lemos, F.; Ramôa Ribeiro, F. Catal. Rev. Sci. Eng. 2013, 55 (4), 454–515. 97. Serna, P.; Gates, B. C. Acc. Chem. Res. 2014, 47 (8), 2612–2620. 98. Szeto, K. C.; Gallo, A.; Hernández-Morejudo, S.; Olsbye, U.; De Mallmann, A.; Lefebvre, F.; Gauvin, R. M.; Delevoye, L.; Scott, S. L.; Taoufik, M. J. Phys. Chem. C 2015, 119 (47), 26611–26619. 99. Nédez, C.; Theolier, A.; Lefebvre, F.; Choplin, A.; Basset, J. M.; Joly, J. F.; Benazzi, E. Microporous Mater. 1994, 2 (4), 251–259. 100. Perego, C.; Millini, R. Chem. Soc. Rev. 2013, 42 (9), 3956–3976. 101. Pérez-Ramírez, J.; Christensen, C. H.; Egeblad, K.; Christensen, C. H.; Groen, J. C. Chem. Soc. Rev. 2008, 37 (11), 2530–2542. 102. Roth, W. J.; Nachtigall, P.; Morris, R. E.; Cejka, J. Chem. Rev. 2014, 114 (9), 4807–4837. 103. Furukawa, H.; Cordova, K. E.; O’Keeffe, M.; Yaghi, O. M. Science 2013, 341, 1230444. 104. Kaskel, S. In The Chemistry of Metal–Organic Frameworks; Kaskel, S., Ed., Wiley-VCH Verlag GmbH & Co: KGaA, 2016; pp 1–3. 105. Farha, O. K.; Eryazici, I.; Jeong, N. C.; Hauser, B. G.; Wilmer, C. E.; Sarjeant, A. A.; Snurr, R. Q.; Nguyen, S. T.; Yazaydin, A.Ö.; Hupp, J. T. J. Am. Chem. Soc. 2012, 134 (36), 15016–15021. 106. Hönicke, I. M.; Senkovska, I.; Bon, V.; Baburin, I. A.; Bönisch, N.; Raschke, S.; Evans, J. D.; Kaskel, S. Angew. Chem. Int. Ed. 2018, 57 (42), 13780–13783. 107. Tanabe, K. K.; Cohen, S. M. Chem. Soc. Rev. 2011, 40 (2), 498–519. 108. Marshall, R. J.; Forgan, R. S. Eur. J. Inorg. Chem. 2016, 4310–4331. 109. Yin, Z.; Wan, S.; Yang, J.; Kurmoo, M.; Zeng, M.-H. Coord. Chem. Rev. 2019, 378, 500–512. 110. Bai, Y.; Dou, Y.; Xie, L.-H.; Rutledge, W.; Li, J.-R.; Zhou, H.-C. Chem. Soc. Rev. 2016, 45 (8), 2327–2367. 111. Rimoldi, M.; Howarth, A. J.; DeStefano, M. R.; Lin, L.; Goswami, S.; Li, P.; Hupp, J. T.; Farha, O. K. ACS Catal. 2017, 7 (2), 997–1014. 112. Cavka, J. H.; Jakobsen, S.; Olsbye, U.; Guillou, N.; Lamberti, C.; Bordiga, S.; Lillerud, K. P. J. Am. Chem. Soc. 2008, 130 (42), 13850–13851. 113. Winarta, J.; Shan, B.; McIntyre, S. M.; Ye, L.; Wang, C.; Liu, J.; Mu, B. Cryst. Growth Des. 2020, 20 (2), 1347–1362. 114. Larabi, C.; Quadrelli, E. A. Eur. J. Inorg. Chem. 2012, 3 (18), 3014–3022. 115. Manna, K.; Ji, P.; Greene, F. X.; Lin, W. J. Am. Chem. Soc. 2016, 138 (24), 7488–7491. 116. Wang, T. C.; Vermeulen, N. A.; Kim, I. S.; Martinson, A. B. F.; Stoddart, J. F.; Hupp, J. T.; Farha, O. K. Nat. Protoc. 2016, 11 (1), 149–162. 117. Mondloch, J. E.; Katz, M. J.; Isley, W. C.; Ghosh, P.; Liao, P.; Bury, W.; Wagner, G. W.; Hall, M. G.; Decoste, J. B.; Peterson, G. W.; Snurr, R. Q.; Cramer, C. J.; Hupp, J. T.; Farha, O. K. Nat. Mater. 2015, 14 (5), 512–516. 118. Mondloch, J. E.; Bury, W.; Fairen-Jimenez, D.; Kwon, S.; Demarco, E. J.; Weston, M. H.; Sarjeant, A. A.; Nguyen, S. T.; Stair, P. C.; Snurr, R. Q.; Farha, O. K.; Hupp, J. T. J. Am. Chem. Soc. 2013, 135 (28), 10294–10297. 119. Yang, D.; Odoh, S. O.; Wang, T. C.; Farha, O. K.; Hupp, J. T.; Cramer, C. J.; Gagliardi, L.; Gates, B. C. J. Am. Chem. Soc. 2015, 137 (23), 7391–7396. 120. Klet, R. C.; Tussupbayev, S.; Borycz, J.; Gallagher, J. R.; Stalzer, M. M.; Miller, J. T.; Gagliardi, L.; Hupp, J. T.; Marks, T. J.; Cramer, C. J.; Delferro, M.; Farha, O. K. J. Am. Chem. Soc. 2015, 137 (50), 15680–15683. 121. Rimoldi, M.; Hupp, J. T.; Farha, O. K. ACS Appl. Mater. Interfaces 2017, 9 (40), 35067–35074. 122. Korzynski, M. D.; Consoli, D. F.; Zhang, S.; Román-Leshkov, Y.; Dinca, M. J. Am. Chem. Soc. 2018, 140 (22), 6956–6960. 123. Desai, S. P.; Ye, J.; Zheng, J.; Ferrandon, M. S.; Webber, T. E.; Platero-Prats, A. E.; Duan, J.; Garcia-Holley, P.; Camaioni, D. M.; Chapman, K. W.; Delferro, M.; Farha, O. K.; Fulton, J. L.; Gagliardi, L.; Lercher, J. A.; Penn, R. L.; Stein, A.; Lu, C. C. J. Am. Chem. Soc. 2018, 140 (45), 15309–15318. 124. Thiam, Z.; Abou-Hamad, E.; Dereli, B.; Liu, L.; Emwas, A.-H.; Ahmad, R.; Jiang, H.; Isah, A. A.; Ndiaye, P. B.; Taoufik, M.; Han, Y.; Cavallo, L.; Basset, J.-M.; Eddaoudi, M. J. Am. Chem. Soc. 2020, 142 (39), 16690–16703. 125. Crabtree, R. H. The Organometallic Chemistry of the Transition Metals, John Wiley & Sons, Inc., 2014; pp 69–97. 126. Crabtree, R. H.; Pi-Complexe. The Organometallic Chemistry of the Transition Metals, John Wiley & Sons, Inc., 2014; pp 134–162. 127. Lappert, M.; Protchenko, A.; Power, P.; Seeber, A. Metal Amide Chemistry, John Wiley & Sons, Ltd., 2008; pp 149–204. 128. Oakton, E.; Vilé, G.; Levine, D. S.; Zocher, E.; Baudouin, D.; Pérez-Ramírez, J.; Copéret, C. Dalton Trans. 2014, 43 (40), 15138–15142. 129. Kaeffer, N.; Liu, H.-J.; Lo, H.-K.; Fedorov, A.; Copéret, C. Chem. Sci. 2018, 9 (24), 5366–5371. 130. Turova, N. Y.; Turevskaya, E. P.; Kessler, V. G.; Yanovskaya, M. I. The Chemistry of Metal Alkoxides, Springer: Boston, 2002; pp 1–10. 131. Krempner, C. Eur. J. Inorg. Chem. 2011, 1689–1698.

84 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197.

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts Vigato, P. A.; Peruzzo, V.; Tamburini, S. Coord. Chem. Rev. 2009, 253 (7–8), 1099–1201. Bhirud, V. A.; Ehresmann, J. O.; Kletnieks, P. W.; Haw, J. F.; Gates, B. C. Langmuir 2006, 22 (1), 490–496. Mania, P.; Verel, R.; Jenny, F.; Hammond, C.; Hermans, I. Chem. A Eur. J. 2013, 19 (30), 9849–9858. D’Elia, V.; Dong, H.; Rossini, A. J.; Widdifield, C. M.; Vummaleti, S. V. C.; Minenkov, Y.; Poater, A.; Abou-Hamad, E.; Pelletier, J. D. A.; Cavallo, L.; Emsley, L.; Basset, J.-M. J. Am. Chem. Soc. 2015, 137 (24), 7728–7739. Hamieh, A.; Dey, R.; Samantaray, M. K.; Abdel-Azeim, S.; Abou-Hamad, E.; Chen, Y.; Pelletier, J. D. A.; Cavallo, L.; Basset, J.-M. Organometallics 2016, 35 (15), 2524–2531. Deguns, E. W.; Taha, Z.; Meitzner, G. D.; Scott, S. L. J. Phys. Chem. B 2005, 109 (11), 5005–5011. Zhu, H.; Ould-Chikh, S.; Dong, H.; Llorens, I.; Saih, Y.; Anjum, D. H.; Hazemann, J. L.; Basset, J.-M. ChemCatChem 2015, 7 (20), 3332–3339. Rimoldi, M.; Fodor, D.; van Bokhoven, J. A.; Mezzetti, A. Cat. Sci. Technol. 2015, 5 (9), 4575–4586. Pelletier, J. D. A.; Basset, J.-M. Acc. Chem. Res. 2016, 49 (4), 664–677. Amakawa, K.; Wrabetz, S.; Kröhnert, J.; Tzolova-Müller, G.; Schlögl, R.; Trunschke, A. J. Am. Chem. Soc. 2012, 134 (28), 11462–11473. Copéret, C.; Estes, D. P.; Larmier, K.; Searles, K. Chem. Rev. 2016, 116 (15), 8463–8505. Mertis, K.; Galyer, L.; Wilkinson, G. J. Organomet. Chem. 1975, 97 (3), C65. Delley, M. F.; Lapadula, G.; Núñez-Zarur, F.; Comas-Vives, A.; Kalendra, V.; Jeschke, G.; Baabe, D.; Walter, M. D.; Rossini, A. J.; Lesage, A.; Emsley, L.; Maury, O.; Copéret, C. J. Am. Chem. Soc. 2017, 139 (26), 8855–8867. Fujdala, K. L.; Brutchey, R. L.; Tilley, T. D. In Surface and Interfacial Organometallic Chemistry and Catalysis. Topics in Organometallic Chemistry; Copéret, C., Chaudret, B., Eds.; vol. 16; Springer-Verlag: Berlin Heidelberg, 2005; pp 69–115. Copéret, C. Acc. Chem. Res. 2019, 52 (6), 1697–1708. Guan, E.; Gates, B. C. ACS Catal. 2018, 8 (1), 482–487. Vansant, E. F., Van Der Voort, P., Vrancken, K. C., Eds.; Studies in Surface Science and Catalysis; vol. 93; Elsevier, 1995; pp 79–91. Lamberti, C.; Zecchina, A.; Groppo, E.; Bordiga, S. Chem. Soc. Rev. 2010, 39 (12), 4951–5001. Zaera, F. Chem. Soc. Rev. 2014, 43 (22), 7624–7663. Bañares, M. A.; Wachs, I. E. In Encyclopedia of Analytical Chemistry; Meyers, R. A., Ed., John Wiley & Sons, Ltd, 2010. Hartman, T.; Wondergem, C. S.; Kumar, N.; van den Berg, A.; Weckhuysen, B. M. J. Phys. Chem. Lett. 2016, 7 (8), 1570–1584. Busca, G. In Metal Oxide Catalysis; Jackson, S. D., Hargreaves, J. S. J., Eds., Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2008; pp 95–175. Van Doorslaer, S.; Murphy, D. M. In EPR Spectroscopy. Topics in Current Chemistry; Drescher, M., Jeschke, G., Eds.; vol. 321; Springer: Berlin Heidelberg, 2011; pp 1–39. Venezia, A. M. Catal. Today 2003, 77 (4), 359–370. Nguyen, L.; Tao, F. F.; Tang, Y.; Dou, J.; Bao, X.-J. Chem. Rev. 2019, 119 (12), 6822–6905. Schoonheydt, R. A. Chem. Soc. Rev. 2010, 39 (12), 5051–5066. Roe, D. C.; Kating, P. M.; Krusic, P. J.; Smart, B. E. Top. Catal. 1998, 5 (1–4), 133–147. Ashbrook, S. E.; Griffin, J. M.; Johnston, K. E. Annu. Rev. Anal. Chem. 2018, 11 (1), 485–508. Kobayashi, T.; Perras, F. A.; Slowing, I. I.; Sadow, A. D.; Pruski, M. ACS Catal. 2015, 5 (12), 7055–7062. Copéret, C.; Liao, W.-C.; Gordon, C. P.; Ong, T.-C. J. Am. Chem. Soc. 2017, 139 (31), 10588–10596. Li, J.; Sun, J. Acc. Chem. Res. 2017, 50 (11), 2737–2745. Schlögl, R. In Advances in Catalysis, vol. 52, Gates, B. C., Knözinger, H., Eds.; vol. 52; Academic Press, 2009; pp 273–338. Garino, C.; Borfecchia, E.; Gobetto, R.; van Bokhoven, J. A.; Lamberti, C. Coord. Chem. Rev. 2014, 277–278, 130–186. Moulijn, J. A., van Leeuwen, P. W. N. M., van Santen, R. A., Eds.; Studies in Surface Science and Catalysis; vol. 79; Elsevier, 1993; pp 419–438. Su, D. S.; Zhang, B.; Schlögl, R. Chem. Rev. 2015, 115 (8), 2818–2882. Thomas, J. M. Proc. R. Soc. A 2017, 473, 20160714. Grogger, W.; Hofer, F.; Kothleitner, G.; Schaffer, B. Top. Catal. 2008, 50 (1–4), 200–207. van Santen, R. A., Sautet, P., Eds.; Computational Methods in Catalysis and Materials Science: An Introduction for Scientists and Engineers, Wiley-VCH Verlag GmbH & Co. KGaA, 2009. Sautet, P.; Delbecq, F. Chem. Rev. 2010, 110 (3), 1788–1806. Chizallet, C.; Raybaud, P. Cat. Sci. Technol. 2014, 4 (9), 2797–2813. Grajciar, L.; Heard, C. J.; Bondarenko, A. A.; Polynski, M. V.; Meeprasert, J.; Pidko, E. A.; Nachtigall, P. Chem. Soc. Rev. 2018, 47 (22), 8307–8348. Hlatky, G. G. Chem. Rev. 2000, 100 (4), 1347–1376. Ballard, D. G. H. In Advances in Catalysis, vol. 23, Eley, D. D., Pines, H., Weisz, P. B., Eds.; vol. 23; Academic Press, 1973; pp 263–325. Zakharov, V. A.; Dudchenko, V. K.; Paukshtis, E. A.; Karakchiev, L. G.; Yermakov, Y. I. J. Mol. Catal. 1977, 2 (6), 421–435. Zakharov, V. A.; Yermakov, Y. I. Catal. Rev. Sci. Eng. 1979, 19 (1), 67–103. Tosin, G.; Santini, C. C.; Basset, J.-M. Top. Catal. 2009, 1203–1210. Millot, N.; Soignier, S.; Santini, C. C.; Baudouin, A.; Basset, J. M. J. Am. Chem. Soc. 2006, 128 (29), 9361–9370. Yang, X.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1991, 113 (9), 3623–3625. Stalzer, M. M.; Delferro, M.; Marks, T. J. Catal. Lett. 2015, 145 (1), 3–14. Dahmen, K.-H.; Hedden, D.; Burwell, R. L.; Marks, T. J. Langmuir 1988, 4 (5), 1212–1214. Motta, A.; Fragalà, I. L.; Marks, T. J. J. Am. Chem. Soc. 2008, 130 (49), 16533–16546. Tafazolian, H.; Culver, D. B.; Conley, M. P. Organometallics 2017, 36 (13), 2385–2388. Culver, D. B.; Venkatesh, A.; Huynh, W.; Rossini, A. J.; Conley, M. P. Chem. Sci. 2020, 11 (6), 1510–1517. Nicholas, C. P.; Ahn, H.; Marks, T. J. J. Am. Chem. Soc. 2003, 125 (14), 4325–4331. Mol, J. C. J. Mol. Catal. A Chem. 2004, 213 (1), 39–45. Astruc, D. In Olefin Metathesis: Theory and Practice; Grela, K., Ed., John Wiley & Sons, Inc., 2014; pp 1–36. Hérisson, J.-L.; Chauvin, Y. Die Makromol. Chemie 1970, 141 (3487), 161–176. Lwin, S.; Wachs, I. E. ACS Catal. 2014, 4 (8), 2505–2520. Schrock, R. R. Tetrahedron 1999, 55 (27), 8141–8153. Herrmann, W. A.; Stumpf, A. W.; Priermeier, T.; Bogdanovic, S.; Dufaud, V.; Basset, J.-M. Angew. Chem. Int. Ed. Engl. 1996, 35 (23/24), 2803–2805. Blanc, F.; Chabanas, M.; Copéret, C.; Fenet, B.; Herdweck, E. J. Organomet. Chem. 2005, 690 (23), 5014–5026. Blanc, F.; Copéret, C.; Thivolle-Cazat, J.; Basset, J.-M.; Lesage, A.; Emsley, L.; Sinha, A.; Schrock, R. R. Angew. Chem. Int. Ed. 2006, 45 (8), 1216–1220. Rhers, B.; Salameh, A.; Baudouin, A.; Quadrelli, E. A.; Taoufik, M.; Copéret, C.; Lefebvre, F.; Basset, J.-M.; Solans-Monfort, X.; Eisenstein, O.; Lukens, W. W.; Lopez, L. P. H.; Sinha, A.; Schrock, R. R. Organometallics 2006, 25 (15), 3554–3557. Blanc, F.; Thivolle-Cazat, J.; Basset, J.-M.; Copéret, C.; Hock, A. S.; Tonzetich, Z. J.; Schrock, R. R. J. Am. Chem. Soc. 2007, 129 (5), 1044–1045. Rendón, N.; Berthoud, R.; Blanc, F.; Gajan, D.; Maishal, T.; Basset, J.-M.; Copéret, C.; Lesage, A.; Emsley, L.; Marinescu, S. C.; Singh, R.; Schrock, R. R. Chem. A Eur. J. 2009, 15 (20), 5083–5089. Mougel, V.; Copéret, C. Chem. Sci. 2014, 5 (6), 2475–2481.

Surface organometallic and coordination chemistry approach to formation of single site heterogeneous catalysts

85

198. Conley, M. P.; Mougel, V.; Peryshkov, D. V.; Forrest, W. P.; Gajan, D.; Lesage, A.; Emsley, L.; Copéret, C.; Schrock, R. R. J. Am. Chem. Soc. 2013, 135 (51), 19068–19070. 199. Pucino, M.; Zhai, F.; Gordon, C. P.; Mance, D.; Hoveyda, A. H.; Schrock, R. R.; Copéret, C. Angew. Chem. Int. Ed. 2019, 58 (34), 11816–11819. 200. Pucino, M.; Inoue, M.; Gordon, C. P.; Schowner, R.; Stöhr, L.; Sen, S.; Hegedüs, C.; Robé, E.; Tóth, F.; Buchmeiser, M. R.; Copéret, C. Angew. Chem. Int. Ed. 2018, 57 (44), 14566–14569. 201. De Jesus Silva, J.; Mance, D.; Pucino, M.; Benedikter, M. J.; Elser, I.; Buchmeiser, M. R.; Copéret, C. Helv. Chim. Acta 2020, 103, e2000161. 202. Pucino, M.; Mougel, V.; Schowner, R.; Fedorov, A.; Buchmeiser, M. R.; Copéret, C. Angew. Chem. Int. Ed. 2016, 55 (13), 4300–4302. 203. Sheldon, R. A. J. Mol. Catal. 1980, 7 (1), 107–126. 204. Sheldon, R. A.; Wallau, M.; Arends, I. W. C. E.; Schuchardt, U. Acc. Chem. Res. 1998, 31 (8), 485–493. 205. Adolfsson, H. In Modern Oxidation Methods; Bäckvall, J.-E., Ed., Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, 2010; pp 37–84. 206. Maschmeyer, T.; Rey, F.; Sankar, G.; Thomas, J. M. Nature 1995, 378 (6553), 159–162. 207. Coles, M. P.; Lugmair, C. G.; Terry, K. W.; Tilley, T. D. Chem. Mater. 2000, 12 (1), 122–131. 208. Jarupatrakorn, J.; Tilley, T. D. J. Am. Chem. Soc. 2002, 124 (28), 8380–8388. 209. Brutchey, R. L.; Ruddy, D. A.; Andersen, L. K.; Tilley, T. D. Langmuir 2005, 21 (21), 9576–9583. 210. Ruddy, D. A.; Brutchey, R. L.; Tilley, T. D. Top. Catal. 2008, 48 (1–4), 99–106. 211. Notestein, J. M.; Iglesia, E.; Katz, A. J. Am. Chem. Soc. 2004, 126 (50), 16478–16486. 212. Notestein, J. M.; Solovyov, A.; Andrini, L. R.; Requejo, F. G.; Katz, A.; Iglesia, E. J. Am. Chem. Soc. 2007, 129 (50), 15585–15595. 213. Sattler, J. J. H. B.; Ruiz-Martinez, J.; Santillan-Jimenez, E.; Weckhuysen, B. M. Chem. Rev. 2014, 114 (20), 10613–10653. 214. Korzynski, M. D.; Dinca, M. ACS Cent. Sci. 2017, 3 (1), 10–12. 215. Szeto, K. C.; Loges, B.; Merle, N.; Popoff, N.; Quadrelli, A.; Jia, H.; Berrier, E.; De Mallmann, A.; Delevoye, L.; Gauvin, R. M.; Taoufik, M. Organometallics 2013, 32 (21), 6452–6460. 216. Kaphan, D. M.; Ferrandon, M. S.; Langeslay, R. R.; Celik, G.; Wegener, E. C.; Liu, C.; Niklas, J.; Poluektov, O. G.; Delferro, M. ACS Catal. 2019, 9 (12), 11055–11066. 217. Conley, M. P.; Delley, M. F.; Núnez-Zarur, F.; Comas-Vives, A.; Copéret, C. Inorg. Chem. 2015, 54 (11), 5065–5078. 218. Delley, M. F.; Silaghi, M. C.; Nuñez-Zarur, F.; Kovtunov, K. V.; Salnikov, O. G.; Estes, D. P.; Koptyug, I. V.; Comas-Vives, A.; Copéret, C. Organometallics 2017, 36 (1), 234–244. 219. Hu, B.; Schweitzer, N. M.; Zhang, G.; Kraft, S. J.; Childers, D. J.; Lanci, M. P.; Miller, J. T.; Hock, A. S. ACS Catal. 2015, 5 (6), 3494–3503. 220. Estes, D. P.; Siddiqi, G.; Allouche, F.; Kovtunov, K. V.; Safonova, O. V.; Trigub, A. L.; Koptyug, I. V.; Copéret, C. J. Am. Chem. Soc. 2016, 138 (45), 14987–14997. 221. Getsoian, A.; Das, U.; Camacho-Bunquin, J.; Zhang, G.; Gallagher, J. R.; Hu, B.; Cheah, S.; Schaidle, J. A.; Ruddy, D. A.; Hensley, J. E.; Krause, T. R.; Curtiss, L. A.; Miller, J. T.; Hock, A. S. Cat. Sci. Technol. 2016, 6 (16), 6339–6353. 222. Searles, K.; Siddiqi, G.; Safonova, O. V.; Copéret, C. Chem. Sci. 2017, 8 (4), 2661–2666. 223. Szeto, K. C.; Jones, Z. R.; Merle, N.; Rios, C.; Gallo, A.; Le Quemener, F.; Delevoye, L.; Gauvin, R. M.; Scott, S. L.; Taoufik, M. ACS Catal. 2018, 8 (8), 7566–7577. 224. Liu, C.; Camacho-Bunquin, J.; Ferrandon, M.; Savara, A.; Sohn, H.; Yang, D.; Kaphan, D. M.; Langeslay, R. R.; Ignacio-de Leon, P. A.; Liu, S.; Das, U.; Yang, B.; Hock, A. S.; Stair, P. C.; Curtiss, L. A.; Delferro, M. Polyhedron 2018, 152, 73–83. 225. Uzun, A.; Bhirud, V. A.; Kletnieks, P. W.; Haw, J. F.; Gates, B. C. J. Phys. Chem. C 2007, 111 (41), 15064–15073. 226. Lu, J.; Serna, P.; Gates, B. C. ACS Catal. 2011, 1 (11), 1549–1561. 227. Uzun, A.; Gates, B. C. Angew. Chem. Int. Ed. 2008, 47 (48), 9245–9248. 228. Uzun, A.; Gates, B. C. J. Am. Chem. Soc. 2009, 131 (43), 15887–15894. 229. Lu, J.; Serna, P.; Aydin, C.; Browning, N. D.; Gates, B. C. J. Am. Chem. Soc. 2011, 133 (40), 16186–16195. 230. Hoffman, A. S.; Debefve, L. M.; Zhang, S.; Perez-Aguilar, J. E.; Conley, E. T.; Justl, K. R.; Arslan, I.; Dixon, D. A.; Gates, B. C. ACS Catal. 2018, 8 (4), 3489–3498.

6.05 Challenges with atomically dispersed supported metal catalysts: Controlling performance, improving stability, and enhancing metal loading lu-O¨ztuluma,b and Alper Uzuna,b,c, a Department of Chemical and Biological Engineering, Koç University, Samira Fatma Kurtog b Istanbul, Turkey; Koç University TÜPRAS¸ Energy Center (KUTEM), Koç University, Istanbul, Turkey; and c Koç University Surface Science and Technology Center (KUYTAM), Koç University, Istanbul, Turkey © 2023 Elsevier Ltd. All rights reserved.

6.05.1 6.05.2 6.05.2.1 6.05.2.1.1 6.05.2.1.2 6.05.2.1.3 6.05.2.1.4 6.05.2.2 6.05.2.3 6.05.3 References

Introduction Challenges with atomically dispersed supported metal catalysts Limited ability to tune catalytic performance Effects of non-support ligands Effects of supports as ligands Effects of metal nuclearity Effects of promoters Limited ability to control stability Limited metal loadings Summary

86 87 87 87 89 94 98 101 106 108 109

Abstract Atomically dispersed supported metal catalysts offer opportunities of maximum utilization of expensive noble metals and provide unprecedented catalytic properties. Thus, these novel materials have received tremendous attention, especially in the last decade. Notwithstanding their advantages and such rapidly growing interest, these novel materials face various challenges, which need to be overcome to make them industrially viable. One of these challenges is the limited ability to control their catalytic properties. Changing the ligand environment, which also includes the support, varying the metal nuclearity, and the use of promoters (also including ionic liquid sheets) have shown to offer broad opportunities for tuning the catalytic performance. The other challenges are mostly related with the limited stability of the active species under reaction conditions, which also limits the metal loadings on the support surfaces. Control of electronic structure on the metal sites and the use of functional groups on support surfaces have been shown to be effective in this direction. In this chapter, some of the recent approaches aiming at overcoming these challenges related with the atomically dispersed supported metal catalysts are presented.

6.05.1

Introduction

Refining of crude oil to valuable products, conversion of renewable biomass to transportation fuels, treatment of exhaust gases, production of fertilizers, high-strength polymers, and medicines, and many other processes would not be possible without the existence of catalysts. These critical materials mediate the reaction pathways, enabling the formation of desired products at commercially viable rates.1 Most of the industrial catalysts include transition metals. However, transition metals are often very expensive; thus, they are being dispersed in the form of nanometer-sized particles on cheap high-surface-area supports. These dispersed moieties have dimensions of 1–50 nm and are often considered as nanoparticles. Catalytic performance of the active metal centers located on the surface of these supported metal nanoparticles depends on many factors, such as the size and shape of the nanoparticles carrying these active sites, location of the active sites on these nanoparticles, i.e., corners, edges, and faces, and the influence of supports on these sites.2,3 In a conventional supported metal catalyst, it is very challenging to track the effects of these individual structural factors on the catalytic properties. Because, the structure of these catalysts is very complex, having a highly non-uniform distribution of active sites with different catalytic properties. Therefore, understanding of the structure-performance relationships in such conventional supported metal catalysts requires fundamental level investigations on model systems. One of the pioneering approaches in this direction is to synthesize the supported metal catalysts as simple and as uniform as possible, just like their molecular analogs.4–6 The idea here is to controllably change the metal nuclearities and metal-support interactions while maintaining the atomic-level dispersion of the active metal centers. Combining the strengths of various characterization techniques and taking the advantage of the high degree of uniformity provide broad opportunities for understanding the behavior of these well-defined structures before, during, and after simple test reactions. These model catalysts are known as the supported molecular metal catalysts, exemplified by site-isolated mononuclear metal complexes7 and extremely small metal clusters of only a few metal atoms uniformly dispersed on supports.8 These novel catalysts have been very useful in investigating the structure-performance relationships at a fundamental level.

86

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00002-9

Challenges with atomically dispersed supported metal catalysts

87

Besides, the metal centers in these materials were found to offer quite interesting catalytic properties,9,10 triggering a tremendous interest in deeper research focusing on these novel materials. With the exciting advancements in the characterization techniques offering atomic level insights,11–13 the subject has now evolved into one of the hottest topics of the catalysis field in the last decade: Atomically dispersed supported metal catalysts. With some slight variations in the coordination environment of the metal on support surfaces, similar materials are also known as single-atom catalysts (SACs) and/or single-site metal catalysts and/or single-metal-atom catalysts and/or site-isolated metal catalysts and the list goes on.14,15 Throughout this chapter “atomically dispersed supported metal catalysts” will be used to account for all these types of novel materials. Atomically dispersed supported metal catalysts are drawing wide attention because they offer new catalytic properties16 along with an efficient use of expensive transition metals, at a metal dispersion of 100%. Notwithstanding the presence of numerous research articles, and even a significant number of recent review articles15–27 summarizing most of these reports, there is still only limited insight into how to control and enhance their catalytic properties, constituting the main challenges of the field. These challenges set the focus of this chapter. Some of the recent breakthrough examples have been selected to demonstrate the recent approaches in overcoming these challenges related with the atomically dispersed supported metal catalysts: (i) tuning their catalytic performance, (ii) improving their stability, and (iii) enhancing their metal loadings.

6.05.2

Challenges with atomically dispersed supported metal catalysts

6.05.2.1

Limited ability to tune catalytic performance

Tuning the catalytic performance of the active metal centers anchored onto supports requires a precise control over their electronic structure.27 In this regard, one approach is changing the metal’s ligand environment. The ligands can be the remnants of the metal precursor used to synthesize the catalyst or they can be the intermediates formed during the reaction or treatments.19 Besides these molecular-sized groups, the ligands can be macro-sized as well, such as supports. Thus, the local coordination environment of the metal centers and the support’s electron-donor properties play an important role in adjusting the electronic structure of the active metal centers.17 Other than changing the supports or the non-support ligands, the reactivity of atomically dispersed supported metal catalysts can also be altered by changing the nuclearity of the metal centers. A change in the metal nuclearity from mononuclear species to small metal clusters consisting of only a few metal atomsdthe active sites on these clusters can still be considered as atomically disperseddgenerally results in significant changes in the electronic structure of the active sites. Thus, precise control of the metal nuclearity also offers broad opportunities for adjusting the catalytic properties. Another approach in controlling the catalytic performance is the addition of promoter atoms or dopants which both change the electronic environment of the active sites by directly or indirectly interacting with the metal and stabilize the atomically dispersed metal centers on support surface.24 The following sections present some of the recent progress in the way of controlling the catalytic properties of atomically dispersed supported metal catalysts.

6.05.2.1.1

Effects of non-support ligands

Ligand environment of the active metal centers in an atomically dispersed supported metal catalyst can be changed by performing treatments under various gas environments. For instance, when atomically dispersed zeolite HY-, zeolite Hb-, zeolite HSSZ-53-, gAl2O3-, and MgO-supported iridium complexes having reactive ethylene ligands, supported Ir(C2H4)2 complexes, were reacted with CO at 298 K and 1 bar, ethylene ligands bonded to iridium were replaced with carbonyl ligands forming an iridium gem-dicarbonyl complex, evident from the formation of new infrared (IR) spectroscopy features representing the CO bands on each support.28–31 This rapid exchange of C2H4 ligands with CO was also observed for Rh(C2H4)2 supported on zeolites and MgO.32–34 When Ir(CO)2 complexes supported on different zeolites and metal oxides were treated in different combinations of CO, C2H4, H2, and He at 298 K and 1 bar, data demonstrated the various extend of ligand exchange reactions.28–30 While MgO-supported Ir(CO)2 complex was found to be stable under various gas environments, the complex supported on zeolite HY with an Si/Al ratio of 30 were found to form various different complexes under these gas environments including Ir(CO)3, Ir(CO)(C2H4), Ir(CO)2(C2H4), Ir(CO)(C2H4)2, and Ir(C2H5)2.29 For example, X-ray absorption fine structure (EXAFS) and IR data confirmed that treating zeolite HY-supported Ir(CO)2 complex under a mixture of C2H4 and CO (2:1 molar ratio) at 300 K results in the formation of Ir(CO)2(C2H4) or Ir(CO)(C2H4)2. Furthermore, treating the zeolite HY-supported Ir(C2H4)2 complex in flowing H2 at 300 K indicated that ethylene ligands on iridium are converted into s-bonded ethyl ligands.29 Fig. 1 demonstrates such changes in the ligand environment on the zeolite HY-supported iridium complex under different flow conditions. Catalytic performance measurements conducted on these catalysts consisting of different ligands showed that while the zeolite HY-supported Ir(C2H4)2 complex was active for ethylene hydrogenation and H2/D2 exchange, the Ir(CO)2 complex supported on the same zeolite did not provide any activity for ethylene hydrogenation and almost negligible activity for H2/D2 exchange (almost one order of magnitude lower compared to the case with the Ir(C2H4)2 complex), confirming that CO ligands are strong inhibitors for the interactions of hydrogen molecule with the metal center.29 As a general trend, the performance of the atomically dispersed complexes for the dissociation of H2 is reduced when the total number of ligands bonded to iridium, especially the number of carbonyl ligands, is increasing. However, the influence of carbonyl ligands on catalytic performance was different on reactions involving the adsorbed CO as an intermediate on the metal, such as CO oxidation. For instance, results on an MgAl2O4-supported iridium catalyst showed that the CO ligands present on the atomically dispersed iridium species offer a high performance for low temperature CO oxidation reaction.35 Moreover, density functional theory (DFT) calculations performed on atomically dispersed Rh supported on ZSM-5 for the

88

Challenges with atomically dispersed supported metal catalysts

Fig. 1 Structural transformations on the ligand environment of a highly dealuminated zeolite HY-supported Ir(C2H4)2 complex under various treatment conditions. Reprinted with permission from Ref. Lu, J.; Serna, P.; Gates, B. C. Acs Catal. 2011, 1(11), 1549–1561. Copyright (2011) American Chemical Society.

partial oxidation of methane showed that CO bonded to Rh strongly promotes CeOH bond formation, which is found to be the rate determining step instead of methane activation.36 However, it was also demonstrated that water was required in the reaction environment to hinder the poisoning of the catalyst by CO.36 Similar influence of the ligand environment on the catalytic performance was also demonstrated on an atomically dispersed Ru catalyst anchored on a polyphenylene framework containing different ligands. Data showed that -[bipy-Ru(II)(CO)2Cl2] anchored on the polyphenylene framework was active for CO oxidation and had an excellent stability in air for over 400 days, while -[bipyRu(III)Cl4] was completely inert.37 -[bipy-Ru(II)(CO)2Cl2] was obtained from -[bipy-Ru(III)Cl4] by Cl
/CO ligand exchange transforming Ru(III) to Ru(II) simultaneously.37 A direct treatment of -[bipy-Ru(III)Cl4] in flowing H2 resulted in the formation of Ru nanoparticles as H2 removes all Cl ligands. Thus, -[bipy-Ru(III)Cl4] was first treated in flowing CO at 433 K. EXAFS fitting confirmed that CO reacts with -[bipy-Ru(III)Cl4] forming mer(Cl)-[Ru(bipy)(CO)Cl3] and HCl. The ligand exchange proceeds by adding H2 to the feed, which reduces Ru and removes the other Cl ligand to obtain the catalytically active Ru site (-[bipy-Ru(II)(CO)2Cl2]) with the desired ligand environment.37 In a similar fashion, catalytic performance of an atomically dispersed Ru catalyst prepared using a layered double hydroxide as support for CO2 hydrogenation to yield formic acid was tuned by the incorporation of hydroxide ligands.38 Catalytic performance data showed that the high catalytic performance of this catalyst was not only ascribed to the support properties but also because of a triad of basic hydroxyl ligands, showing that a special placement of hydroxyl ligands is crucial for increasing the electron density on Ru.38 As in the case of homogenous-phase catalysts, the ligands might be large-sized organic groups as well. These organic ligands not only play an important role in controlling the electronic structure of the active species but also induce steric effects on the catalytic performance. For instance, CeO2-supported isolated Pt sites coordinated with a variety of N-based bidentate ligands (3,6(2-pyridyl)-1,2,4,5-tetrazine [DPTZ], 1,10-phenanthroline-5,6-dione [PDO], bathophenanthroline [BPhen], 4,40 -dinonyl-2,20 dipyridyl [C9BP], and 2,20 -bipyridine-4,40 -dicarboxylic acid [4,40 -BPDCA]) were prepared to understand the influence of the ligand environment on the catalytic performance for hydrosilylation of 1-octane and dimethoxymethylsilane, an important reaction in silicon chemistry.39 While the Pt-4,40 -BPDCA/CeO2 did not exhibit any catalytic performance, the other Pt-based atomically dispersed complexes with various N-based ligands provided activity for the reaction. Comparison of the reaction rates demonstrated a strong influence of the ligand characteristics on the catalytic performance. The hydrosilylation yield varied with the following order of the ligands: 4,40 -BPDCA < C9BP < BPhen < BPhen þ DPTZ < PDO < DPTZ.39 Switching from DPTZ to PDO decreased the yield from 90% to 66%, however it significantly enhanced the reusability of the catalyst. Besides, inactivity of the catalyst with

Challenges with atomically dispersed supported metal catalysts

89

4,40 -BPDCA ligand was attributed to the strong interaction between eCOOH and CeO2 causing the BPDCA to push Pt away from the support, demonstrating the steric influence of ligands on the catalytic performance.39 These examples demonstrate the strong influence of the non-support ligands on the catalytic performance of the atomically dispersed supported metal catalysts. The ligands directly influence the electronic structure of the active centers, and in addition to this electronic effect, they control the catalytic performance through steric effects as well. Thus, their design is crucial in tuning the catalytic properties as much as in the case of homogenous-phase catalysts.

6.05.2.1.2

Effects of supports as ligands

Supports act as macro-sized ligands to adjust the electronic environment of the atomically dispersed active metal centers.40 Quantification of the electronic structure on the metal centers is crucial in the way of unrevealing this influence. Data obtained by X-ray absorption near-edge structure (XANES) or X-ray photoelectron spectroscopies (XPS) might provide the required insights; however, the related features that can be measured in these techniques, in general, are strongly influenced by the coordination environment of the metal centers, which includes the non-support ligands as well. Thus, the data should be complemented with characterization of other structural information sensitive to the electronic structure of the metal centers. In this respect, simple non-support ligands can be used to probe the electron density on the metal centers. In this regard, CO is a perfect probe molecule, as the IR features associated with its adsorbed form on the metal site are sensitive to the electronic structure on the metal center. Because an excess electron density on the metal center can be shared with the carbon atom of the carbonyl ligand through p-back-donation. This electron exchange results in the weakening of the CeO bond leading to a red shift in the corresponding stretching frequency of the CO ligand in the IR spectrum, n(CO). Thus, n(CO) band position can be used to quantify the electron density on metal centers supported on different supports to unravel the extent of their ligand effect.41,42 As already mentioned in Section 6.05.2.1.1, reacting supported Ir(C2H4)2 complexes (or analogous supported Rh(C2H4)2 complexes) with CO results in the rapid exchange of ethylene ligands with CO forming supported Ir(CO)2 (or Rh(CO)2) complexes, irrespective of the support type. According to both n(CO) band positions in their IR spectra and X-ray absorption spectroscopy (XAS) data at the Ir LIII edge, it was inferred that the electron density on Ir increases as follows: Ir(CO)2/zeolite HY (Si/Al atomic ratio ¼ 30) < Ir(CO)2/zeolite Hb (Si/Al atomic ratio ¼ 19) < Ir(CO)2/zeolite HSSZ-53 (Si/Al atomic ratio ¼ 23)  Ir(CO)2/g-Al2O3 < Ir(CO)2/MgO.28,29 It was also shown that the degree of ligand exchange when treating these complexes in flowing C2H4 at 298 K and 1 bar depends strongly on the type of support. Data showed that the number of C2H4 ligands that can bound on each Ir site is correlated with the n(CO) positions: As iridium centers become more electron-rich with an increase in the electrondonor strength of the support, such as when they are supported on MgO, the number of C2H4 ligands that can replace CO decreases (Fig. 2).28 The choice of support as a ligand enables tuning of the electronic state of iridium by controlling the amount of electron they donate and by dictating the number of non-support ligands that the metal can hold.29 In this regard, quantifying the support’s electron-donor character offers opportunities for fine tuning of the electronic structure over the active metal centers. Point of zero charge (PZC), the pH at which the surface becomes neutral, of a metal oxide support offers a broad prospect. For instance, Ir(CO)2 complexes were synthesized on various high-surface-area metal oxide supports (SiO2, TiO2, Fe2O3, CeO2, MgO, and La2O3).43 Electron density over iridium on each support was characterized by XANES data collected at the Ir LIII edge and the corresponding positions of the symmetric n(CO) bands. The XANES data demonstrated that the edge energy of iridium in these metal oxide-supported complexes decreases with an increase in the PZC of the support (Fig. 3A). This trend shows that the electron-donor property of the metal oxides gets stronger (with an increase in the PZC value)

Fig. 2 Average number of C2H4 ligands that can bond to iridium when supported Ir(CO)2 complexes are treated in flowing C2H4 at 298 K and 1 bar. Reprinted with permission from Ref. Lu, J.; Aydin, C.; Browning, N.D.; Gates, B.C. Langmuir 2012, 28(35), 12806–12815. Copyright (2012) American Chemical Society.

90 Challenges with atomically dispersed supported metal catalysts

Fig. 3 Correlations describing the electron donation capabilities of the supports: (A) Ir LIII edge of the supported Ir(CO)2 complexes and the PZC of the supports, (B) Ir LIII edge and symmetric n(CO) band positions of the supported Ir(CO)2 complexes, and (C) symmetric n(CO) band positions of the supported Ir(CO)2 complexes and the PZC of the supports. Published by the Royal Society of Chemistry (under the terms of the CC-BY Creative Commons Attribution 3.0 Unported License). Babucci, M.; Fang, C.-Y.; Perez-Aguilar, J. E.; Hoffman, A. S.; Boubnov, A.; Guan, E.; Bare, S. R.; Gates, B. C.; Uzun, A. Chem. Sci. 2019, 10(9), 2623–2632.

Challenges with atomically dispersed supported metal catalysts

91

in the following order: SiO2 < TiO2 < Fe2O3 < CeO2 < MgO < La2O3.43 Fig. 3B and C show that both XANES and IR data provide consistent information about the change of electron density on iridium with a change in support’s PZC. After treating these supported Ir(CO)2 complexes in C2H4 at 373 K and partially replacing some of the inert CO ligands with C2H4, the catalysts were tested for the selective hydrogenation of 1,3-butadiene, a test reaction that is sensitive to the electronic environment over the metal centers. Data showed that the electron density on iridium controls the partial hydrogenation selectivity toward total butenes. The highest selectivity to butenes was observed when using a support having a strong electron-donor character, such as La2O3. In contrast, when using a support having a weak electron-donor character, such as SiO2, the selectivity to butenes was the lowest. A correlation of the selectivity to partial hydrogenation and the electron density on Ir probed by the corresponding n(CO) band position demonstrates a broad prospect for tuning the catalytic performance by the choice of support.43 Similar influence of the supports on metal’s catalytic properties was observed on Rh(CO)2 and Rh(C2H4)2 complexes supported on zeolite HY (with an Si/Al ratio of 30) and MgO for the same reaction.44 Rh(C2H4)2 supported on MgO provided a lower turnover frequency (TOF) compared to that of its counterpart supported on zeolite HY. The selectivity to total butenes for both catalysts was similar. However, independent of the non-support ligand environment of Rh, the selectivity to 1-butene was always higher on the strong-electron donating MgO. When considering the samples containing carbonyls as the ligands, MgO-supported Rh(CO)2 provided more than 99% selectivity toward total butenes, while zeolite HY-supported Rh(CO)2 provided a lower selectivity.44 Other than metal oxides, porous crystalline materials, such as MOFs, are also used as supports to synthesize atomically dispersed supported metal complexes. The electron-donor property of the support is crucial in determining the catalytic performance in these materials as well. Correlations between the electron density on the MgO-, ZrO2-, NU-1000-, UiO-67-, UiO-66-, and zeolite HYsupported Ir(C2H4)2 complexes (probed by the n(CO) band positions of their analogous supported Ir(CO)2 complexes) and their catalytic performances for ethylene hydrogenation and dimerization are provided in Fig. 4. Data showed that the catalytic performance for both ethylene hydrogenation and dimerization (TOF) increase as the iridium sites become less electron-rich, such as on zeolite HY having an Si/Al ratio of 30 (Fig. 4A). The same trend is valid for the selectivity of the dimerization (Fig. 4B).45 Silicoaluminophosphates (SAPOs) were also used as supports to atomically dispersed metal complexes.46 A comparison of the catalytic performance of analogous Rh complexes prepared on both SAPO and zeolites demonstrated that the SAPO supportedRh(C2H4)2 complexes were more active than their zeolite HY-supported counterparts. Besides, while the major product (with a selectivity of 99.6%) on Rh(C2H4)2/SAPO was ethane (remaining products being butane and butenes), the corresponding ethane selectivities were only 61.3% and 52.9% for the Rh(C2H4)2 complexes supported on zeolite HY with a Si/Al ratio of 2.6 and zeolite HY with a Si/Al ratio of 15, respectively.46 These findings showed that, even though the frameworks of SAPO and zeolite HY are isostructural, their electronic environment differ significantly because of their framework composition. EXAFS fits showed that RheO distance was shorter in SAPO-supported complexes (2.10 Å) compared to that in zeolite HY-supported complex (2.15 Å), implying that the electron withdrawing property of zeolite HY (with a Si/Al ratio of 15) is higher, explaining the reason for the selectivity difference. A longer distance between Rh and support’s oxygen signifies a weaker interaction of Rh and O in zeolite HY. Furthermore, the weaker acid strength of SAPO results in a lower tendency to form carbonaceous coke formation, which explains the enhanced lifetime of Rh(C2H4)2/SAPO.46 The formation of different products on isostructural zeolites implies the different effect of supports as macro-sized ligands. Revealing the effect of supports as ligands might become challenging when the support offers non-identical binding sites, as in the case of metal oxides. On these non-uniform supports, the binding of atomically dispersed metal complexes might occur on different sites, even though some of these sites might be more favorable compared to others, their surface densities might be limited. Thus, once they are all occupied with metal complexes, the other sites might act as the secondary adsorption sites with an increase in the metal loading. Consequently, low degree of uniformity of the binding sites might become an issue on non-crystalline supports especially at high metal loadings. For instance, MgO-supported Ir(CO)2 complexes prepared at an Ir loading of 1 wt% were characterized with relatively broad n(CO) bands with a full width at half maxima (FWHM) value of more than 25 cm
1, indicative of their low degree of uniformity, while those supported on zeolites were characterized with quite narrow n(CO) bands with a FWHM value in the range of only 5 cm
1, demonstrating their high degree of uniformity.47 It was shown that when Ir(C2H4)2 complexes were anchored on MgO at a significantly reduced Ir loading of 0.01 wt%, each iridium atom was bonded to three O atoms.48 At higher metal loadings, the terrace sites were also started to be populated with iridium complexes characterized by an IreO coordination number of 2.48 The bonding environment of iridium influences its catalytic performance. Iridium centers bonded to three oxygen atoms were found to be less active than those bonding to the support through two IreO bonds.48 This difference in performance was inferred to be associated with the limited availability of orbitals for electron sharing with the reactants when the complex is anchored on the support with three O atoms.48 Thus, other than the electron-donor character of the support, the local coordination environment that the support offers for anchoring the metal complexes is also crucial for tuning the catalytic performance of the atomically dispersed catalysts. To tune the coordination environment and the oxidation state of atomically dispersed active metal centers, treatment conditions of the catalyst can be varied. For instance, a Pt precursor was adsorbed on very small 5-nm-diameter anatase TiO2 nanoparticles at a Pt loading corresponding to an average 0.5 Pt atoms per TiO2 nanoparticle. The resulting sample was treated at three different conditions: oxidation at 300  C, mild reduction at 250  C, and harsh reduction at 450  C. By complementing experimental data (XAS, IR, high-angle annular dark field scanning transmission electron microscopy (HAADF-STEM)) with DFT calculations, it was proposed that the oxidative environment forces Pt atoms to substitute into Ti6c positions, mild reduction created (PtO2)ads species by pulling Pt out of the lattice, and harsh reduction created mobile (PtOH)ads moieties present at both step and terrace sites of the TiO2 nanoparticle. These differences observed in the local coordination environment of Pt sites upon various treatments were

92

Challenges with atomically dispersed supported metal catalysts

Fig. 4 Correlation between (A) TOF for ethylene conversion and (B) selectivity toward ethylene dimerization catalyzed by supported Ir(C2H4)2 complexes and n(CO) band positions of their analogous supported Ir(CO)2 complexes. Data were measured at 298 K and 1 bar using 30 mg of catalyst (1 wt% Ir loading). The feed partial pressures were 100 mbar C2H4, 200 mbar H2, 700 mbar He, where the total flow rate was 100 ml min
1. Reprinted with permission from Ref. Yang, D.; Odoh, S. O.; Borycz, J.; Wang, T. C.; Farha, O. K.; Hupp, J. T.; Cramer, C. J.; Gagliardi, L.; Gates, B. C. Acs Catal. 2016, 6(1), 235–247. Copyright (2016) American Chemical Society.

shown to significantly influence the catalytic performance for CO oxidation.49 The catalytic performance results of the catalysts following oxidation and mild reduction treatments provided similar reactivities with apparent activation energies of 72 and 78 kJ mol
1, respectively. However, the catalyst formed by the harsh reduction provided an apparent activation energy of 48 kJ mol
1 with an increase in reaction rate ranging from two- to five-fold depending on the reaction temperature.49 In a similar manner, a family of atomically dispersed Pt catalysts supported on Fe2O3 was prepared with a PteO(support) coordination number varying from 1.8 to 3.8 as a response to the changes in the treatment temperature. The coordination environment of Pt was linearly correlated with both the oxidation state of Pt sites and their corresponding TOF for chemoselective hydrogenation of 3-nitrostyrene (Fig. 5).50 Results illustrated that electronic environment over the metal is mainly dictated by their coordination environment on the support surface.50 Likewise, differences in treatment conditions were shown to result in changes in the coordination environment of atomically dispersed Fe in a metal-Cx-Ny catalyst. After pyrolyzing the starting material at 500, 600, and 700  C, Fe-N4 SAs/N-C, Fe-N3C1 SAs/N-C, and Fe-N2C2 SAs/N-C catalysts were obtained, respectively (Fig. 6A–C), where N-C

Challenges with atomically dispersed supported metal catalysts

93

Fig. 5 Correlation of TOF for chemoselective hydrogenation of 3-nitrostyrene (blue line, star) and the corresponding oxidation state of Pt (red line) with the corresponding coordination environment of atomically dispersed Pt sites on Fe2O3. Oxidation states were determined by both XPS (tringle) and XANES (circle). Catalytic performance data were measured at 40  C, using 10 mg of catalyst and 90 mg Fe2O3. The reaction mixture included 3nitrostyrene, toluene, and internal standard o-xylene pressurized with 3 bar H2. Reproduced with permission from Ref. Ren, Y.; Tang, Y.; Zhang, L.; Liu, X.; Li, L.; Miao, S.; Sheng Su, D.; Wang, A.; Li, J.; Zhang, T. Nat. Commun. 2019, 10(1), 4500 (under the terms of the CC-BY Creative Commons Attribution 4.0 International License). Copyright (2019) Nature Publishing Group.

Fig. 6 HAADF-STEM images of as-prepared catalysts consisting of different coordination environment: (A) Fe-N4 SAs/N-C, (B) Fe-N3C1 SAs/N-C, (C) Fe-N2C2 SAs/N-C, where N-C represents the N-doped carbon matrix and the subscripts, x and y, in “Fe-Cx-Ny” stand for the coordination environment of the Fe sites, (D) Catalytic performance for benzene oxidation of the as-prepared atomically dispersed samples in comparison with the bare N-C and N-C-supported iron nanoparticles. Data were measured after 24 h using 50 mg catalyst, 0.4 ml benzene, 6 ml H2O2 (30 wt%), and 3.0 ml CH3CN. After 24 h reaction at 30  C, 63.4 mg n-hexadecane was used as the internal standard for analysis. Reproduced with permission from Ref. Pan, Y.; Chen, Y.; Wu, K.; Chen, Z.; Liu, S.; Cao, X.; Cheong, W. C.; Meng, T.; Luo, J.; Zheng, L.; Liu, C.; Wang, D.; Peng, Q.; Li, J.; Chen, C. Nat. Commun. 2019, 10(1), 4290 (under the terms of the CC-BY Creative Commons Attribution 4.0 International License). Copyright (2019) Nature Publishing Group.

represents the N-doped carbon matrix and the subscripts, x and y, in “Fe-Cx-Ny” stand for the coordination environment of the Fe sites in single atom catalysts (SAs).51 Among these catalysts, Fe-N4 SAs/N-C exhibited the highest conversion for benzene oxidation with 100% selectivity to phenol (Fig. 6D), along with a higher adsorption capacity of O2 compared to the other catalysts.51 Likewise, performing calcination on an ethylene glycolate containing atomically dispersed TiO2-supported Pd catalyst showed that

94

Challenges with atomically dispersed supported metal catalysts

Fig. 7 (A) Structure of DPTZ, demonstration of (B) pristine TiO2 (101), and (C) defective TiO2-supported Pt-DPTZ catalyst. Reproduced from Ref. Zhou, X.; Chen, L.; Sterbinsky, G. E.; Mukherjee, D.; Unocic, R. R.; Tait, S. L. Catal. Sci. Technol. 2020, 10(10), 3353–3365 with permission from The Royal Society of Chemistry.

more Ti(III)eOePd interfaces were exposed to the reactants upon the removal of ethylene glycolate during calcination, facilitating the activation of O2 and resulting in a better activity for CO oxidation.52 Furthermore, oxidation state of atomically dispersed Pt supported on a defective ceria was controlled by changing the reduction temperature, which played a major role in its catalytic performance for CO oxidation, CH4 combustion, and NO oxidation.53 In a different investigation, Pt metal with DPTZ and Cl ligands were anchored on defective titania and pristine titania.54 The defective titania enabled the coordination of Pt with more chloride ligands, resulting in a shortened induction period, enabling the dispersion of DPTZ on the defective sites easing the contact between Pt and reactants for alkene hydrosilylation reaction (Fig. 7).54 Besides the treatment conditions, changing the synthesis method can also significantly alter the electronic properties on the active centers. For instance, to prepare atomically dispersed Ag atoms on microporous hollandite manganese oxide, two different methods were used: anti-Ostwald ripening and conventional wet impregnation.55 Synchrotron X-ray powder diffraction data and EXAFS fit results confirmed that the interatomic distances of AgdO were shorterdindicating a stronger oxygen-silver interactiondon the sample prepared by anti-Ostwald ripening compared to its counterpart in the sample prepared by the conventional wet impregnation. The stronger electronic metal-support interaction caused a lower activation energy leading to a higher catalytic performance for the oxidation of formaldehyde.55 These examples demonstrate that the supports play a significant role in controlling the electronic structure of the atomically dispersed metal. The influence depends strongly on their electron-donor character and the coordination environment that they offer for anchoring the active metal centers.

6.05.2.1.3

Effects of metal nuclearity

As highlighted above, electronic structure of the active metal centers depends on their coordination environment set by the neighboring atoms, such as ligands including supports. However, sometimes small active metal centers can be present as sub-nanometersized clusters of only a few metal atoms still offering atomic dispersion of the active centers. In such clusters, the metal nuclearity plays an important role in controlling the catalytic properties. Formation of small clusters from mononuclear metal complexes in various reactive environments is well-characterized by operando XAS measurements for supported Ir(C2H4)2 and Rh(C2H4)2 complexes.56–59 The nuclearity of iridium sites supported on MgO and dealuminated zeolite HY (Si/Al atomic ratio ¼ 30) influenced the catalytic performance for ethylene hydrogenation.58 For both supports, the TOF for ethylene hydrogenation was higher when the metal centers were in the form of Ir4 clusters, for which all metal atoms were assumed to be available for the reaction.58 However, the performance was also strongly dependent on the type of support as mentioned in Section 6.05.2.1.2. The increase in TOF was 6-fold as Ir(C2H4)2 was converted to Ir4 when the support was MgO, but it was only 1.2-fold when the support was dealuminated zeolite HY.58 Ir4 clusters formed on UiO-67 provided twice as high TOF compared to the UiO-67-supported Ir(C2H4)2 complexes for ethylene conversion.60 Ir4 clusters on UiO-67 were less selective for dimerization compared to mononuclear complexes.60 The rate-determining step for UiO-67 supported Ir4 clusters and Ir(C2H4)2 complexes differed: On the isolated iridium centers the rate-determining step was H2 activation on iridium, while it was the hydrogenation of iridium bound ethyl ligands on the clusters.60 MgO-supported Rh2(CO)6 clusters were found to provide 99% selectivity for n-butenes at a 1,3-butadiene conversion as high as 97%, which makes this catalyst even more selective than gold catalysts.61 The mononuclear Rh(CO)2 complexes supported on MgO, on the other hand, provided less than 1% conversion and only 80% selectivity to n-butenes under identical conditions.61 Similar differences in catalytic performance with a change in metal nuclearity was also demonstrated on metal oxide-supported Ga catalysts. Reaction of Ga(i-Bu)3 (where i-Bu represents CH2CH-(CH3)2) with alumina and silica showed that grafting Ga(i-Bu)3 on g-Al2O3 leads to the formation of site-isolated Ga complexes (Ga-(iBu)2), whereas grafting it on SiO2 resulted in the formation of Ga dimers (Ga2(i-Bu)3).62 The mononuclear Ga sites supported on alumina provided much higher propane dehydrogenation activity than the dimers supported on SiO2, confirming the importance of the proximity of Ga and Al sites trough AleOeGa bonds to promote heterolytic CeH bond activation.62 Moreover, these active species with different metal nuclearities can be interconverted into each other upon the changes in the reactant gas composition, offering opportunities for tuning the catalytic properties with the partial pressures of reactants. For instance, XANES data measured during a H2 treatment of Ir(C2H4)2 complexes supported on dealuminated zeolite HY with an Si/Al ratio of 30 during a ramp to 353 K at 1 bar showed the presence of isosbestic points indicating a clear change from one uniform structure into another (Fig. 8A).56 EXAFS data measured real-time during the treatment showed the appearance of an IreIr shell accompanied by a decrease in the IreO and IreC coordination numbers associated with a decrease in those characterizing

Challenges with atomically dispersed supported metal catalysts

Fig. 8 (A) Normalized XANES spectra recorded at the Ir LIII edge of the dealuminated zeolite-Y-supported Ir(C2H4)2 complex during cluster formation when the temperature was increasing from 298 K to 353 K following a ramp rate of 1 K min
1 (inset represent the change in white line intensity with respect to time on stream), (B) EXAFS analysis showing the change in coordination numbers of dealuminated zeolite Y-supported Ir(C2H4)2 during cluster formation as the temperature was increasing from 298 K to 353 K following a ramp rate of 1 K min
1, and (C) proposed model for cluster formation. Reproduced from Ref. Uzun, A.; Gates, B. C. Angew. Chem. Int. Ed. 2008, 47(48), 9245–9248. Copyright 2008 Wiley-VCH.

95

96

Challenges with atomically dispersed supported metal catalysts

Ir-support and the Ir-ligand bonds, respectively (Fig. 8B). The final IreIr coordination number was 3, indicating the formation of Ir4 clusters at the end of the temperature ramp.56 A model was constructed by combining EXAFS, XANES, and IR data to show the cluster formation demonstrating the changes in metal–metal and metal–ligand coordination (Fig. 8C). Breaking-up of the Ir4 clusters by oxidative fragmentation to form site-isolated metal atoms was possible by treating the sample consisting of Ir4 clusters in flowing C2H4.56 Such reversible interconversion was also observed during ethylene hydrogenation on the same catalyst, where the feed composition was changed continuously between an ethylene-rich (C2H4/H2 ¼ 4) to an equimolar (C2H4/H2 ¼ 1) composition and then to H2-rich (C2H4/H2 ¼ 0.3) composition.57 Data demonstrated that under the ethylene-rich feed flow, the iridium was present as mononuclear complexes on the support (zeolite HY, Si/Al atomic ratio ¼ 30), while switching the gas composition to H2-rich feed caused the formation of Ir4 clusters (Fig. 9). When the feed composition was cycled between ethylene-rich and H2rich, data demonstrated that all these structural changes were reversible and that it is possible to tune the nuclearity of the atomically dispersed metals in the working state by the choice of feed composition (Fig. 9). Such reversible changes in metal nuclearity occurring as a response to the changes in the feed gas composition were also observed on zeolite-supported Rh.59,63 For instance, in a recent example, zeolite HY (Si/Al ¼ 30)-supported Rh(CO)2 complexes were converted into Rh4(CO)12 clusters under water-gas shift conditions at 308 K with a high yield. Upon a switch of flow composition to pure He data showed redispersion of the clusters at 353 K.63 These changes in metal nuclearity can potentially induce significant changes in the catalytic properties. For example, mononuclear Rh complexes supported on zeolite HY (Si/Al ¼ 30) were found to catalyze both ethylene dimerization and ethylene hydrogenation, with n-butene being the main products, under an ethylene-rich reactant composition (C2H4/H2 ¼ 4).59 When switching to a H2-rich feed (C2H4/H2 ¼ 0.25), the selectivity to the hydrogenation products (butane and ethane) increased. Next, switching to a pure H2 mixture resulted in the formation of extremely small clusters of Rh, with a final coordination number of 1.9. Then, switching to a H2-rich feed mixture (C2H4/H2 ¼ 0.25) resulted in a high selectivity for hydrogenation and a significantly high catalytic activity on the clusters. The clusters were then converted to mononuclear Rh by switching to an ethylene-rich feed. This change in metal nuclearity led to a decrease in hydrogenation rate accompanied by an increase in the selectivity for n-butene.59 These results demonstrate that the metal nuclearity in atomically dispersed supported metal catalyst is sensitive to the reactant gas composition and it imposes a strong influence on the catalytic properties. Thus, the active species should be identified in detail at different reaction conditions to be able to evaluate the structure-performance relationships. Besides, these results can be further extended to traditional supported metal catalysts, for which a conventional approach in evaluating the reaction kinetics relies directly on the measurement of the changes in reaction rate as a response to the changes in the partial pressures of the reactants. And for these measurements, the conventional approach is to assume that the structure of the active sites remains the same throughout the measurements. However, the data obtained on the atomically dispersed supported metal catalysts resolving the structural changes with gas composition illustrate that this assumption might not be valid all the time, suggesting that the active species under different reaction conditions should be identified for a proper kinetic analysis. Similarly, structural transformations on a mononuclear Pt catalyst supported on g-Al2O3 was investigated during CO oxidation by in-situ XAS and IR spetroscopies.64 Fig. 10 demonstrates the white line intensities (Fig. 10A) and PtePt coordination numbers (Fig. 10B) during the calcination in air, CO oxidation reaction, reduction in hydrogen, and CO oxidation measured on 0.3 wt% Pt/ g-Al2O3 and 1.0 wt% Pt/g-Al2O3 in O2-rich or O2-poor conditions.64 During calcination and CO oxidation conditions (at high

Fig. 9 The change in coordination numbers obtained by three shell model EXAFS analysis of dealuminated zeolite Y-supported iridium catalyst initially present as supported Ir4 clusters. The coordination numbers in the ethylene-rich (red background), equimolar (green background), and H2rich (blue background) region show the continuous change of the coordination numbers indicating the formation of mononuclear Ir(C2H4)2 sites or Ir4 clusters depending on the feed composition. Reprinted with permission from Ref. Uzun, A.; Gates, B. C. J. Am. Chem. Soc. 2009, 131(43), 15887– 15894. Copyright (2009) American Chemical Society.

Challenges with atomically dispersed supported metal catalysts

97

Fig. 10 (A) White line intensities obtained from XANES and b) PtePt coordination number during subsequent calcination, CO oxidation, reduction, and CO oxidation environments measured on 0.3 wt% Pt/g-Al2O3 in oxygen-rich conditions (denoted as COOX10, CO:O2:He ¼ 2:10:88%) and oxygen-poor conditions (denoted as COOX2, CO:O2:He ¼ 2:2:96%) and 1.0 wt% Pt/g-Al2O3 in oxygen-rich conditions (COOX10). CO oxidation performance data were measured by cycling the temperature between 50  C and 280  C, using a total flow rate of 50 ml min
1 of oxygen rich or oxygen-poor conditions. The Pt amount was constant in all catalytic testing (500 mg). Reprinted with permission from Ref. Dessal, C.; Len, T.; Morfin, F.; Rousset, J.-L.; Aouine, M.; Afanasiev, P.; Piccolo, L. Acs Catal. 2019, 9(6), 5752–5759. Copyright (2019) American Chemical Society.

conversion), white line intensity was the highest. During the reduction treatment, white line intensity decreased accompanied by a decrease in the PteO coordination number along with a significant increase in PtePt coordination number exceeding 5 (Fig. 10B), indicating the cluster formation.64 Increasing Pt loading or decreasing the O2 concentration accelerates the cluster formation and decreases the oxidation state. As in the case of Ir, clusters of Pt were found to be more active compared to single Pt atoms. The

98

Challenges with atomically dispersed supported metal catalysts

Fig. 11 State of the Pt species tracked by in-situ TEM at various reaction conditions. Reproduced with permission from Ref. Liu, L.; Zakharov, D. N.; Arenal, R.; Concepcion, P.; Stach, E. A.; Corma, A. Nat. Commun. 2018, 9(1), 574 (under the terms of the CC-BY Creative Commons Attribution 4.0 International License). Copyright (2018) Nature Publishing Group.

pre-reduced catalyst, consisting of clusters, was found to be much more active than the pre-calcined catalyst, which consisted mostly of atomically dispersed Pt, along with some clusters. However, when the CO conversion reached 100%, no PtePt contribution could be observed (Fig. 10B, second CO þ O2 region).64 Disappearance of PtePt contribution at high CO conversion was attributed to the oxidation of Pt clusters to amorphous PtOx species rather than redispersion, which was also verified by STEM images indicating PtOx particles with an average size of 1.1  0.3 nm. At 100% conversion, only O2 and inert gas is present in the reactor, creating an oxidative environment and leading to PtOx formation. Cooling down the reactor resulted in the appearance of PtePt contribution (Fig. 10B). Thus, changing the metal nuclearity with pretreatments influenced the catalytic performance for CO oxidation, which was evidenced by in-situ characterization techniques. The dynamic changes of Pt species confined in MCM-22 zeolite were investigated by in-situ TEM at different reaction conditions.65 TEM images obtained under various oxidizing and reducing environments at different temperatures along with CO oxidation performance measurements showed that these conditions control the cluster formation or redispersion behavior as summarized in Fig. 11.65 Reductive environment between 100  C and 300  C (CO þ H2O and CO þ O2) leads to Pt cluster formation. A higher temperature (400  C) in these environments causes a further growth of Pt clusters in CO þ H2O environment, whereas a Pt redispersion into smaller clusters or even atomically dispersed sites is observed for CO þ O2 condition.65 In an oxidative atmosphere (NO þ H2 and NO þ CO) Pt species initially present as atomically dispersed form formed clusters, when increasing the temperature to 200  C. An increase in temperature to 400  C caused a redispersion of these clusters, whereas a further increase to 800  C or higher led to an agglomeration of atomically dispersed Pt sites.65

6.05.2.1.4

Effects of promoters

Promoters or dopants influence the electronic properties of the active metals either directly or indirectly. One approach is the addition of alkali metals, which both stabilize the atomically dispersed metals and enhance their catalytic properties. For example, it was demonstrated that addition of Naþ ions enabled the Pt sites to bond to Naþ ions through eO ligands (PteOxeNa). This approach was effective in immobilizing the Pt sites on active and inert supports, such as anatase (TiO2), microporous K-type L-zeolite (KLTL), and mesoporous silica MCM-41.66 These Naþ containing atomically dispersed supported Pt catalysts exhibited a significant enhancement in the catalytic performance for the water-gas shift reaction (Fig. 12A).66 Increase in the reaction rate was associated with the eOH groups on the surface (Fig. 12B), and the active sites were determined as Pt–O(OH)x-species.66 These eOH groups can be supplied by the active metal oxides (such as TiO2) or by the presence of Naþ when the support was inert (which are not able to supply active eOH groups, such as KLTL and MCM-41). Thus, by the addition of alkali Naþ, the structure of the support could be tuned so that inert supports, such as zeolites or SiO2, could act as TiO2.66 Similarly, propylene selectivity of an atomically dispersed Pt/CeO2 catalyst for propane dehydrogenation could be significantly enhanced by the addition of Sn to CeO2.67 Characterization data suggested that Pt sites underwent a structural transformation with Sn atoms on the surface to form subnanometer PteSn clusters during the reaction. Catalytic performance results indicated that this structural transformation of the Pt sites with the dopant improved the propylene selectivity from 0% to 84.5%.67 Upon oxidation of these CeO2 supported PteSn clusters, single Pt centers and SnO2 particles were formed on the surface, showing that these transformations were reversible.67 When phosphorus was used as a promoter in a catalyst consisting of CeO2-supported isolated Pt centers, spectroscopy data indicated that Pt centers have a higher valence state in the presence of P.68 This charge transfer from P to Pt centers resulted in a 10-fold enhancement for styrene, cyclohexane, phenylacetylene, and nitrobenzene hydrogenation performance.68

Challenges with atomically dispersed supported metal catalysts

99

Fig. 12 (A) Water gas-shift reactivities of supported atomically dispersed Pt catalysts with and without the addition of Na measured using a reformate-type gas mixture (11% CO, 26% H2O, 7% CO2, and 26% H2, balance He) at a flow rate of 207 ml min
1 on a catalyst bed containing 100–500 mg of each catalyst to reach a CO conversion of 10 wt%) and high porosity of the support (> 1000 m2 g
1).115 Noble metals were loaded on S-doped mesoporous carbon by thermal reduction under Ar or H2. The PteS coordination number determined by EXAFS was found to be as 3.6  0.3 for the 10 wt% Pt-loaded catalyst, while increasing the

Fig. 17 Cascade anchoring method to synthesize N-doped carbon-supported atomically dispersed metal catalysts. Reproduced with permission from Ref. Zhao, L.; Zhang, Y.; Huang, L. B.; Liu, X. Z.; Zhang, Q. H.; He, C.; Wu, Z. Y.; Zhang, L. J.; Wu, J.; Yang, W.; Gu, L.; Hu, J. S.; Wan, L. J. Nat. Commun. 2019, 10(1), 1278 (under the terms of the CC-BY Creative Commons Attribution 4.0 International License). Copyright (2019) Nature Publishing Group.

108

Challenges with atomically dispersed supported metal catalysts

Pt loading to 20 wt% resulted in aggregation, characterized by a PtePt coordination number of 2.3  0.7.115 Likewise, S-doped zeolite-templated carbon with a high amount of sulfur (17 wt% S) was used to anchor atomically dispersed Pt sites (5 wt%) by forming PteS4 complexes.128 Pyrolysis of metal containing porous materials, such as MOFs, is another approach for increasing the metal loading. For instance, coordinatively unsaturated NieN sites in porous carbon were prepared by the pyrolysis of Zn/Ni bimetallic ZIF-8, resulting in a Ni loading of 5.4 wt% and exhibiting a high catalytic activity toward CO2RR.119 Similarly, the pyrolysis of Fe-modified ZIF-8 resulted in FeN4 sites on carbon.109 However, unlike the conventional ZIF-8 structure, a novel steric tetra-imidazole structure was used to supply the high density of N-rich sites enabling the anchoring of Fe effectively.109 The catalyst obtained provided a metal loading of 2.6 wt% and was active for ORR with a high stability.109 Fe112 and Co105 anchored on porous porphyrinic triazine-based frameworks led to metal loadings of 8.3 and 5.2 wt%, respectively, and providing active sites characterized as FeeN4 and CoeN4, respectively. A method involving the pyrolysis of a coordinated polymer was developed to reach extremely high metal loadings.110 For instance, an atomically dispersed Fe-based catalyst with an Fe loading of 30.0 wt% was prepared by a coordinated polymer strategy, where formaldehyde and dicyandiamide was polymerized and Fe ions were coordinated with the resin of dicyandiamideformaldehyde.110 The solid material was then treated first in Ar at 600  C and then in 5% H2 (in balance Ar) at 400  C. The resulting catalyst exhibited a high catalytic performance for the epoxidation of styrene with 64% yield and a selectivity of 89% toward styrene oxide, while iron nanoparticles and ironporphyrin were found to be inactive. The catalyst provides high stability even after five runs. The strategy was applicable for many other metals as well. Following it, 21.6, 22.4, 21.1, 13.5, 3.5, 3.8, 3.2, and 4.4 wt% loadings were reached using Ni, Cu, Zn, Ru, Rh, Pd, Pt, and Ir, respectively.110 Apart from these carbon-based support materials, there exist several other examples utilizing various strategies to enhance the metal loading of isolated metals supported on metal oxides. For instance, using a commercially available CeO2 support, 3 wt% of Pt was anchored by atom trapping.124 Specifically, after depositing the Pt precursor by wet impregnation in the pores of CeO2, the catalyst was dried and then calcined at 800  C in air for 10 h, leading to the reaction of PtO2 with under-coordinated Ce cations at CeO2(111) step edges.124 Another strategy involves the use of alkaline ions to stabilize Pt sites at a loading of 2 wt% prepared in a single-step incipient wetness impregnation on ordinary supports, such as SiO2 (Pt1-Ox-Cs/SiO2) or Al2O3 (Pt1-Ox-K/Al2O3), as discussed above in the section related to the stability.122 Metal sulfides are also interesting support materials allowing the anchoring of atomically dispersed metals at a high metal loading. For instance, Pt supported on MoS2 with Pt loadings up to 7.5 wt% characterized with the lack of any PtePt coordination were synthesized and used for CO2 hydrogenation to produce methanol.127 For the catalyst exhibiting a Pt loading of 7.5 wt%, it was found that two Pt atoms were replacing the neighboring Mo atoms present in the MoS2 nanosheets, while the Pt sites in 0.2 wt% Pt-loaded catalysts occupied only one Mo atom.127 Different intermediates were observed during CO2 hydrogenation on these two catalysts with different loadings. In-situ XPS and diffuse reflectance infrared Fourier transform analysis evidenced that the intermediate for the 0.2 wt% Pt-loaded catalyst containing the isolated Pt sites was CH2OH*, while on the 7.5 wt% Pt-loaded catalyst having neighboring Pt sites the intermediate was COOH*. Furthermore, H2-temperature programmed desorption showed that on the sample with a Pt loading of 7.5 wt%, H2 desorption temperature was decreased, indicating that H2 dissociation is promoted on neighboring Pt sites. Combining these characterization results with DFT calculations, it was revealed that the neighboring Pt monomers worked synergistically making them highly active for CO2 hydrogenation and they reduced the activation energy barrier as compared to the isolated monomers by a sequential transformation of CO2 to first formic acid and then methanol.127 In contrast, mechanistic studies showed that on the isolated sites, CO2 was hydrogenated directly to methanol.127

6.05.3

Summary

Atomically dispersed supported metal catalysts contain well-defined and structurally uniform active metal centers, which ease the understanding of structure-performance relationships. These materials have gained a tremendous attention, especially in the last decade, because of their unique catalytic properties and opportunities that they offer for maximum utilization of expensive noble metals. However, the need for tuning of the catalytic properties, maintaining the atomic level dispersion under reaction conditions, and increasing the metal loading to obtain a high density of active centers set the main challenges related with the atomically dispersed supported metal catalysts. This chapter provides an account of some of the breakthrough examples from the recent literature focusing on overcoming these challenges. Control of the electronic structure on the active metal centers by the ligand environment including supports, adjusting the metal nuclearity, addition of dopants, and even coating the surfaces with a thin layer of IL coatings are some of the strategies demonstrated to be useful in tuning the catalytic performance. Furthermore, adjusting the coordination environment of the active metal on supports also provides a flexibility in controlling the catalytic performance. Next, the limited ability to enhance the stability of atomically dispersed metal catalysts was considered. In this regard, consequences of the above-mentioned approaches employed to tune the electronic structure of the metal on the stability of the active centers under reaction conditions were investigated. It is highlighted that use of supports with a high density of defects, mostly reducible oxides, is key in maintaining the atomic dispersion. Atom trapping is a widely utilized method which allows atomic dispersion at elevated temperatures as high as 800  C under oxidative or inert environments. Other strategies for successfully stabilizing single atoms on supports at even reaction conditions were lowering the metal loading, introducing dopants, such as K, Na, Ba, and even IL coatings, the use of MOFs or zeolites to trap metal atoms, or the use of mesoporous supports, such as mesoporous Al2O3. Finally, the limited ability of increasing the metal loadings in the way of obtaining a high density of active centers in

Challenges with atomically dispersed supported metal catalysts

109

a unit reactor volume is considered. In this regard, especially the progress with the use of carbon-based supports having adjustable functional groups were presented. O-, N-, and S-containing functional groups were found to provide strong binding sites for atomically dispersing the noble metals on carbon-based supports enabling high metal loading (> 5 wt%). The examples covered in this chapter present creative strategies to overcome some of the major challenges related with the atomically dispersed supported metal catalysts. They help in setting the guidelines for moving the research further in this fairly old, yet freshly and rapidly developing field.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58.

Bell, A. T. Science 2003, 299 (5613), 1688–1691. Gates, B. C. J. Mol. Catal. A: Chem. 2000, 163 (1), 55–65. Liu, L.; Corma, A. Chem. Rev. 2018, 118 (10), 4981–5079. Cooper, C.; Dooley, K. M.; Fierro-Gonzalez, J. C.; Guzman, J.; Jentoft, R.; Lamb, H. H.; Ogino, I.; Runnebaum, R. C.; Sapre, A.; Uzun, A. Acs Catal. 2020, 10 (20), 11912– 11935. Copéret, C.; Chabanas, M.; Petroff Saint-Arroman, R.; Basset, J.-M. Angew. Chem. Int. Ed. 2003, 42 (2), 156–181. Thomas, J. M.; Raja, R.; Lewis, D. W. Angew. Chem. Int. Ed. 2005, 44 (40), 6456–6482. Gates, B. C. Top. Catal. 2000, 14 (1), 173–180. Uzun, A.; Dixon, D. A.; Gates, B. C. Chemcatchem 2011, 3 (1), 95–107. Resasco, J.; Christopher, P. J Phys Chem Lett 2020, 10114–10123. Mitchell, S.; Vorobyeva, E.; Perez-Ramirez, J. Angew. Chem. Int. Ed. 2018, 57 (47), 15316–15329. Uzun, A.; Ortalan, V.; Browning, N. D.; Gates, B. C. Chem. Commun. 2009, (31), 4657–4659. Ortalan, V.; Uzun, A.; Gates, B. C.; Browning, N. D. Nat. Nanotechnol. 2010, 5 (7), 506–510. Liu, L.; Lopez-Haro, M.; Calvino, J. J.; Corma, A. Nat. Protoc. 2020. https://doi.org/10.1038/S41596-020-0366-9. Gates, B. C.; Flytzani-Stephanopoulos, M.; Dixon, D. A.; Katz, A. Catal. Sci. Technol. 2017, 7 (19), 4259–4275. Qin, R.; Liu, P.; Fu, G.; Zheng, N. Small Methods 2018, 2 (1), 1700286. Yang, X. F.; Wang, A.; Qiao, B.; Li, J.; Liu, J.; Zhang, T. Acc. Chem. Res. 2013, 46 (8), 1740–1748. Lang, R.; Du, X.; Huang, Y.; Jiang, X.; Zhang, Q.; Guo, Y.; Liu, K.; Qiao, B.; Wang, A.; Zhang, T. Chem. Rev. 2020. https://doi.org/10.1021/Acs.Chemrev.0c00797. Ji, S.; Chen, Y.; Wang, X.; Zhang, Z.; Wang, D.; Li, Y. Chem. Rev. 2020. https://doi.org/10.1021/Acs.Chemrev.9b00818. Babucci, M.; Guntida, A.; Gates, B. C. Chem. Rev. 2020. https://doi.org/10.1021/Acs.Chemrev.0c00864. Lai, W. H.; Miao, Z.; Wang, Y. X.; Wang, J. Z.; Chou, S. L. Adv. Energy Mater. 2019, 9 (43), 1900722. Liu, J. Acs Catal. 2017, 7 (1), 34–59. Flytzani-Stephanopoulos, M.; Gates, B. C. Annu. Rev. Chem. Biomol. 2012, 3, 545–574. Wang, K.; Wang, X.; Liang, X. Chemcatchem 2020. https://doi.org/10.1002/Cctc.202001255. Qin, R.; Liu, K.; Wu, Q.; Zheng, N. Chem. Rev. 2020. https://doi.org/10.1021/Acs.Chemrev.0c00094. Piccolo, L. Catal. Today 2020. https://doi.org/10.1016/J.Cattod.2020.03.052. Liu, L.; Corma, A. Trends In Chemistry 2020. https://doi.org/10.1016/J.Trechm.2020.02.003. Liu, D.; He, Q.; Ding, S.; Song, L. Adv. Energy Mater. 2020, 10 (32), 2001482. Lu, J.; Aydin, C.; Browning, N. D.; Gates, B. C. Langmuir 2012, 28 (35), 12806–12815. Lu, J.; Serna, P.; Gates, B. C. Acs Catal. 2011, 1 (11), 1549–1561. Lu, J.; Aydin, C.; Liang, A. J.; Chen, C.-Y.; Browning, N. D.; Gates, B. C. Acs Catal. 2012, 2 (6), 1002–1012. Uzun, A.; Bhirud, V. A.; Kletnieks, P. W.; Haw, J. F.; Gates, B. C. J. Phys. Chem. C 2007, 111 (41), 15064–15073. Liang, A. J.; Bhirud, V. A.; Ehresmann, J. O.; Kletnieks, P. W.; Haw, J. F.; Gates, B. C. J. Phys. Chem. B 2005, 109 (51), 24236–24243. Ogino, I.; Gates, B. C. J. Phys. Chem. C 2010, 114 (6), 2685–2693. Bhirud, V. A.; Ehresmann, J. O.; Kletnieks, P. W.; Haw, J. F.; Gates, B. C. Langmuir 2006, 22 (1), 490–496. Lu, Y.; Wang, J.; Yu, L.; Kovarik, L.; Zhang, X.; Hoffman, A. S.; Gallo, A.; Bare, S. R.; Sokaras, D.; Kroll, T.; Dagle, V.; Xin, H.; Karim, A. M. Nat. Catal. 2019, 2 (2), 149–156. Bunting, R. J.; Thompson, J.; Hu, P. Phys. Chem. Chem. Phys. 2020, 22 (20), 11686–11694. Kang, L.; Wang, B.; Thetford, A.; Wu, K.; Danaie, M.; He, Q.; Gibson, E.; Sun, L. D.; Asakura, H.; Catlow, R.; Wang, F. R. Angew. Chem. Int. Ed. 2020. https://doi.org/ 10.1002/Anie.202008370. Mori, K.; Taga, T.; Yamashita, H. Acs Catal. 2017, 7 (5), 3147–3151. Chen, L.; Ali, I. S.; Tait, S. L. Chemcatchem 2020, 12 (13), 3576–3584. Witzke, R. J.; Chapovetsky, A.; Conley, M. P.; Kaphan, D. M.; Delferro, M. Acs Catal. 2020, 10 (20), 11822–11840. Zhao, A.; Gates, B. C. J. Am. Chem. Soc. 1996, 118 (10), 2458–2469. Babucci, M.; Uzun, A. J. Mol. Liq. 2016, 216, 293–297. Babucci, M.; Fang, C.-Y.; Perez-Aguilar, J. E.; Hoffman, A. S.; Boubnov, A.; Guan, E.; Bare, S. R.; Gates, B. C.; Uzun, A. Chem. Sci. 2019, 10 (9), 2623–2632. Yardimci, D.; Serna, P.; Gates, B. C. Acs Catal. 2012, 2 (10), 2100–2113. Yang, D.; Odoh, S. O.; Borycz, J.; Wang, T. C.; Farha, O. K.; Hupp, J. T.; Cramer, C. J.; Gagliardi, L.; Gates, B. C. Acs Catal. 2016, 6 (1), 235–247. Perez-Aguilar, J. E.; Chen, C.-Y.; Hughes, J. T.; Fang, C.-Y.; Gates, B. C. J. Am. Chem. Soc. 2020, 142 (26), 11474–11485. Hoffman, A. S.; Fang, C. Y.; Gates, B. C. J. Phys. Chem. Lett. 2016, 7 (19), 3854–3860. Hoffman, A. S.; Debefve, L. M.; Zhang, S.; Perez-Aguilar, J. E.; Conley, E. T.; Justl, K. R.; Arslan, I.; Dixon, D. A.; Gates, B. C. Acs Catal. 2018, 3489–3498. Derita, L.; Resasco, J.; Dai, S.; Boubnov, A.; Thang, H. V.; Hoffman, A. S.; Ro, I.; Graham, G. W.; Bare, S. R.; Pacchioni, G.; Pan, X.; Christopher, P. Nat. Mater. 2019, 18 (7), 746–751. Ren, Y.; Tang, Y.; Zhang, L.; Liu, X.; Li, L.; Miao, S.; Sheng Su, D.; Wang, A.; Li, J.; Zhang, T. Nat Commun 2019, 10 (1), 4500. Pan, Y.; Chen, Y.; Wu, K.; Chen, Z.; Liu, S.; Cao, X.; Cheong, W. C.; Meng, T.; Luo, J.; Zheng, L.; Liu, C.; Wang, D.; Peng, Q.; Li, J.; Chen, C. Nat. Commun. 2019, 10 (1), 4290. Liu, P.; Zhao, Y.; Qin, R.; Gu, L.; Zhang, P.; Fu, G.; Zheng, N. Sci. Bull. 2018, 63 (11), 675–682. Jeong, H.; Shin, D.; Kim, B. S.; Bae, J.; Shin, S.; Choe, C.; Han, J. W.; Lee, H. Angew. Chem. Int. Ed. 2020. https://doi.org/10.1002/Anie.202009776. Zhou, X.; Chen, L.; Sterbinsky, G. E.; Mukherjee, D.; Unocic, R. R.; Tait, S. L. Catal. Sci. Technol. 2020, 10 (10), 3353–3365. Hu, P.; Huang, Z.; Amghouz, Z.; Makkee, M.; Xu, F.; Kapteijn, F.; Dikhtiarenko, A.; Chen, Y.; Gu, X.; Tang, X. Angew. Chem. Int. Ed. 2014, 53 (13), 3418–3421. Uzun, A.; Gates, B. C. Angew. Chem. Int. Ed. 2008, 47 (48), 9245–9248. Uzun, A.; Gates, B. C. J. Am. Chem. Soc. 2009, 131 (43), 15887–15894. Lu, J.; Serna, P.; Aydin, C.; Browning, N. D.; Gates, B. C. J. Am. Chem. Soc. 2011, 133 (40), 16186–16195.

110 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119.

Challenges with atomically dispersed supported metal catalysts Serna, P.; Gates, B. C. J. Am. Chem. Soc. 2011, 133 (13), 4714–4717. Yang, D.; Gaggioli, C. A.; Conley, E.; Babucci, M.; Gagliardi, L.; Gates, B. C. J. Catal. 2020, 382, 165–172. Yardimci, D.; Serna, P.; Gates, B. C. Chemcatchem 2012, 4 (10), 1547–1550. Szeto, K. C.; Jones, Z. R.; Merle, N.; Rios, C.; Gallo, A.; Le Quemener, F.; Delevoye, L.; Gauvin, R. M.; Scott, S. L.; Taoufik, M. Acs Catal. 2018, 8 (8), 7566–7577. Fang, C.-Y.; Zhang, S.; Hu, Y.; Vasiliu, M.; Perez-Aguilar, J. E.; Conley, E. T.; Dixon, D. A.; Chen, C.-Y.; Gates, B. C. Acs Catal. 2019, 9 (4), 3311–3321. Dessal, C.; Len, T.; Morfin, F.; Rousset, J.-L.; Aouine, M.; Afanasiev, P.; Piccolo, L. Acs Catal. 2019, 9 (6), 5752–5759. Liu, L.; Zakharov, D. N.; Arenal, R.; Concepcion, P.; Stach, E. A.; Corma, A. Nat. Commun. 2018, 9 (1), 574. Yang, M.; Liu, J.; Lee, S.; Zugic, B.; Huang, J.; Allard, L. F.; Flytzani-Stephanopoulos, M. J. Am. Chem. Soc. 2015, 137 (10), 3470–3473. Xiong, H.; Lin, S.; Goetze, J.; Pletcher, P.; Guo, H.; Kovarik, L.; Artyushkova, K.; Weckhuysen, B. M.; Datye, A. K. Angew. Chem. Int. Ed. 2017, 56 (31), 8986–8991. Ma, Y.; Chi, B.; Liu, W.; Cao, L.; Lin, Y.; Zhang, X.; Ye, X.; Wei, S.; Lu, J. Acs Catal. 2019, 9 (9), 8404–8412. Li, Q.; Chen, W.; Xiao, H.; Gong, Y.; Li, Z.; Zheng, L.; Zheng, X.; Yan, W.; Cheong, W.-C.; Shen, R.; Fu, N.; Gu, L.; Zhuang, Z.; Chen, C.; Wang, D.; Peng, Q.; Li, J.; Li, Y. Adv. Mater. 2018, 30 (25), 1800588. Zhao, Y.; Sohn, H.; Hu, B.; Niklas, J.; Poluektov, O. G.; Tian, J.; Delferro, M.; Hock, A. S. Acs Omega 2018, 3 (9), 11117–11127. Liu, P.; Zhao, Y.; Qin, R.; Mo, S.; Chen, G.; Gu, L.; Chevrier, D. M.; Zhang, P.; Guo, Q.; Zang, D. Science 2016, 352 (6287), 797–800. Babucci, M.; Fang, C.-Y.; Hoffman, A. S.; Bare, S. R.; Gates, B. C.; Uzun, A. Acs Catal. 2017, 7 (10), 6969–6972. Babucci, M.; Hoffman, A. S.; Debefve, L. M.; Kurtoglu, S. F.; Bare, S. R.; Gates, B. C.; Uzun, A. J. Catal. 2020, 387, 186–195. Xue, Z.; Qin, L.; Jiang, J.; Mu, T.; Gao, G. Phys. Chem. Chem. Phys. 2018, 20 (13), 8382–8402. Akçay, A.; Babucci, M.; Balci, V.; Uzun, A. Chem. Eng. Sci. 2015, 123, 588–595. Hayes, R.; Warr, G. G.; Atkin, R. Chem. Rev. 2015, 115 (13), 6357–6426. Hulsey, M. J.; Zhang, J.; Yan, N. Adv. Mater. 2018, 30 (47), 1802304. Dessal, C.; Sangnier, A.; Chizallet, C.; Dujardin, C.; Morfin, F.; Rousset, J. L.; Aouine, M.; Bugnet, M.; Afanasiev, P.; Piccolo, L. Nanoscale 2019, 11 (14), 6897–6904. Duan, S.; Wang, R.; Liu, J. Nanotechnology 2018, 29 (20), 204002. Kurtoglu, S. F.; Hoffman, A. S.; Akgül, D.; Babucci, M.; Aviyente, V.; Gates, B. C.; Bare, S. R.; Uzun, A. Acs Catal. 2020, 10 (21), 12354–12358. O’connor, N. J.; Jonayat, A. S. M.; Janik, M. J.; Senftle, T. P. Nature Catalysis 2018, 1 (7), 531–539. Wan, J.; Chen, W.; Jia, C.; Zheng, L.; Dong, J.; Zheng, X.; Wang, Y.; Yan, W.; Chen, C.; Peng, Q.; Wang, D.; Li, Y. Adv. Mater. 2018, 30 (11), 1705369. Qiao, B.; Liang, J.-X.; Wang, A.; Xu, C.-Q.; Li, J.; Zhang, T.; Liu, J. J. Nano Research 2015, 8 (9), 2913–2924. Zhang, J.; Wu, X.; Cheong, W. C.; Chen, W.; Lin, R.; Li, J.; Zheng, L.; Yan, W.; Gu, L.; Chen, C.; Peng, Q.; Wang, D.; Li, Y. Nat. Commun. 2018, 9 (1), 1002. Jones, J.; Xiong, H.; Delariva, A. T.; Peterson, E. J.; Pham, H.; Challa, S. R.; Qi, G.; Oh, S.; Wiebenga, M. H.; Pereira Hernández, X. I.; Wang, Y.; Datye, A. K. Science 2016, 353 (6295), 150–154. Lang, R.; Xi, W.; Liu, J. C.; Cui, Y. T.; Li, T.; Lee, A. F.; Chen, F.; Chen, Y.; Li, L.; Li, L.; Lin, J.; Miao, S.; Liu, X.; Wang, A. Q.; Wang, X.; Luo, J.; Qiao, B.; Li, J.; Zhang, T. Nat. Commun. 2019, 10 (1), 234. Nelson, N. C.; Chen, L.; Meira, D.; Kovarik, L.; Szanyi, J. Angew. Chem. Int. Ed. Engl. 2020. https://doi.org/10.1002/Anie.202007576. Wei, S.; Li, A.; Liu, J. C.; Li, Z.; Chen, W.; Gong, Y.; Zhang, Q.; Cheong, W. C.; Wang, Y.; Zheng, L.; Xiao, H.; Chen, C.; Wang, D.; Peng, Q.; Gu, L.; Han, X.; Li, J.; Li, Y. Nat Nanotechnol. 2018, 13 (9), 856–861. Qiao, B.; Wang, A.; Yang, X.; Allard, L. F.; Jiang, Z.; Cui, Y.; Liu, J.; Li, J.; Zhang, T. Nat. Chem. 2011, 3 (8), 634–641. Liu, P.; Chen, J.; Zheng, N. Chinese J. Catal. 2017, 38 (9), 1574–1580. Derita, L.; Dai, S.; Lopez-Zepeda, K.; Pham, N.; Graham, G. W.; Pan, X.; Christopher, P. J. Am. Chem. Soc. 2017, 139 (40), 14150–14165. Kwak, J. H.; Hu, J.; Mei, D.; Yi, C.-W.; Kim, D. H.; Peden, C. H. F.; Allard, L. F.; Szanyi, J. Science 2009, 325 (5948), 1670. Zhang, Z.; Zhu, Y.; Asakura, H.; Zhang, B.; Zhang, J.; Zhou, M.; Han, Y.; Tanaka, T.; Wang, A.; Zhang, T.; Yan, N. Nat. Commun. 2017, 8, 16100. Guo, X.; Fang, G.; Li, G.; Ma, H.; Fan, H.; Yu, L.; Ma, C.; Wu, X.; Deng, D.; Wei, M.; Tan, D.; Si, R.; Zhang, S.; Li, J.; Sun, L.; Tang, Z.; Pan, X.; Bao, X. Science 2014, 344 (6184), 616–619. Zhai, Y.; Pierre, D.; Si, R.; Deng, W.; Ferrin, P.; Nilekar, A. U.; Peng, G.; Herron, J. A.; Bell, D. C.; Saltsburg, H.; Mavrikakis, M.; Flytzani-Stephanopoulos, M. Science 2010, 329 (5999), 1633–1636. Yang, M.; Li, S.; Wang, Y.; Herron, J. A.; Xu, Y.; Allard, L. F.; Lee, S.; Huang, J.; Mavrikakis, M.; Flytzani-Stephanopoulos, M. Science 2014, 346 (6216), 1498–1501. Wang, H.; Dong, J.; Allard, L. F.; Lee, S.; Oh, S.; Wang, J.; Li, W.; Shen, M.; Yang, M. Appl. Catal. B Environ. 2019, 244, 327–339. Liu, L.; Diaz, U.; Arenal, R.; Agostini, G.; Concepcion, P.; Corma, A. Nat. Mater. 2017, 16 (1), 132–138. Li, Z.; Schweitzer, N. M.; League, A. B.; Bernales, V.; Peters, A. W.; Getsoian, A. B.; Wang, T. C.; Miller, J. T.; Vjunov, A.; Fulton, J. L.; Lercher, J. A.; Cramer, C. J.; Gagliardi, L.; Hupp, J. T.; Farha, O. K. J. Am. Chem. Soc. 2016, 138 (6), 1977–1982. Noh, H.; Cui, Y.; Peters, A. W.; Pahls, D. R.; Ortuno, M. A.; Vermeulen, N. A.; Cramer, C. J.; Gagliardi, L.; Hupp, J. T.; Farha, O. K. J. Am. Chem. Soc. 2016, 138 (44), 14720–14726. Wang, J.; Li, Z.; Wu, Y.; Li, Y. Adv. Mater. 2018, 30 (48), 1801649. Wang, A.; Li, J.; Zhang, T. Nat. Rev. Chem. 2018, 2 (6), 65–81. Zhou, Y.; Tao, X.; Chen, G.; Lu, R.; Wang, D.; Chen, M.-X.; Jin, E.; Yang, J.; Liang, H.-W.; Zhao, Y.; Feng, X.; Narita, A.; Müllen, K. Nat. Commun. 2020, 11 (1). Zhao, L.; Zhang, Y.; Huang, L. B.; Liu, X. Z.; Zhang, Q. H.; He, C.; Wu, Z. Y.; Zhang, L. J.; Wu, J.; Yang, W.; Gu, L.; Hu, J. S.; Wan, L. J. Nat. Commun. 2019, 10 (1), 1278. Yi, J.-D.; Xu, R.; Chai, G.-L.; Zhang, T.; Zang, K.; Nan, B.; Lin, H.; Liang, Y.-L.; Lv, J.; Luo, J.; Si, R.; Huang, Y.-B.; Cao, R. J. Mater. Chem. A 2019, 7 (3), 1252–1259. Wu, J.; Zhou, H.; Li, Q.; Chen, M.; Wan, J.; Zhang, N.; Xiong, L.; Li, S.; Xia, B. Y.; Feng, G.; Liu, M.; Huang, L. Adv. Energy Mater. 2019, 9 (22), 1900149. Han, G.; Zheng, Y.; Zhang, X.; Wang, Z.; Gong, Y.; Du, C.; Banis, M. N.; Yiu, Y.-M.; Sham, T.-K.; Gu, L.; Sun, Y.; Wang, Y.; Wang, J.; Gao, Y.; Yin, G.; Sun, X. Nano Energy 2019, 66, 104088. Li, F.; Han, G.-F.; Noh, H.-J.; Kim, S.-J.; Lu, Y.; Jeong, H. Y.; Fu, Z.; Baek, J.-B. Energy Environ. Sci. 2018, 11 (8), 2263–2269. Lin, X.; Peng, P.; Guo, J.; Xie, L.; Liu, Y.; Xiang, Z. Nano Energy 2021, 80, 105533. Xiong, Y.; Sun, W.; Xin, P.; Chen, W.; Zheng, X.; Yan, W.; Zheng, L.; Dong, J.; Zhang, J.; Wang, D.; Li, Y. Adv. Mater. 2020, 32 (34), 2000896. Cheng, Y.; He, S.; Lu, S.; Veder, J. P.; Johannessen, B.; Thomsen, L.; Saunders, M.; Becker, T.; De Marco, R.; Li, Q.; Yang, S. Z.; Jiang, S. P. Adv Sci. 2019, 6 (10), 1802066. Yi, J.-D.; Xu, R.; Wu, Q.; Zhang, T.; Zang, K.-T.; Luo, J.; Liang, Y.-L.; Huang, Y.-B.; Cao, R. Acs Energy Lett. 2018, 3 (4), 883–889. Wang, Q.; Huang, X.; Zhao, Z. L.; Wang, M.; Xiang, B.; Li, J.; Feng, Z.; Xu, H.; Gu, M. J. Am. Chem. Soc. 2020, 142 (16), 7425–7433. Babucci, M.; Sarac Oztuna, F. E.; Debefve, L. M.; Boubnov, A.; Bare, S. R.; Gates, B. C.; Unal, U.; Uzun, A. Acs Catal. 2019, 9 (11), 9905–9913. Wang, L.; Chen, M.-X.; Yan, Q.-Q.; Xu, S.-L.; Chu, S.-Q.; Chen, P.; Lin, Y.; Liang, H.-W. Sci. Adv. 2019, 5 (10). Eaax6322. Yang, H.; Shang, L.; Zhang, Q.; Shi, R.; Waterhouse, G. I. N.; Gu, L.; Zhang, T. Nat. Commun. 2019, 10 (1), 4585. Lu, P.; Yang, Y.; Yao, J.; Wang, M.; Dipazir, S.; Yuan, M.; Zhang, J.; Wang, X.; Xie, Z.; Zhang, G. Appl. Catal. B 2019, 241, 113–119. Zhao, S.; Cheng, Y.; Veder, J.-P.; Johannessen, B.; Saunders, M.; Zhang, L.; Liu, C.; Chisholm, M. F.; De Marco, R.; Liu, J.; Yang, S.-Z.; Jiang, S. P. Acs Appl. Energy Mater. 2018, 1 (10), 5286–5297. Yan, C.; Li, H.; Ye, Y.; Wu, H.; Cai, F.; Si, R.; Xiao, J.; Miao, S.; Xie, S.; Yang, F.; Li, Y.; Wang, G.; Bao, X. Energy Environ. Sci. 2018, 11 (5), 1204–1210.

Challenges with atomically dispersed supported metal catalysts

111

120. Cheng, Y.; Zhao, S.; Johannessen, B.; Veder, J. P.; Saunders, M.; Rowles, M. R.; Cheng, M.; Liu, C.; Chisholm, M. F.; De Marco, R.; Cheng, H. M.; Yang, S. Z.; Jiang, S. P. Adv. Mater. 2018, 30 (13), 1706287. 121. Liu, Y.; Zhou, Y.; Li, J.; Wang, Q.; Qin, Q.; Zhang, W.; Asakura, H.; Yan, N.; Wang, J. Appl. Catal. B 2017, 209, 679–688. 122. Cao, S.; Zhao, Y.; Lee, S.; Yang, S.; Liu, J.; Giannakakis, G.; Li, M.; Ouyang, M.; Wang, D.; Sykes, E. C. H.; Flytzani-Stephanopoulos, M. Sci. Adv. 2020, 6 (25). Eaba3809. 123. Zeng, Z.; Su, Y.; Quan, X.; Choi, W.; Zhang, G.; Liu, N.; Kim, B.; Chen, S.; Yu, H.; Zhang, S. Nano Energy 2020, 69, 104409. 124. Kunwar, D.; Zhou, S.; Delariva, A.; Peterson, E. J.; Xiong, H.; Pereira-Hernández, X. I.; Purdy, S. C.; Ter Veen, R.; Brongersma, H. H.; Miller, J. T.; Hashiguchi, H.; Kovarik, L.; Lin, S.; Guo, H.; Wang, Y.; Datye, A. K. Acs Catal. 2019, 9 (5), 3978–3990. 125. Zhu, Y.; Cao, T.; Cao, C.; Luo, J.; Chen, W.; Zheng, L.; Dong, J.; Zhang, J.; Han, Y.; Li, Z.; Chen, C.; Peng, Q.; Wang, D.; Li, Y. Acs Catal. 2018, 8 (11), 10004–10011. 126. Li, T.; Liu, J.; Song, Y.; Wang, F. Acs Catal. 2018, 8 (9), 8450–8458. 127. Li, H.; Wang, L.; Dai, Y.; Pu, Z.; Lao, Z.; Chen, Y.; Wang, M.; Zheng, X.; Zhu, J.; Zhang, W.; Si, R.; Ma, C.; Zeng, J. Nat. Nanotechnol. 2018, 13 (5), 411–417. 128. Choi, C. H.; Kim, M.; Kwon, H. C.; Cho, S. J.; Yun, S.; Kim, H. T.; Mayrhofer, K. J.; Kim, H.; Choi, M. Nat. Commun. 2016, 7, 10922. 129. Tian, J.; Yang, D.; Wen, J.; Filatov, A. S.; Liu, Y.; Lei, A.; Lin, X. M. Nanoscale 2018, 10 (3), 1047–1055. 130. Wang, Z.; Xu, S. M.; Xu, Y.; Tan, L.; Wang, X.; Zhao, Y.; Duan, H.; Song, Y. F. Chem. Sci. 2019, 10 (2), 378–384. 131. Li, J.; Chen, S.; Yang, N.; Deng, M.; Ibraheem, S.; Deng, J.; Li, J.; Li, L.; Wei, Z. Angew. Chem. Int. Ed. 2019, 58 (21), 7035–7039.

6.06

Metal containing nanoclusters in zeolites

Guanna Li and Evgeny A. Pidkoc, a Biobased Chemistry and Technology, Wageningen University & Research, Wageningen, The Netherlands; b Laboratory of Organic Chemistry, Wageningen University & Research, Wageningen, The Netherlands; and c Inorganic Systems Engineering group, Department of Chemical Engineering, Faculty of Applied Sciences, Delft University of Technology, Delft, The Netherlands a,b

© 2023 Elsevier Ltd. All rights reserved.

6.06.1 6.06.2 6.06.2.1 6.06.2.2 6.06.3 6.06.3.1 6.06.3.2 6.06.3.3 6.06.3.4 6.06.4 6.06.4.1 6.06.4.2 6.06.4.3 6.06.5 6.06.5.1 6.06.5.2 6.06.5.3 6.06.6 Acknowledgment References

Introduction Synthesis Encapsulation of nanoclusters in zeolite Isolated single metal atom sites in zeolites Advanced characterization techniques for zeolite encapsulated metal species Electron microscopy X-ray absorption spectroscopy Vibrational spectroscopy Solid-state nuclear magnetic resonance Catalytic applications C1 molecules conversion Active site cooperation and multifunctionality in confined space Confined space for selectivity control Computational modeling Structure prediction by operando thermodynamic analysis Reactivity scaling relationship and beyond Micro-kinetic modeling and dynamics Conclusion and perspective

112 113 113 117 120 121 122 126 129 131 131 132 135 136 136 136 139 141 141 141

Abstract The molecular-sized void space of the zeolitic micropores is perfect matrices to encapsulate and stabilize multicomponent and multifunctional complexes that can be used as active sites for a wide range of important catalytic transformations. In this article, we discuss and analyze the key developments of the last decade in the catalytic chemistry of metal-containing nanoclusters confined in zeolite micropores. We will present a concise summary of the recent developments in the tailored synthesis strategies, the advanced in-situ and operando characterization techniques, the enhanced performances of zeolite stabilized nanoclusters in various catalytic processes, and the application of computational modeling approaches for addressing the puzzle of catalyst-reactivity relationships. The article will be concluded with a brief discussion on the perspective for future developments anticipated for this field.

6.06.1

Introduction

Zeolites are a versatile class of inorganic microporous materials with a wide range of research and industrial applications in catalysis, ion-exchange, adsorption, and separation technology. In modern catalysis, zeolites represent one of the most important groups of robust supports and solid catalysts for a wide range of industrial catalytic processes, due to their unique physiochemical properties that persist under harsh reaction conditions (acidic or basic system, high temperature, and pressure) and the high flexibility of finely tunable functionalities by the target-driven design of their textural and chemical characteristics. The widespread application and the central position of zeolites in the field of heterogeneous catalysis attribute not only to their microporous structures contributing to the unique shape selectivity but also to the possibility of introducing and manipulating the active components confined within the micropores contributing thus with the nanoconfinement, cooperativity and unique electronic effects as the additional reactivity gears to tailor the activity and selectivity of the resulting composite catalysts.1 The molecular-sized void space of the zeolitic micropores are particularly suitable to encapsulate and stabilize multicomponent and multifunctional complexes2,3 that can be used as the active sites for a wide range of important catalytic transformations such as selective biomass conversion,4,5 valorization of natural gas6–8 and CO29 recycling as well as petrochemical conversion10 and selective catalytic reduction of NOx.11,12 In this article, we discuss and analyze the key developments of the last decade in the catalytic chemistry of metal-containing nanoclusters confined in zeolite micropores. We will present a concise summary of the recent developments in the tailored synthesis

112

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00022-4

Metal containing nanoclusters in zeolites

113

strategies, the advanced in-situ and operando characterization techniques, the enhanced performances of zeolite stabilized nanoclusters in various catalytic processes, and the application of computational modeling approaches for addressing the puzzle of catalyst-performance relationships. The article will be concluded with a brief discussion on the perspective for future developments anticipated for this field.

6.06.2

Synthesis

Zeolites are three-dimensional microporous crystalline materials assembled by corned-shared tetrahedral TO4 building blocks, with T atoms usually being Si or Al. A large variety of zeolite structures can be synthesized by tuning the synthesis parameters such as the composition of the precursors, the types of the structure-directing agents, or the crystallization conditions. The micropores of the resulting solids can be used to encapsulate isolated metal sites, nanoclusters, and nanoparticles in local micro-environments with nanoconfinement effects.13 The size of the encapsulated species is one of the crucial factors governing their distinct chemical properties and catalytic performance (Fig. 1). A reduction of the size usually contributes to a higher reactivity due to the increase of the fraction of the coordination-unsaturated metal sites in smaller metal entities. Furthermore, due to the quantum size effects, subtle changes in the size or geometry of a molecular- or nanosized cluster can give rise to substantial changes in their physicochemical characteristics and, accordingly, their catalytic reactivity.14–17 Therefore, much effort has been put into the improved control over the formation of cluster species inside the zeolite micropores. Below, we review the state-of-the-art synthesis strategies for encapsulating metal-containing nanoclusters and even single atom sites inside the zeolite matrices.

6.06.2.1

Encapsulation of nanoclusters in zeolite

Small metal particles often show high activity in many types of reactions due to the coordination-unsaturated metal centers, high facet exposure, and large surface-to-volume ratios. However, one of the biggest challenges is the low stability of ultra-small metal clusters and particles. Due to the high surface energy, they tend to aggregate into big nanoparticles (> 5 nm) when dispersed on open surfaces, resulting in decreased catalytic activity.18,19 The formation and stabilization of the metal nanoparticles inside the zeolite microspores is one of the practical solutions to circumvent such aggregation process. Furthermore, the zeolite encapsulation strategy allows the tailored formation of multifunctional reactive environments with uniquely tunable catalytic properties. The rigid and robust zeolite framework provides the unique stabilizing micro-environment that not only prevents sintering of the metal nanoclusters but also defines their size, shape, and electronic properties.20–22 Post synthetic protocols such as ion-exchange, impregnation, and chemical vapor deposition have been widely applied to encapsulate metal clusters within large pores of the zeolite crystals. The ion-exchange method is one of the most widely used approaches to incorporate metal species into the zeolites with controlled exchange numbers. In this method, the proton or alkali metal cations compensating the framework charge imbalance induced by lattice Al sites in the parent material are being replaced by other metal species through the ion-exchange reaction in an aqueous solution. However, in practice, these conventional techniques often give rise to heterogeneous metal speciation with the metal being deposited both in the micropores and on the external surface of the zeolite crystals. Post-synthetic reduction or calcination treatments are usually used to mobilize the deposited species and facilitate their diffusion to the extraframework sites inside the zeolite channels. Such common strategies have been widely used in zeolite chemistry and catalysis to prepare materials containing extraframework species such as Ga, Fe, Mo, etc. By redox treatment, the

Fig. 1 Noble metal particle size effect in zeolite catalysis. (A) The change of surface free energy and specific activity with the size of noble metal particles on the zeolite support; (B) The comparison of TOF values of H2 formation upon the dehydrogenation of formic acid–sodium formate (1:1) at 25
C (black) and 50
C (red) with Pd nanoparticles of different average size on either carbon (Pd/C) or S-1 zeolite (Pd/S-1) (im ¼ impregnation, in ¼ in situ encapsulation). Insets present the TEM images and corresponding size distributions of Pd clusters (scale bars: 20 nm). Adopted with permission from Wu, S.-M.; Yang, X.-Y.; Janiak, C. Confinement Effects in Zeolite-Confined Noble Metals. Angew. Chem. Int. Ed. 2019, 58(36), 12340–12354. Copyright 2019, Wiley.

114

Metal containing nanoclusters in zeolites

transition metal species can give rise to various oxide, sulfide, and carbide clusters. Integrated strategies for engineering such metaloxo, metal-sulfide, and metal-carbide nanoclusters in zeolite matrices have been recently reviewed by Kosinov et al. in Ref. 23. However, for small-pore zeolites, these methods often show low efficiency because of the limited accessibility of the narrowed channels by the solvated or gaseous metal precursors. Moreover, the ion-exchange and impregnation in aqueous solutions may give rise to heterogenous particle size distribution and poor catalytic activity. In the text below we will rather focus on alternative and modified strategies that allow for improved control of cluster speciation. Several new strategies combining the conventional ion-exchange or impregnation approaches with more advanced synthetic methodologies have been introduced for the synthesis of complex inorganic architectures confined in the zeolite pores. Zhang et al. combined the ion-exchange with the layer reassembling process to encapsulate metal nanoclusters in zeolites.24 The flexible distance of the layered zeolite intermediates offers opportunities for the direct introduction of metal precursors during the zeolite synthesis. This approach has been used to selectively introduce Pd particles in the FER zeolite matrix. First, FER layers were swelled into their layered precursors with the aid of cetyltrimethylammonium (CTAþ) surfactant. Then, the CTAþ cations were exchanged by Pd precursors. The topotactic transformation from the layered precursor to the 3-dimensional FER framework resulted in the effective encapsulation of the metal species, which were transformed to well-defined Pd clusters of ca. 1.4 nm by subsequent calcination and reduction treatments. Román-Leshkov and co-workers combined ion-exchange with selective demetalation to prepare bimetallic nanoclusters within the micropores of MFI zeolite.25 Zn-containing MFI zeolite (Zn-MFI) was firstly synthesized by the hydrothermal method. Then Pd2þ ions were introduced into the pores by ion-exchange to form Pd@Zn-MFI, taking the advantages of the charge imbalance of the Zn-containing framework. Finally, by temperature-programmed calcination and reduction treatments, the PdZnx bimetallic nanoclusters were generated within the MFI channels by alloying of Pd with Zn ions extracted from the framework. The procedure yielded bimetallic PdZnx nanoclusters of ca. 1 nm highly active in selective hydrogenation of halogenated nitroarenes. Moliner et al. complemented the incipient wetness impregnation with the atom-trapping technique (Scheme 1) to prepare

Scheme 1 Atom-trapping methodology when the AleCHA is mixed with a Pt-containing SiO2 or Al2O3 support. Reproduced with permission from Moliner, M.; Gabay, J.; Kliewer, C.; Serna, P.; Corma, A., Trapping of Metal Atoms and Metal Clusters by Chabazite under Severe Redox Stress. ACS Catal. 2018, 8(10), 9520–9528. Copyright 2018, American Chemical Society.

a series of sintering-resistant Pt, Pd, and bimetallic PtePd nanoparticles inside non-reducible Al-containing CHA zeolite. The method involves the selective transfer of weakly bound MOx precursors from open SiO2 or Al2O3 supports into the zeolite channels upon treatment at 650
C in the O2 atmosphere. The high affinity of Al-containing CHA zeolite to trap and stabilize noble singlemetal atoms and metal clusters is the driving force for this process.26 The direct encapsulation of metal precursors during the hydrothermal zeolite synthesis is a promising strategy for the modification of small-pore zeolites and to improve the particle size distribution. In this method, the zeolite framework is assembled around the metal precursors. Unfortunately, this promising methodology fails often when applied to common transition metal precursors, because of their rapid precipitation asynchronous with the slow hydrothermal zeolite crystallization under the highly basic synthesis conditions. To overcome this problem, Iglesia and co-workers developed a general and versatile synthetic strategy using bifunctional organosilane ligands that enabled encapsulation of different metal clusters (Pt, Pd, Ir, Rh, and Ag) within the small pores of NaA zeolite. Modification of this zeolite type could not be achieved by the post-synthetic approaches. The authors have shown that the bifunctional mercaptosilane protecting ligands prevent the metal precursors from the precipitation and promote the condensation of silicate structures around the ligated precursors (Fig. 2A).27 This strategy was further expanded to other small-pore zeolites such as SOD and GIS.28 Furthermore, the same group developed a robust low-temperature hydrothermal direct synthesis of MFI zeolites modified with uniform 1.3–1.7 nm Pt, Ru, and Rh nanoparticles via interzeolite transformation. In this method, the large-pore zeolites of BEA and FAU containing metal nanoparticles within their microporous voids are synthesized by the direct hydrothermal routes. Subsequent recrystallization of such parent zeolite precursors yields the modified MFI daughter material with a higher framework density without the loss of encapsulation (Fig. 2B). The interzeolite transformation involves the nucleation and crystallization of the MFI without using any organic structure-directing agents, which represents a potentially more economical and environmentally favorable route compared to the conventional procedures.29 These methods require that the zeolite synthesis and encapsulation are carried out simultaneously under mild template-free crystallization conditions. As a result, they are normally used for the encapsulation of metal clusters into zeolites with a relatively low Si to Al ratio (Si/Al < 3) resulting in low hydrothermal stability of the resulting composites.

Metal containing nanoclusters in zeolites

115

Fig. 2 (A) Schematic process for mercaptosilane-assisted metal encapsulation during zeolite crystallization. (B) Encapsulation of metal clusters within MFI by exchanging cationic metal precursors into a parent zeolite (BEA, FAU) and transforming these zeolites into daughter structures of higher framework density (MFI) under hydrothermal conditions. (A) Reproduced with permission from Choi, M.; Wu, Z.; Iglesia, E. MercaptosilaneAssisted Synthesis of Metal Clusters within Zeolites and Catalytic Consequences of Encapsulation. J. Am. Chem. Soc. 2010, 132(26), 9129–9137. Copyright 2010, American Chemical Society. (B) Reproduced with permission from Goel, S.; Zones, S. I.; Iglesia, E. Encapsulation of Metal Clusters within MFI Via Interzeolite Transformations and Direct Hydrothermal Syntheses and Catalytic Consequences of their Confinement. J. Am. Chem. Soc. 2014, 136(43), 15280–15290. Copyright 2014, American Chemical Society.

To overcome this limit of low hydrothermal stability, new strategies were developed to encapsulate noble metals into high-silica zeolite matrices. Corma and co-workers reported a method for incorporating Pt metal clusters and single atoms in thermally stable high-silica zeolites. Purely siliceous layered MWW zeolite precursors were first prepared via the conventional hydrothermal route and expanded by the organic surfactant of hexadecyltrimethylammonium (CTMACþ OH). Then, the sub-nanometer Pt species were introduced into the interlayer space. At the next step, the removal of the organic compounds and the formation of the 3D zeolite structure upon high-temperature calcination 540
C yielded sub-nanometer-sized Pt clusters encapsulated in zeolite cages of the 3D MCM-22 (Fig. 3A and B).30 More recently, the same group reported another approach to regioselectively encapsulate sub-nanometer Pt species into purely siliceous MFI zeolite by a one-pot synthesis strategy. The location of the Pt nanocluster was specified by limiting the access of the Pt species to the sinusoidal channels with the intersectional void occupied by the tetrapropylammonium hydroxide (TPAþ OH) template molecules.31 A similar direct encapsulation approach was also employed to highly selectively deposit Ir clusters into the 10 membered rings connecting the two neighboring 12 membered ring supercages of the MWW zeolite.32 Xu and co-workers developed a cationic polymer-assisted synthesis allowing the encapsulation of Pt nanoparticles into the micropores of both siliceous and aluminosilicate MFI zeolites.33 Their method was based on the use of polydiallydimethylammonium chloride template, which limits the mobility of PtCl62 precursors and at the same time promotes the

116 Metal containing nanoclusters in zeolites

Fig. 3 (A) Encapsulation of Pt nanoclusters during the transformation of a 2D MWW zeolite precursors into daughter 3D MCM-22 structure. (B) HAADF-HRSTEM image of Pt@MCM-22. (C) Cationic polymer-assisted synthetic strategy to encapsulate Pt nanoclusters into MFI zeolite. (D) High-resolution TEM image of Pt@HZSM-5. (B) Reproduced with permission from Liu, L.; Diaz, U.; Arenal, R.; Agostini, G.; Concepcion, P.; Corma, A. Generation of Subnanometric Platinum with High Stability during Transformation of a 2D Zeolite into 3D. Nat. Mater. 2017, 16(1), 132–138. Copyright 2017, Springer Nature. (D) Reproduced with permission from Cho, H. J.; Kim, D.; Li, J.; Su, D.; Xu, B. Zeolite-Encapsulated Pt Nanoparticles for Tandem Catalysis. J. Am. Chem. Soc. 2018, 140(41), 13514–13520. Copyright 2018, American Chemical Society.

Metal containing nanoclusters in zeolites

117

encapsulation of metal precursors during zeolite crystallization in an alkaline environment. The driving force for the selective encapsulation is the electrostatic interactions with both the anionic metal precursor and the silicate or aluminosilicate building blocks of the zeolites in the synthesis gel (Fig. 3C and D). Yu et al. developed a facile and direct hydrothermal approach and prepared ultrasmall Pd nanoclusters with a uniform particle size ca. 1.8 nm well dispersed at the channel intersections of the MFI structure. The nanoclusters were encapsulated by in-situ confinement within nanosized silicalite-1 (S-1) zeolite using [Pd-(NH2CH2CH2NH2)2]Cl2 as a metal precursor and TPAþ OH as a template. The preferential localization of the TPAþ template ions in the middle of the straight channel resulted in the selective deposition of the Pd nanoclusters at the intersection sites. Such a selective encapsulation of nanoparticles resulted in their enhanced resistance to sintering and stability toward high-temperature redox and hydrothermal treatments.34 Later, the same group also successfully synthesized ultra-small bimetallic PdeM(OH)2 (M ¼ Ni, Co) clusters encapsulated within purely siliceous zeolites with superior thermal stability and high activity for formic acid dehydrogenation.35 Subnanometer bimetallic PtZn cluster was also confined into the micropores of S-1 zeolite by the similar direct hydrothermal crystallization. The introduction of Zn remarkably improved the stability of confined Pt clusters. The resulting PtZn@S-1 catalyst exhibited excellent activity and selectivity in the propane dehydrogenation reaction.36 Zhu et al. reported the ultrafast in-situ encapsulation of metal nanoclusters into MFI zeolite during its crystallization. With this method, Pt@Sn-MFI materials were synthesized within just 5 min. The ultrafast route for zeolite synthesis allows circumventing the mismatch in the time-scales of the precipitation of metal nanoclusters and the crystallization of zeolite matrix (Fig. 4).37 Xiao’s group developed another generic solvent-free strategy by mixing, grinding, and heating solid raw materials to synthesize bimetallic AuPd nanoparticles within the micropore of S-1 zeolite showing a very high efficiency of noble metal utilization. This method is viewed as a greener and more environmentally benign alternative to the conventional hydrothermal methods as it simultaneously eliminates the need of the organic templates and water solvent during the synthesis as well as the subsequent hightemperature calcination step commonly required to remove the template molecules from the zeolite pores. (Fig. 5A).38 The metal nanoparticles are immobilized onto amorphous silica, which is then crystallized into a zeolite framework in the solid phase. The efficiency and yield of the zeolite product are remarkably enhanced by such a simple synthesis protocol.39–42 It is shown that more than 96% of gold and palladium could be successfully loaded into the zeolite, which is much higher than the 36% encapsulation efficiency obtained in the hydrothermal process. Catalytic test revealed a high activity, selectivity, and stability of Au-Pd@S-1 catalyst in the aerobic oxidation of bioethanol even in the presence of 90% water.43 This solvent-free method was later expanded to the encapsulation of monometallic Pd nanoparticles in S-1. The thus formed Pd@S-1 material catalyzes furfural hydrodeoxygenation with furan selectivity of 99% at 91% conversion, which is superior to the values obtained with the conventionally prepared supported Pd/S-1 catalyst.44 The same group reported a seed-directed organic template-free synthesis route toward Pd@BEA materials. The Pd nanoparticles supported on BEA zeolite nanocrystals function as zeolite seeds in the synthesis gel for the growth of new BEA zeolite crystals so that the Pd nanoparticles were encapsulated inside the extended BEA zeolite structure (Pd@BEA).45,46 The crystallization process effectively controls the size and stability of the metal nanoclusters. During the synthesis, an amorphous aluminosilicate gel covers and protects the metal nanoparticles deposited on the zeolite seeds. This results in a crystalline zeolite sheath that prevents sintering and stabilizes the nanoparticles even during catalytic operation under severe conditions (Fig. 5B–J).47 The sizes of the metal nanoparticles in the final metal@zeolite catalysts can be adjusted by fine tuning the particle size in the initial seed precursor. Such a method was generalized to synthesize a variety of transition metal-loaded silica and aluminosilicate zeolite catalysts such as Pd@S-1, Pt@BEA, Pt@MOR, Rh@BEA, Rh@MOR, Ag@BEA, Ag@MOR, and [email protected],47 An analogous seed-directed solvent-free method was also employed for the encapsulation of Au nanoclusters in TS-1 zeolite. The encapsulated Au nanoparticles are anchored at the framework Ti sites resulting in the AueTi synergy.48

6.06.2.2

Isolated single metal atom sites in zeolites

Single metal atom catalysis has been emerging as a new frontier in catalysis science and technology since 2011.49–51 Theoretically, isolated single-atom site (SAS) catalysts have the maximum atom-utilization efficiency and offer uniquely tunable catalytic properties combining the benefits of both heterogeneous and homogeneous catalysis. Therefore, the selective generation of SASs with high activity and stability is of great importance from the academic and industrial points of view.52 Besides, SASs can be used as the ideal model systems for the investigation of the structure-activity relationships of various practical catalytic systems.7,53,54 Stabilization of the active phase in the single-atom state and prevention of their agglomeration is one of the key challenges in the field.55 The aggregation of the active metal sites into bulk-like clusters or leaching into reaction solution decrease the recyclability and catalytic performance of such catalysts. Among different support materials, zeolites were identified as promising nano-containers for the stabilization of SASs. An important advantage over alternative support materials is the unique shape selectivity and defined multifunctional reaction environment provided by the zeolite channels encapsulating SASs.56 Pioneering works by Gates and co-workers have introduced the view on the zeolite framework as a macro-ligand providing defined donor sites with defined steric constraints and variable basicity capable of coordinating and stabilizing various transition metal species. Well-defined transition metal species have been incorporated into large-pore FAU-type Y zeolite via the formal ligand exchange reaction of the M(L)2(acac)x (M ¼ Rh,57,58 Ru,59 Ir,60,61 Au,62 L ¼ CO, C2H4, CH3, acac ¼ acetylacetonate, and x ¼ 1 or 2) precursors with the zeolite Brønsted acid sites.

118

Metal containing nanoclusters in zeolites

Fig. 4 (A) The mismatching between metal precipitation and zeolite crystallization. (B) The strategies for making metal precipitation and zeolite crystallization to match in time scale. (C) Scheme of the proposed ultrafast encapsulation. (D) X-ray diffraction patterns of the Pt@Sn-ZSM-5 samples synthesized at 190
C for different times. (E) and (F) HAADF-STEM image and corresponding Pt/Sn particle size distribution of the Pt@Sn-ZSM-5 zeolite synthesized at 190
C for 5 min. (G) Relative crystallinity and compositional evolution of the Pt@Sn-ZSM-5 samples synthesized over different periods. (H) UV–Vis spectra of the Pt@Sn-ZSM-5 samples synthesized over different times. Reproduced with permission from Zhu, J.; Osuga, R.; Ishikawa, R.; Shibata, N.; Ikuhara, Y.; Kondo, J. N.; Ogura, M.; Yu, J. H.; Wakihara, T.; Liu, Z. D.; Okubo, T. Ultrafast Encapsulation of Metal Nanoclusters into MFI Zeolite in the Course of its Crystallization: Catalytic Application for Propane Dehydrogenation. Angew. Chem. Int. Ed. 2020, 59, 19669–19674. Copyright 2020, Wiley.

Metal containing nanoclusters in zeolites

119

Fig. 5 (A) Synthesis of Pd@S-1 by solvent-free procedure. (B) Synthesis of Pd@Beta via the seed-directed route. (C) Proposed model, (D) STEM image, (E) HR-TEM, (F) tomographic image of Pt@BEA, (G) proposed model, (H) HR-TEM tomographic image of Pt@BEA after calcination at 600
C for 240 min, (I) TEM tomographic image of Pt/BEA. (J) TEM image of Pt/BEA after calcination at 600
C for 240 min. (A) Reproduced with permission from Wang, C.; Wang, L.; Zhang, J.; Wang, H.; Lewis, J. P.; Xiao, F.-S. Product Selectivity Controlled by Zeolite Crystals in Biomass Hydrogenation over a Palladium Catalyst. J. Am. Chem. Soc. 2016, 138(25), 7880–7883. Copyright 2016, American Chemical Society. (H) Reproduced with permission from Zhang, J.; Wang, L.; Zhang, B. S.; Zhao, H. S.; Kolb, U.; Zhu, Y. H.; Liu, L. M.; Han, Y.; Wang, G. X.; Wang, C. T.; Su, D. S.; Gates, B. C.; Xiao, F. S. Sinter-Resistant Metal Nanoparticle Catalysts Achieved by Immobilization Within Zeolite Crystals Via Seed-Directed Growth. Nat. Catal. 2018, 1(7), 540–546. Copyright 2018, Springer Nature.

More conventional catalyst synthesis methods such as aqueous ion-exchange and incipient wetness impregnation can also be used to introduce isolated SASs in zeolite matrices. For instance, Fe, Cu, Ni, Zn, Ga can be selectively exchanged into the sixmembered rings of Beta, SSZ-13, CHA, and ZSM-5 zeolites.6,63–66 Pt@LTL, Rh@ZSM-5, and InH2@CHA zeolites with confined single metal sites were also fabricated by ion-exchange followed by oxidation or reduction treatment, which show significant reactivity for CO oxidation, methane transformation, and nonoxidative dehydrogenation of ethane, respectively.67–69 The ion-exchange method is intrinsically limited by the ion-exchange capacity of the zeolite host and the solubility of the metal precursors. The impregnation appears therefore a more versatile technique for encapsulating metals in the zeolite matrices. Shan and co-workers employed incipient wetness impregnation (IWI) to synthesize isolated Rh sites in the zeolite of ZSM-5.7 A similar approach was used for the selective synthesis of Pt(II)eO(OH)x single site in LTL and isolated WO4 species entrapped in USY zeolite.70,71 Yang and co-workers synthesized single site AueO(OH)xe(Na/K) mononuclear species inside the LTL zeolite by IWI followed by high-temperature calcination treatment.72 The presence of alkali cations enhanced the stability of the supported Au species similar to the mechanism observed for other single-atom-type catalyst systems.73,74 Such alkali-stabilized single Au sites exhibited catalytic activities similar to the more conventional ceria and titania-supported single-atom gold catalysts. In analogy to nanoparticle encapsulation described in the previous section, the direct in situ immobilization of single metal atom sites during the zeolite synthesis has also been reported. Li and co-workers developed a general strategy to prepare zeolite Y-based catalysts containing such SASs as Pt, Pd, Ru, Rh, Co, Ni, and Cu.75 Transition metal ethanediamine (EDA) complexes could be selectively dispersed among the b-cages of zeolite Y during the crystallization process. Subsequent thermal reduction treatment decomposes the confined M-EDA precursors and anchors the metal sites at the 6-membered rings making them accessible to reagents from the zeolite supercages (Fig. 6A).

120

Metal containing nanoclusters in zeolites

(A)

Pt-ISAS @ Y zeolite

Pt Precursor @ Y zeolite

In situ

Air and H2

Assembly

Thermal Treatment

Pt Precursor

E - cage

Pt Si/Al O N O

SiO4/AlO4

(C)

(D)

(B) N

N

Ru

N

O O

Si O

N

N N

O

5 nm

723 K

443 K N

Synthesis Gel

(E)

2 nm

(F)

Vacuum decomposition

Hydrothermal reaction

Ru(en)3 @S-1

Ru SAs/S-1 5 nm

2 nm

Fig. 6 (A) In situ separation and confinement of a platinum precursor in a b-cage of Y zeolite followed by thermal treatment. (B) Synthesis procedure of Ru single atom in S-1. (C and D) AC HAADF-STEM images of Ru SASs/S-1. (E and F) AC HAADF-STEM images of Ru SASs/S-1-used. Reproduced with permission from Liu, Y.; Li, Z.; Yu, Q.; Chen, Y.; Chai, Z.; Zhao, G.; Liu, S.; Cheong, W.-C.; Pan, Y.; Zhang, Q.; Gu, L.; Zheng, L.; Wang, Y.; Lu, Y.; Wang, D.; Chen, C.; Peng, Q.; Liu, Y.; Liu, L.; Chen, J.; Li, Y. A General Strategy for Fabricating Isolated Single Metal Atomic Site Catalysts in Y Zeolite. J. Am. Chem. Soc. 2019, 141(23), 9305–9311; Qiu, J. Z.; Hu, J. B.; Lan, J. G.; Wang, L. F.; Fu, G. Y.; Xiao, R. J.; Ge, B. H.; Jiang, J. X. Pure Siliceous Zeolite-Supported Ru Single-Atom Active Sites for Ammonia Synthesis. Chem. Mater. 2019, 31(22), 9413–9421. Copyright 2019, American Chemical Society.

Yu and co-workers encapsulated single Rh atoms within S-1 and aluminosilicate ZSM-5 zeolites by using one-pot hydrothermal synthesis followed by ligand-protected direct hydrogen reduction.76 Purely siliceous S-1 supported Ru SASs were also successfully synthesized by the one-pot hydrothermal method. The mononuclear [Ru(NH2CH2CH2NH2)3]Cl3 metal precursor was successfully confined within the S-1 under hydrothermal zeolite synthesis conditions at 443 K. Subsequent vacuum treatment at 723 K for 12 h resulted in the removal of the organic structure-directing agents and ligands to generate Ru SASs inside the zeolite pores (Fig. 6B–F). The applicability of this methodology for the functionalization of other zeolite topologies with various noble metal SASs has been proposed.77

6.06.3

Advanced characterization techniques for zeolite encapsulated metal species

To confirm the successful encapsulation and the well-defined characteristics of the clusters and single-atom sites inside the zeolite pores, their comprehensive characterization is necessary. Most standard bulk characterization techniques fail to address the complexity of the metal-zeolite composite materials because of the relatively low metal concentration and intrinsic heterogeneity of the solid materials. To overcome the limitations of the individual techniques, the complementary use of several characterization tools such as electron microscopy, X-ray absorption spectroscopy, nuclear magnetic resonance (NMR), vibrational and UV–Vis spectroscopies aided by computational modeling is often practiced to identify the size, distribution, structure, and electronic properties of such active metal centers encapsulated inside zeolite materials. Such structural insights are also required to understand the mechanism of catalytic transformations by such materials and support the rational design and optimization of the zeolite-based catalysts. However, due to the structural flexibility, nanoclusters and single atom metal species can exhibit significant structural changes when interacting with reactants, complicating thus the determination of the active site structure under the catalyst working conditions. The combination of comprehensive ex-situ, in-situ, and operando characterization techniques provides insights into the physiochemical properties of metal-zeolite composites and helps to elucidate the structure-reactivity relationships to guide the engineering of superior zeolite catalyst.78

Metal containing nanoclusters in zeolites 6.06.3.1

121

Electron microscopy

To determine the location and structure of the confined metal nanoparticles, high-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM) is typically used, which provides a high resolution and contrast between the heavy metal guest atoms and the light atoms of the zeolite framework. Gates and co-workers have pioneered the application of a damage-reduced aberration-corrected scanning transmission electron microscopy (STEM) in combination with infrared and extended X-ray absorption fine structure (EXAFS) spectroscopies to unravel the locations of the single metal atoms and small metal nanoclusters, such as Ir, Au, La, Os, Rh, inside zeolite micropores.79–81 For instance, they identified the location and structure of hexairidium clusters confined in the NaY supercages by atomic-resolution (HAADF) aberration-corrected STEM with a high signal-to-noise ratio. The diameter of the cluster was precisely measured and the preferred locations of the Ir6 cluster were identified (25%) (Fig. 7).82

Fig. 7 (A) Unprocessed aberration-corrected HAADF-STEM image of zeolite NaY containing 10 wt% iridium after decarbonylation of clusters, showing the zeolite framework in the (110) direction, and a histogram showing the distribution of the nanocluster diameters. (B) A magnified view of the dashed-rectangular area in (A) containing the two Ir6 clusters encircled. A cluster at a T5 site is represented in red, and a cluster at a T6 site is represented in blue. Simulations of the FAU in the (110) projection are superimposed on the experimental image indicating Ir6 clusters. The arrows indicate the relative orientations of the T5 and T6 sites with respect to the (110) direction of the zeolite crystal. (C) Simulated FAU models in the (111) projection corresponding to the location of the two Ir6 clusters shown in (B). (D) Frequencies of Ir6 clusters located in T5 and T6 sites. (E) Fast Fourier transform of the experimental image shown in (A). (F) Theoretical diffraction pattern of zeolite NaY in the (110) direction. Reproduced with permission from Aydin, C.; Lu, J.; Shirai, M.; Browning, N. D.; Gates, B. C. Ir-6 Clusters Compartmentalized in the Supercages of Zeolite NaY: Direct Imaging of a Catalyst with Aberration-Corrected Scanning Transmission Electron Microscopy. ACS Catal. 2011, 1(11), 1613–1620. Copyright 2011, American Chemical Society.

122

Metal containing nanoclusters in zeolites

Furthermore, the aggregation of the isolated Ir atoms into nanoclusters in the 14-ring channels of zeolite SSZ-13 was also tracked and directly visualized. The authors proposed the autocatalytic agglomeration and sintering mechanism based on the electron microscopy data (Fig. 7).83 The location of the mononuclear gold complexes in the zeolite pores was also confirmed by the same technique. These gold species were identified to be the catalytic active sites for CO oxidation at 298 K and 1 bar.62 The presence of oxygen-bridged La pairs with LaeLa distance of ca. 3.84 Å in zeolite Y cages was observed consistent with the proposal based on X-ray photoelectron spectroscopy81 and earlier periodic DFT studies.84 STEM was also used to identify the location and structure of other metal clusters such as Ag, Cd, and Zn in different zeolites.85–89 Both light alkali cations and ion-exchanged rare earth metals (Eu, Tb, and Gd) could be unambiguously detected by aberrationcorrected STEM with atomic resolution, providing unique information about the structure and location of both metal and zeolite support.90 Very recently, Mayoral and co-workers reported the atomic-level observations of all framework atoms including heteroatom Fe and extraframework Na cations by spherical aberration-corrected STEM coupled with annular dark field (ADF) and annular bright-field (ABF) image modes, which have brought the sensitivity to single heteroatoms or point defects with extremely low density (Fig. 8).91 Corma et al. developed a strategy to characterize precisely the location of sub-nanometer Pt and Sn species in MFI zeolite by STEM technique that records simultaneously the high-angle annular dark-field (HAADF) images for heavy metals and integrated differential phase-contrast (iDPC) images for the zeolite framework in the same area.92,93 By combining with other spectroscopies this approach could provide essential information about the spatial distribution of the metal species encapsulated in zeolites and also reveal the locations of framework Al sites. Such newly developed technique has been successfully applied to in-situ imaging with atomic resolution of the structural changes of the zeolite framework induced by adsorption of aromatic molecules.94 TEM tomography was also developed to provide 3-dimensional quantitative descriptions of the size, distribution, and interparticle distance of the metal particles within individual zeolite crystals.19,95,96 For instance, Pt nanoclusters as small as 1 nm can be visualized by electron tomography and image analysis.19

6.06.3.2

X-ray absorption spectroscopy

Synchrotron X-ray absorption spectroscopy (XAS) is a powerful and widely used characterization technique to study the structure and electronic properties of zeolite-supported metal catalysts both ex-situ and under catalytic working conditions. There are two spectral regions that carry different kinds of information. The X-ray absorption near-edge structure (XANES) provides an insight into the electronic structure and oxidation state of the metal atoms. The processed data from extended X-ray absorption fine structure (EXAFS) contains information about the local environment of the elements of interest, e.g. the bond distances and coordination numbers of metal centers within the zeolite. The recent developments of in situ/operando XAS techniques were instrumental in unraveling the dynamics and structural flexibility of intrazeolite clusters under realistic conditions of catalytic reactions. The oxidation state of Cu species during methane oxidation to methanol was characterized by Sushkevich et al. by a combination of in situ XAS and Fourier transform infrared spectroscopy (FTIR) of adsorbed probe molecules over Cu@MOR catalyst (Fig. 9).97 The activated catalyst was dominated by Cu species with an oxidation state þ 2 as revealed by XANES. Upon reaction with methane, up to 70% of these CuII species were gradually reduced to a CuI state with simultaneous production of the oxygenated surface intermediates. The reaction with methane was followed by the introduction of steam into the reactor to extract the methanol product. This treatment resulted in the decrease of the peak due to CuI with a concomitant increase of the CuII signals in the spectrum. The interconversion of CuI and CuII species was also confirmed by FTIR using CO and NO as molecular probes. The speciation of Cu clusters in SSZ-13 zeolite has been intensively studied by XAS.63,98–101 Cu/SSZ-13 has been industrialized for the selective catalytic reduction (SCR) of NOx with NH3. However, the nature and composition of Cu species are still not fully understood. Many spectroscopic studies have been devoted to the problem of the nature and speciation of Cu sites in these catalysts with several excellent reviews on the subject published in recent years.11,102–104 The complexity of the structural problem of catalysis by confined transition metal clusters in zeolites can be well illustrated by the recent studies by Paolucci et al.104 A comprehensive study of the active site speciation and evolution under the catalytic conditions has been carried out by a combination of site-sensitive experimental characterization (ex-situ, in-situ, and operando XAS) and computational modeling (density functional theory (DFT), ab initio free energy and molecular dynamics).63 Two distinct structures were identified, namely, Cu2þ and (CuOH)þ charge-balanced by the lattice Al sites of the zeolite framework. It was shown that H2O solvation at ambient or NH3 solvation under the SCR reaction condition can liberate the Cu species from the initial binding sites and induce their dynamic behavior and self-organization inside the zeolite pores.105 It was demonstrated that the isolated Cu ions solvated by NH3 (CuI(NH3)2) can diffuse from one cage to the adjacent cage far away from the charge compensated Al center via the 8-membered ring window. The resulting binuclear complex promotes the O2 activation. Such dynamic paring process is revisable and the catalytic cycle is closed by the subsequent reduction of CuII by NO and NH3 to produce N2, H2O, and regenerate CuI(NH3)2.98 A similar Cu site dynamics was observed by Dinh and co-workers during continuous partial oxidation of methane to methanol.106 Based on the kinetic and in situ spectroscopy studies, the authors proposed the binuclear (Cu2O)2þ clusters, formed by the diffusion of hydrated Cu ions along a proton-paved highway, as the major active site for selective oxidation of methane to methanol. Such active site migration can be blocked by the introduction of NH3, which strongly binds with both the zeolitic protons and Cu ions. It was found that the optimal conditions for the catalytic reaction and the formation of such transient dimeric Cu moieties combine the high CH4 and water partial pressures and high density of Brønsted acid sites on the framework of zeolite

Metal containing nanoclusters in zeolites 123

Fig. 8 Cs-corrected STEM ADF images and EEL spectrum of Fe-MFI. (A) High-resolution ADF image. (B) EEL spectrum. (C–E) Enlarged images corresponding to the three regions marked by rectangles in (A) together with surface plots of 2D-intensity distribution map, where bright dots in (A), (C), (D), and (E) are marked by arrows with T-site symbols. (F) Simulated images of Fe-MFI, where two single Fe atoms are located at T2 and T5 sites corresponding to two Fe atoms per unit cell, under the conditions of probe size: 1.0 Å and specimen thickness: 105 Å. Reproduced with permission from Mayoral, A.; Zhang, Q.; Zhou, Y.; Chen, P.; Ma, Y.; Monji, T.; Losch, P.; Schmidt, W.; Schüth, F.; Hirao, H.; Yu, J.; Terasaki, O. Direct Atomic-Level Imaging of Zeolites: Oxygen, Sodium in Na-LTA and Iron in Fe-MFI. Angew. Chem. Int. Ed. 2020, 59(44), 19510–19517. Copyright 2020, Wiley.

124

Metal containing nanoclusters in zeolites

Fig. 9 (A) In situ XANES spectra recorded during the interaction of Cu@MOR (pretreated in helium flow) with methane at 473 K. (B) In situ XANES spectra recorded during the interaction of water vapor with Cu@MOR at 473 K and 1 bar after the methane reaction. (C) FTIR spectra of CO adsorbed at 100 K onto Cu@MOR that was vacuum-activated (bottom), reacted with methane (middle), and reoxidized with water vapor (top). (D) FTIR spectra of NO adsorbed at 100 K onto Cu@MOR that was vacuum-activated (bottom), reacted with methane (middle) and reoxidized with water vapor (top). Reproduced with permission from Sushkevich, V. L.; Palagin, D.; Ranocchiari, M.; van Bokhoven, J. A. Selective Anaerobic Oxidation of Methane Enables Direct Synthesis of Methanol. Science 2017, 356(6337), 523–527. Copyright 2017, Science.

(Fig. 10). The higher mobility of Cu hydroxide species compared to Cu-oxo species was also proposed by DFT calculations.107 These observations underscore the fact that reaction conditions can have significant influence on the structure and properties of the active site. Therefore, the true active site catalyzing the reaction should be deduced by taking environment-induced structural reorganization into serious consideration. The nature of the active sites and the mechanism of methane dehydroaromatization (MDA) by Mo@ZSM-5 zeolite catalysts has been revisited recently based on the data from advanced characterization techniques.108 The reaction is operated under extremely harsh conditions. Additionally, the fresh catalyst needs to be carburized before the catalytic reaction takes place. Thus, the characterization of the active Mo species in the working catalyst has been a big challenge. Recently, a combination of operando XAS, STEM, and solid-sate NMR shed new light onto this “old” catalyst system.109 Beale et al. utilized XRD, high energy resolution fluorescence detection (Ka-detected) X-ray absorption near-edge spectroscopy (HERFD-XANES) and X-ray emission spectroscopy (XES) under operando reaction conditions to study the nature of the molybdenum species during the MDA reaction over Mo/H-ZSM-5.110 The nature of the Mo species contributing to methane CeH bond activation and the formation of the aromatic products has been a subject of long-term debate in the scientific community. Both MoCx and MoCxOy have been proposed to be active sites because it was not possible to distinguish the elements of carbon and oxygen.111 By the combination of X-ray emission spectroscopy (XES) with HERFD-XANES, it was possible to unambiguously distinguish Mo-carbide and Mo-oxocarbide during the course of the reaction. For the first time, evidence was presented that Mo-oxocarbide species are responsible for the formation of C2Hx/C3Hx light hydrocarbons, while MoC3 formed by MoCxOy carburization prompts benzene production. It was also proposed that the death of

Metal containing nanoclusters in zeolites

125

Fig. 10 (A) XANES and (B) EXAFS of CueCHA (0.11) and (C) XANES and (D) EXAFS of 1AleCueCHA and CueCHA (0.11) under methanol synthesis and NH3 flows. All spectra were collected at 543 K. The catalyst was pretreated in 1 kPa O2, bal. He from 298 to 543 K at 5 K min 1 (pretreat). At 543 K, the catalyst was exposed to methanol synthesis flows of PCH4 ¼ 18 kPa, PO2 ¼ 0.09 kPa, PH2O ¼ 3.14 kPa, bal. He (SS), an additional NH3 cofeed of PNH3 ¼ 0.16 kPa (SS þ NH3), and then the removal of NH3 (SS post-NH3). Following NH3 removal, the catalyst was brought to 673 at 6 K min 1 and held in dry He before cooling to 543 K and exposure to methanol synthesis flows (SS post-purge). (E) Depiction of hypothesized Hþ/H2O-aided diffusion of Cuþ and NH3 inhibition within SSZ-13 to form Cu dimers relevant to the proposed catalytic methane oxidation cycle. Single O atoms may correspond to framework zeolite O atoms or coordinating H2O molecules. Reproduced with permission from Dinh, K. T.; Sullivan, M. M.; Narsimhan, K.; Serna, P.; Meyer, R. J.; Dinca, M.; Román-Leshkov, Y. Continuous Partial Oxidation of Methane to

126

Metal containing nanoclusters in zeolites

the catalyst is mainly caused by the sintering of MoC3 and the deposition of large hydrocarbon species on the external surface of the zeolite.110,112 Kosinov, Hensen, and coworkers investigated the same reaction and catalyst with a specific focus on the early stages of MoOx activation/carburization and induction period of benzene formation by operando spectroscopic tools.109,113–117 They designed a high-resolution pulse reaction method to track the reaction.113 The nature of the intrazeolite sites in Mo/ZSM-5 catalysts formed at the different stages of the reaction was probed by operando XANES while pulsing methane, high-resolution high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and energy-dispersive X-ray (EDX) elemental map of the catalysts (Fig. 11).116,118 The results of nuclear magnetic resonance (NMR) and electron paramagnetic resonance (EPR) spectroscopy pointed to the crucial role of carbonaceous deposits formed in the course of the reaction for the conversion of methane to aromatic products.116 Such polyaromatic species were proposed to act as the hydrocarbon pool for benzene production. Vollmer and co-workers applied an alternative CO carburization strategy avoiding the formation of carbonaceous species.119 The active Mo species can thus be probed directly while eliminating the impact of the aromatic carbon. Based on the results of in situ XANES and 13 C NMR spectroscopy, the authors proposed that the active site is composed of two types of Mo oxycarbidic species. Moliner et al. combined the “average” spectroscopy of in situ quantitative transient EXAFS and “local” microscopy of aberrationcorrected HAADF-STEM to investigate the dynamic and reversible interconversion of Pt nanoclusters and single Pt metal atoms encapsulated in a nanocrystalline high-silica CHA zeolite under reduction and oxidation thermal treatments.120 Pt nanoparticles of around 1 nm were found to be remarkably stable in the presence of H2 and steam at 650
C. However, these nanoparticles decomposed into site-isolated Pt single atoms upon switching to O2 atmosphere during the calcination at 650
C. Furthermore, such aggregationddecomposition process was found to be reversible enabling the efficient regeneration of the metal-zeolite composite catalysts. For instance, the coke deposited on the active site could be removed by calcination in O2 and the H2 treatment re-activated the metal catalytic sites (Fig. 12). Further studies on the dynamic interconversion between highly dispersed SAS Pt species, sub-nanometer Pt clusters, and small Pt nanoparticles revealed a general nature of the observed phenomenon that is highly sensitive to the specific reaction conditions, atmosphere, and operating temperature.121 The development of in-situ and operando characterization is of great importance to provide comprehensive understandings of the dynamic transformation of the catalyst and true active site under catalyst working conditions and thus provide a fundamental basis for the rational design of tailor-made catalysts.122

6.06.3.3

Vibrational spectroscopy

The most widely applied vibrational spectroscopies include Raman and IR, which have been intensively used in catalysis to provide a structural fingerprint of the active metal species in zeolites.123 The speciation of Cu-oxo complexes confined in ZSM-5 zeolite was investigated by resonance Raman spectroscopy by Solomon and co-workers in 2009.124 The most intensive isotope sensitive vibration at the 456 cm1 was assigned to the symmetric stretch of the oxo group bridging two Cu metal center. The weak band at 870 cm1 was assigned to the antisymmetric Cu-oxo stretch vibration mode (Fig. 13). Thus the intrazeolite site responsible for methane activation was exclusively attributed to the binuclear (Cu2O)2þ species, while all other proposed to date binuclear Cu clusters were ruled out. This result is consistent with the observation that methane activation is correlated with the disappearance of the UV–Vis band at 22700 cm1, that was also attributed to (Cu2O)2þ site.125 The same group further identified two types of (Cu2O)2þ species located at different positions within mordenite. These species showed similar geometric and electronic properties but distinguishable activity toward methane activation. The subtle structural differences cannot be distinguished from the resonance Raman spectra.126 However, by using probe molecules with different molecular sizes, it was indicated that the location of the reactive (Cu2O)2þ is very likely to be the 8 membered ring channels and side pockets of the MOR.127 Besides the assignment of the active Cu-oxo species, resonance Raman was also used to identify the formation mechanism of the (Cu2O)2þ species by oxygen activation with pre-reduced Cu precursors. The characteristic resonance Raman bands at 735 and 269 cm1 were assigned to isotope sensitive OeO and insensitive CueC vibrations of an intermediate of m-(h2:h2) peroxo dicopper (II) species (Cu2O2)2þ. The combined results of UV–Vis spectroscopy and O2 temperature-programmed desorption (O2-TPD) led to a proposal that the (Cu2O)2þ active site is formed by incorporation of the excessive oxygen into the framework of zeolite with simultaneous oxidation of two Cuþ spectators into Cu2þ.128 However, this mechanism is still under debate because other types of Cu-peroxo species and radicals were also identified in similar Cu-containing zeolite materials.129,130 The nature of the Cr species confined in ZSM-5 was characterized by Gao et al. using in-situ UV–Vis, IR, operando Raman spectroscopy under methane dehydroaromatization reaction condition of 773–1123 K. Both isolated Cr6þ dioxo and Cr3þ mono-oxo species were identified as the dominant species.131 The same methods were also used to study the speciation of molybdenum encapsulated in ZSM-5. Fig. 14A shows the electronic edge values based on in situ UV–Vis spectra of reference Mo oxide compounds and Mo-oxo species in ZSM-5. It was demonstrated that isolated mononuclear Mo-oxo species were relevant to the state of Mo in ZSM-5. Single metal Mo(]O)22þ compensated by two Al sites corresponding to a Raman band at 993 cm1 was observed for the ZSM-5 with Si/Al ¼ 15. With increasing the Si/Al ratio, the new bands at 975 and 984 cm1 were assigned to Mo(]O)2OH and

=

Methanol Catalyzed by Diffusion-Paired Copper Dimers in Copper-Exchanged Zeolites. J. Am. Chem. Soc. 2019, 141(29):11641–11650. Copyright 2019, American Chemical Society.

Metal containing nanoclusters in zeolites

127

Fig. 11 (A) Operando Mo K-edge XANES spectra and (B) corresponding DXANES spectra recorded while pulsing methane at 700
C over the 2% Mo zeolite catalyst. DXANES spectra were obtained using a spectrum recorded before pulsing methane as the background for subtraction. (C) Comparison of the intensity of the main DXANES feature at 20.015 keV and the benzene yield values obtained simultaneously by MS analysis for 1% Mo, 2% Mo, and 5% Mo catalysts. (D) HAADF-STEM images and corresponding EDX elemental maps of Mo/ZSM-5 catalysts obtained with the catalysts quenched at the maximum benzene yield. (E) Proposed evolution of the active Mo phase. Reproduced with permission from Kosinov, N.; Wijpkema, A. S. G.; Uslamin, E.; Rohling, R.; Coumans, F. J. A. G.; Mezari, B.; Parastaev, A.; Poryvaev, A. S.; Fedin, M. V.; Pidko, E. A.; Hensen, E. J. M. Confined Carbon Mediating Dehydroaromatization of Methane over Mo/ZSM-5. Angew. Chem. Int. Ed. 2018, 57(4), 1016–1020. Copyright 2018, Wiley.

(SieOe)2Mo(]O)2 species anchored on sites with one Al atom and on the external surface, respectively (Fig. 14B). Importantly, operando Raman spectroscopy indicated that the Mo-oxo sites gradually reduced to oxycarbide or carbide species in the course of methane dehydroaromatization reaction, and they could be restored and even enhanced by the oxidation treatment with oxygen (Fig. 14C and D).132 The nature of the reactive extraframework Fe confined in zeolite, the so-called a-Fe and the related reaction intermediate of a-O, was for the first time distinguished from the inactive spectators of other iron-containing species by UV–Vis and variable-temperature variable-field magnetic circular dichroism (VTVH-MCD) spectroscopy. The a-Fe was assigned to be a mononuclear high-spin Fe2þ site stabilized in the 6-membered ring of zeolite framework in a square planar configuration, while a-O is a mononuclear Fe(IV)]O species generated by a-Fe reacting with N2O. The exceptionally high reactivity of the a-O

128

Metal containing nanoclusters in zeolites

Fig. 12 (A) Time-resolved XANES spectra of PteCHA-2 (previously calcined in air at 500
C) in 4% H2 as the temperature is ramped from 25 to 400
C at 10
C/min; (C) Time-resolved XANES spectra of PteCHA-2 (previously reduced in H2 at 400
C) in 20% O2 as the temperature is ramped from 25 to 500
C at 10
C/min; (B) and (D) show the degree of reduction and oxidation, respectively, as a function of the temperature. The solid arrows point to the formation of well-defined isosbestic points during the treatments. (E) FT-EXAFS spectra of PteCHA-2 after various thermal treatments (the arrows indicate the direction of the sequence and corresponding changes in structural features). (F) Schematic illustration of the revisable transformation. Reproduced with permission from Moliner, M.; Gabay, J. E.; Kliewer, C. E.; Carr, R. T.; Guzman, J.; Casty, G. L.; Serna, P.; Corma, A. Reversible Transformation of Pt Nanoparticles into Single Atoms Inside High-Silica Chabazite Zeolite. J. Am. Chem. Soc. 2016, 138(48), 15743–15750. Copyright 2016, American Chemical Society.

intermediate toward methane CeH bond activation is due to the geometrical constraints enforced by the zeolite lattice.6 A mononuclear Cu species anchored to the silanol defect site was also identified by deconvolution of the experimental UV–Vis spectra by simulated excitation signals of different Cu/SSZ-13 models.133 However, precise assignment of the UV–Vis spectroscopic features to specific Cu structures is precluded, for the broad peaks usually result in the overlap of signals from the heterogeneous mixture of Cu motifs.134

Metal containing nanoclusters in zeolites

129

Fig. 13 Resonance Raman spectra (lex ¼ 457.9 nm) of Cu-ZSM-5. 16O2 (red), 18O2 (blue). Inset A: Absorption spectrum of oxygen-activated CuZSM-5. Inset B: “16,18O2” (green), and 1:1 normalized sum of 16O2 and 18O2 (black). Reproduced with permission from Woertink, J. S.; Smeets, P. J.; Groothaert, M. H.; Vance, M. A.; Sels, B. F.; Schoonheydt, R. A.; Solomon, E. I. A [Cu2O]2þ Core in Cu-ZSM-5, the Active Site in the Oxidation of Methane to Methanol. Proc. Natl. Acad. Sci. U.S.A. 2009, 106(45), 18908–18913. Copyright 2009, the United States National Academy of Sciences.

IR spectroscopy is another major and well-established tool for the characterization of zeolite materials.135 The detailed structure of the Brønsted acid sites and their interaction with Lewis sites can be directly distinguished by the signals of hydroxyl groups. Besides, CO is the most widely used IR probe molecule for characterizing cationic sites in zeolites, while different probe molecules could discriminate different coordination and oxidations state, and often the combined use of distinctly different probes is of importance for achieving a comprehensive understanding of the local environments of zeolite materials.136 Qualitative as well as quantitative information of the active site and reaction mechanism may be obtained simultaneously attributing to the development of time and space resolved operando IR spectroscopy.18,137,138

6.06.3.4

Solid-state nuclear magnetic resonance

NMR is a very powerful characterization tool since it is extremely sensitive to the chemical surrounding of the element investigated. Particularly, solid-state NMR (SSNMR) has emerged as a powerful spectroscopic technique capable of providing atomic-level resolution of the structure of the metal active sites in zeolite micropores, their local environment, and dynamic behavior.139,140 Detection of the signal of nuclei of interests with low natural abundance becomes possible because of the improvement of the sensitivity of SSNMR due to the advanced instruments and in-situ and operando SSNMR methods recently developed.141–146 For example, the speciation of such transition metals as Zn, Ga, and Mo have been investigated by SSNMR. To overcome the challenges of the low natural abundance of the NMR active isotopes as well as low gyromagnetic ratios, a 21.1 T ultrahigh field 95Mo NMR spectrometer was used to investigate the nature of the Mo species of 95Mo isotopically enriched Mo/ZSM-5. By comparison of the fresh and carburized catalyst, it was concluded that the carburized Mo species originating from the ion-exchanged Mo species had a good correlation with the aromatics formation rate during catalytic methane dehydroaromatization.147,148 The location, concentration, and internuclear distance of the synergetic active sites of metal-modified zeolite catalysts can be quantified by 1HeM rotational-echo saturation-pulse double-resonance (S-RESPDOR) NMR experiment.140 Deng et al. applied sensitivity enhanced HS-QCPMG (hyperbolic secant-quadrupolar Carr-Purcell-Meiboom-Gill) NMR technique at a high magnetic field of 18.8 T to characterize the nature of Zn species in HZSM-5 for methane CeH bond activation.149 The signal of 67Zn NMR exacerbated by the low loading of zinc species in HZSM-5 zeolite was improved by 16-fold using HS-QCPMG NMR. It was further improved by using 67Zn (67Zn, 89.6%) enriched precursors in preparation of the 67Zn/HZSM-5 samples. The results indicated that there were two 67Zn signals at 224 and 238 ppm which can be assigned to ion-exchanged Zn2þ species located in zeolite channels and highly dispersed ZnO particles, respectively (Fig. 15A). Furthermore, it was found that the strong acid site detected by 1H magic angle spinning (MAS) NMR originated from the synergetic interaction between the Zn Lewis acid and the Brønsted acidic proton evidenced by the 1He67Zn S-RESPDOR NMR data (Fig. 15B). The 1He67Zn internuclear distance between the zeolite Brønsted acidic protons and the Zn2þ Lewis acid site was determined to be 2.70–3.34 Å (Fig. 15C). The enhanced Brønsted acidity and the reactivity Zn-HZSM-5 for CeH bond activation is a result of such a Lewis-Brønsted acid synergy. The same double-resonance technique was also used to unravel the Lewis-Brønsted acid synergy between acidic protons and Ga species in Ga/ZSM-5 zeolite contributing to its reactivity in the methanol-to-aromatics conversion.150,151 Two types of Ga species, i.e. GaOþ ions and GaOx clusters, were proposed based on the results of the wideband uniform-rate smooth truncation and QCPMG (WURST-QCPMG) 71Ga

130

Metal containing nanoclusters in zeolites

Fig. 14 (A) Electronic edge values based on in situ UV–Vis spectra of reference Mo oxide compounds exhibit a linear correlation with the number of bridging MoeOeMo covalent bonds around the central Mo cation. The value of 4.8 eV for 2 wt% Mo/ZSM-5 (Si/Al ¼ 15) corresponds to Mo oxide species with a single Mo atom. (B and C) In situ Raman spectra of Mo/ZSM-5 catalysts under oxygen flow at 773 K as a function of (B) Mo loading for constant Si/Al ¼ 15 and (C) Si/Al ratio for constant 1.3 wt% Mo loading with band assignments to Mo oxide species based on DFT calculations. A. u., arbitrary units. (D) Operando Raman spectra of 2 wt% Mo/ZSM-5 (Si/Al ¼ 15). (E) Proposed initial Mo(]O)22þ nanostructure anchored on double Al-atom framework sites. Reproduced with permission from Gao, J.; Zheng, Y. T.; Jehng, J. M.; Tang, Y. D.; Wachs, I. E.; Podkolzin, S. G. Identification of Molybdenum Oxide Nanostructures on Zeolites for Natural Gas Conversion. Science 2015, 348(6235), 686–690. Copyright 2015, Science.

NMR (Fig. 15D). The quantification of the Ga Lewis acid-Brønsted acid pairs was carried out by the 1He71Ga S-RESPDOR NMR (Fig. 15E). The spectroscopy evidenced the localization of such pair sites at the 6-membered ring inside the zeolite channels (Fig. 15F). The presence of such synergistic sites has been correlated with aromatics selectivity. Extraframework Al (EFAl) species generated upon the dealumination of the zeolite lattice can serve as Lewis acid sites and have significant influences on the catalytic properties of the zeolites. The chemical composition, structure, location, and distribution of EFAl have been intensively studied by various spectroscopic tools including solid-state NMR spectroscopy. The formation, coordination environment, and local structure of EFAl were investigated by one-dimensional single-pulse 27Al MAS NMR and twodimensional multiple-quantum magic angle spinning (MQ-MAS) NMR.152,153 Many EFAl species such as Al(OH)3, AlOOH, AlOþ, Al(OH)2þ, and AlOH2þ species have been directly identified in dealuminated zeolites by 1H double-quantum MAS NMR (DQ-MAS NMR)154 and 27Al DQ-MAS NMR.141 Very recently, the presence of three-coordinate EFAl-Al3þ species and their interaction with adjacent Brønsted acid site resulting in the enhanced acidity and catalyst activity has also been confirmed by the SSNMR combined with suitable probe molecules.155,156 Spatial proximity of three types of EFAl, i.e. Al(OH)3, Al(OH)2þ, and AlOH2þ, in close association with the framework aluminum was confirmed by two-dimensional 27Al DQ-MAS NMR.141 The dealumination mechanism of EFAl species in zeolite during the high-temperature activation can be probed in situ by the high field sensitivityenhanced 27Al DQ-MAS NMR technique. The interaction of the framework aluminum pairs and cationic EFAl species was detected by a combination of multiple-quantum MAS (MQMAS) and two-dimensional double-quantum homonuclear NMR correlation spectroscopy.157 The structure and location of other extraframwork metal species such as Ag and Ti confined in micropore of zeolites have also been examined by SSNMR.158,159 TS-1 has been widely used in industry as a catalyst for the epoxidation of propylene with hydrogen peroxide (H2O2). Over the past 40 years, there had been a consensus that the active site was a mononuclear Ti site on the framework of TS-1 zeolite until very

Metal containing nanoclusters in zeolites

131

Fig. 15 (A) 67Zn HS-QCPMG NMR spectrum of Zn/ZSM-5, (B) 1He67Zn S-RESPDOR NMR spectrum, and (C) illustration of spatial interaction between Zn species and Brønsted acid site. (D) 71Ga WURST-QCPMG NMR spectrum of Ga/ZSM-5, (E) 1He71Ga S-RESPDOR build-up curve of Brønsted acidic proton and (F) model of proximate Ga species and Brønsted acid site in ZSM-5 channel.140 Reproduced with permission from Li, S.; Lafon, O.; Wang, W.; Wang, Q.; Wang, X.; Li, Y.; Xu, J.; Deng, F., Recent Advances of Solid-State NMR Spectroscopy for Microporous Materials. Adv. Mater. 2020, 32(44), 2002879. Copyright 2016, Wiley.

recently Gordon et al. proposed a new binuclear Ti site.160 The catalyst was studied by solid-state 17O NMR in combination with other analytical techniques and computational modeling. It was found that the reaction intermediate detected by 17O NMR is similar to that of the Berkessel–Katsuki catalyst, i.e. a homogeneous titanium catalyst that contains a binuclear Ti active site being able to catalyze the epoxidation reactions as well. This observation together with the simulated NMR spectra and mechanistic computational analysis confirmed that the cooperativity of two Ti sites within such a binuclear cluster embedded inside the zeolite lattice is key to providing the low-energy path for propylene epoxidation. This result is very interesting because binuclear Ti site was proposed earlier to be the active site based on the EXAFS results. However, the conclusion was not accepted by the catalysis community because the interpretation of the EXAFS peak was believed to be wrong.161 It is thus suggested that the EXAFS data for TS-1 should be re-evaluated, and even further more-advanced characterizations should be carried out to further confirm the presence of the binuclear site. The reactivity of other Ti sites such as framework and extraframework defects as well as small titanium oxide clusters should not be assertively excluded and the composition dependence of TS-1 on the reaction conditions should be further investigated.162 It proved that 17O NMR is a very powerful technique that can be used to study the nuclearity of metal active sites in zeolites and the virtue of homogeneous catalyst can be realized by heterogenous catalysis.

6.06.4

Catalytic applications

Zeolites have been widely applied as robust catalysts for promoting various types of reactions by introducing nanoclusters and isolated metal sites as active components confined in the micropores. The design and development of bifunctional and multifunctional catalysts by encapsulation of metal-containing species in zeolite have significantly broadened the application of zeolite-based catalysts beyond the Brønsted acid-catalyzed reactions. Encapsulation of metal nanoclusters in zeolites has applications in diverse processes such as catalytic hydrogenation, hydrodeoxygenation, and oxidation reactions.24,163,164 In the previous sections we were not able to avoid references to specific catalytic applications or mechanistic conclusions relevant to the advanced synthetic and characterization efforts. In this section, we discuss in more detail some selected catalytic applications in an attempt to highlight recent key mechanistic and conceptual discoveries relevant to catalysis by metal-containing clusters encapsulated in zeolite micropores.

6.06.4.1

C1 molecules conversion

Zeolite-supported metal nanoclusters and single-atom sites exhibit promising reactivity for the transformation of C1 molecules into various hydrocarbons and oxygenates.165 For example, both Fe- and Cu-containing zeolites are active in partial oxidation of methane to methanol at mild conditions.106,166–169 The activity of Cu@ZSM-5 was attributed by Sels and co-workers to the presence of binuclear (Cu2O)2þ sites in zeolite channels based on the results of in-situ UV–Vis, resonance Raman combined with DFT calculations.124,128 Later, the trinuclear (Cu3O3)2þ single site stabilized at the entrance of MOR side pocket was proposed as the alternative highly active methane oxygenation site capable of contributing with up to two out of three mu-oxo ligands in its

132

Metal containing nanoclusters in zeolites

structure for the oxidation reaction at elevated pressure.170,171 Methane conversion by Cu-zeolites is normally carried out in a stepwise process including the high-temperature activation of the Cu species, low-temperature methane oxidation, and methanol extraction with the assistance of water. The heating and cooling procedures limit the practical application of this approach. The first isothermal continuous cyclic process was reported by van Bokhoven’s group.172 The reaction procedure was optimized so that both Cu activation and methane oxidation can be operated at a relatively low temperature with 8 h of 1 bar oxygen and subsequent reaction under 6 bar of methane feeding. Methanol can be obtained directly by on-line water extraction at the same temperature.172 It was further proposed that the molecular water can serve as an alternative soft oxidant oxidizing the Cu species to active Cu-oxo sites for activation of methane to methanol.173 In such a scenario, the methanol selectivity over Cu@MOR can reach 97% with 0.202 mol CH3OH per mole Cu methanol productivity.97 An alternative novel concept called “molecular fence” was established by Xiao et al.174 The idea is to increase the local concentration of H2O2 oxidant around the encapsulated AuPd alloy nanoparticles in ZSM-5 zeolite by constructing a hydrophobic external surface around the zeolite. The organic substituent of the silane sheath can prevent the diffusion of H2O2 from the zeolite nanoreactor and thus increase the H2O2 concentration around the AuPd alloy. Meanwhile, hydrophobic methane molecules can easily access the AuPd active site through the hydrophobic silane sheath, and be oxidized to methanol over the AuPd active site with precisely concentrated on-site generated H2O2 oxidant. The catalyst exhibits high activity, excellent selectivity, and good recyclability. Methane conversion can reach 17.3% with methanol selectivity of 92%, corresponding to methanol productivity up to 91.6 millimoles per gram of AuPd per hour (Fig. 16). Conversion of syngas to hydrocarbon by Fe, Co, and Zn-containing zeolites benefits from the bifunctional nature of these catalysts.165 Water-gas shift reaction with high conversion of CO (70%) can be achieved by Pt nanoclusters encapsulated in NaA zeolite.175 Catalytic selectivity in CO2 hydrogenation can be tailored by adjusting the microporous environments. Rh@S-1 catalyst shows high activity for CO2 conversion and maximized selectivity toward CO without deep hydrogenation. In contrast, the MFIsupported Rh nanoparticles exhibit high methane selectivity under similar conditions, indicating a strong correlation between the microporous environment and the catalyst selectivity.176 Besides the superior catalytic activity and selectivity, various synthesis and post-synthetic approaches were developed to obtain robust nanoclusters or single-atom site confined in the zeolites with enhanced long-time stability and sintering/coke resistance in a variety of industrial conditions.47 Bimetallic nanoclusters encapsulate in zeolites were synthesized with excellent stability and reactivity. AuPd@S-1 synthesized by the solvent-free method can convert bioethanol to acetic acid with > 80% conversion and 95% selectivity in the presence of 90% water in the system.43 The metal utilization rate in such applications reaches 98%. Xiao et al. reported a bimetallic RhMn core-shell nanoparticles RhMn@S-1 that can directly convert CO and H2 into C2-oxygenate with superior long-term durability.177 Selectivity to ethanol of 88% at CO conversion of 42% was achieved.177 The excellent catalytic performance was attributed to the sintering hindrance of the S-1 zeolite support which contributes to the stabilization of the active Mn-O-Rhþ species even under reductive atmospheres.

6.06.4.2

Active site cooperation and multifunctionality in confined space

The bifunctionality of zeolite-based catalysts originating from the simultaneous presence of the metal and the Brønsted acid sites in the confined space of the zeolite micropores was found to be critically important for various catalytic reactions. By adjusting the density of each component, the reactivity and selectivity can be fine-tuned to optimize the catalyst performance.70,163,178 Encapsulation of the isolated WO4 sites in the immediate proximity with the Brønsted acid sites dramatically enhanced the productivity of propene by cross-metathesis of ethene and trans-2-butene. Such a cooperative catalyst gave rise to a production rate is ca. 7300 times that of the industrial WO3/SiO2-based catalyst. No observable side products and coke significantly formed under a wide range of reaction conditions.70 The synergy between ultrasmall Ru clusters and zeolitic Brønsted acid sites was also found to be crucial for significant enhancement of the hydrogen evolution from ammonia borane hydrolysis. The hydrogen evolution rate was promoted by the formation of bifunctional RueH active sites between Ru clusters and adjacent Brønsted acid sites with tunable acidity, which could synergistically activate ammonia borane and water molecules. The TOF values of Ru/SAPO-34 and Ru/FAU are much higher than that of the commercial Ru/C catalyst, and are among the highest overall metal-based heterogeneous catalysts for ammonia borane decomposition reported so far under similar conditions (Fig. 17).179 Pt nanoclusters were encapsulated in HZSM-5 efficiently by a cationic polymer-assisted synthetic strategy.33 The catalytic test indicated that Pt@HZSM-5 can promote tandem aldol condensation of furfural with acetone and the subsequent hydrogenation reaction by the synergy of metal and Brønsted acid sites, in contrast to the hydrogenation and decarbonylation of furfural promoted by Pt nanoparticles supported on HZSM-5 under identical conditions. The encapsulated metal clusters were found to be highly thermostable without significant metal sintering or leaching observed during catalyst regeneration. The same group further identified that this selective multistep tandem reaction is favorable only when the encapsulated Pt nanoparticles and Brønsted acid sites are adjacent with each other facilitating the access of the reaction intermediate to the active site of the subsequent reaction and reducing the catalyst deactivation.180 Besides the Brønsted acids, the heteroatom of zeolite framework, for instance, Ti sites of TS-1 zeolite can also synergistically interact with encapsulated metal nanoclusters to enhance the catalytic performances. The electron transfer from the Ti site of TS1 to encapsulated Au nanoparticles weakens the O2 adsorption and therefore promotes the in-situ formation of H2O2 for the vapor phase propylene epoxidation.48 Atomically dispersed Au species can be stabilized and activated in zeolites by alkali ions (Naþ and Kþ) additives. The catalytic performance for the water-gas shift reaction at a temperature below 200
C was comparable to the Au species on reduced metal oxide supports.72 S-1 supported Ru single-atom catalyst shows activity in the catalytic ammonia synthesis

Metal containing nanoclusters in zeolites

133

Fig. 16 (A) Data characterizing the oxidation of methane with H2 and O2 over various catalysts. Reaction conditions: 10 mL of water, 30 min, 70
C, 27 mg of catalyst, feed gas at 3.0 MPa with 3.3% H2/6.6% O2/1.6% CH4/61.7% Ar/26.8% He, and 1200 rpm (rpm). (B and C) Dependences of the methane conversion (Conv.), methanol selectivity (Sel.), methanol productivity (Prod.), and H2O2 concentration in water solution on reaction time over (B) AuPd@ZSM-5-C16 and (C) AuPd@ZSM-5 catalysts. Reaction conditions: 10 mL of water, 70
C, 27 mg of catalyst, feed gas at 3.0 MPa with 3.3% H2/6.6% O2/1.6% CH4/61.7% Ar/26.8% He, and 1200 rpm. Each reaction was tested eight times to obtain the error bars. Reproduced with permission from Jin, Z.; Wang, L.; Zuidema, E.; Mondal, K.; Zhang, M.; Zhang, J.; Wang, C.; Meng, X.; Yang, H.; Mesters, C.; Xiao, F.-S. Hydrophobic Zeolite Modification for In Situ Peroxide Formation in Methane Oxidation to Methanol. Science 2020, 367(6474), 193–197. Copyright 2020, Science.

134

Metal containing nanoclusters in zeolites

Fig. 17 (A and B) Volume of the H2 generated from AB (1 M) hydrolysis versus time at 25
C catalyzed by various catalysts (nRu/nAB ¼ 0.007). (C) The proposed mechanism for NH3BH3 hydrolysis over Ru/SAPO-34 catalysts. (D) Volume of the H2 generated from AB (1 M) hydrolysis versus time and (E) corresponding TOF values at different temperatures catalyzed by Ru/SAPO-34-0.2Si catalyst (nRu/nAB ¼ 0.007), inset of (D): Arrhenius plot (ln TOF versus 1/T). (F) Durability tests for the AB (1 M) hydrolysis at 25
C over Ru/SAPO-34-0.2Si catalysts (nRu/nAB ¼ 0.007). Reproduced with permission from Sun, Q.; Wang, N.; Bai, R.; Hui, Y.; Zhang, T.; Do, D. A.; Zhang, P.; Song, L.; Miao, S.; Yu, J. Synergetic Effect of Ultrasmall Metal Clusters and Zeolites Promoting Hydrogen Generation. Adv. Sci. 2019, 6(10), 1802350. Copyright 2019, Wiley.

higher than the most active conventional CsRu/MgO catalyst, and the ammonia synthesis rate can be further improved by 2 orders of magnitude by introducing a Ba promoter.77 Yu and coworkers discovered that the bimetallic PtZn cluster encapsulated in S-1 has a very high propane dehydrogenation activity with 40.4% propane conversion with 99.3% selectivity to propene even after 13,000 min on stream without co-feeding H2.36 However, the performance of mono-metallic Pt@S-1 treated by conventional calcination/reduction process decreases to 19.8% and 2% after 2900 and 340 min on stream due to significant catalyst deactivation. The incorporation of Zn significantly enhances the stability of the subnanometer Pt clusters compared to that of PtZn/Al2O3 under identical conditions. Similar enhanced catalytic activity and sintering resistance were also observed in other bimetallic zeolite catalyst systems, attributed to the electronic interaction between Pt species and the second metal species in proximity as a stabilizer.37,181

Metal containing nanoclusters in zeolites

135

Long-term catalyst regeneration and recyclization by the introduction of Cs cations into the system make PtZn@S-1 a promising new catalyst for industrial applications.36,182

6.06.4.3

Confined space for selectivity control

Encapsulation of metals into zeolites can selectively hydrogenate one specific functional group presented in one molecule adsorbing on the metal clusters. This represents an effective strategy to control the catalytic selectivity through the unique adsorption conformation with the target functional group on the metal active sites. By encapsulating Pd nanoclusters into zeolites, the multiplesubstituted nitroarenes reactants can selectively adsorb to the Pd nanocluster via the nitro group. The unique end-on adsorption conformation constrains the activation of only one adsorbed nitro group with the other one in the molecule intact, giving extraordinary hydrogenation selectivity and stability toward aniline compared to conventional supported Pd catalysts (Fig. 18).46,183 Selective hydrogenation of nitro group was achieved by the uniform bimetallic PtZnx nanoclusters encapsulated in MFI zeolite as well.25 Xing et al. prepared Pt nanoparticles encapsulated in Y zeolite and found that the hydrogenation of the benzene ring of acetophenone can be completely suppressed due to the thermodynamically unfavorable end-on adsorption of acetophenone on the encapsulated Pt nanoparticles in Y zeolite.184 The elusive para regioselectivity of oxidative coupling of toluene CeH bond of aromatics to produce biarylic compounds can be achieved by encapsulation of single Pd sites in Beta zeolite.185 The selectivity to specific products can also be modulated by encapsulation of nano species into different zeolite frameworks or steering the steric hindrance and electronic property of the metal clusters by surface modification with alkali cation.186 By fixation of TiO2 in the micropore, the synergy between the two functionalities of photocatalytic activity of TiO2 and the hydrophobic property of zeolite micropore

Fig. 18 Substrate conversions (+) and product selectivities (colored columns) for the hydrogenation of (A) 4-nitrochlorobenzene and (B) 4nitrobenzaldehyde on various catalysts. Reaction conditions: substrate (1 mmol), Pd catalyst (0.2 mol %), toluene (10 mL), H2 (1 MPa), 110
C for 45 min (4-nitrochlorobenzene) or 80
C for 2 h (4-nitrobenzaldehyde). Proposed models for the adsorption of 4-nitrochlorobenzene on (C) Pd/C and (D) Pd@Beta. C gray, Cl light green, H white, N blue, O red, Pd dark green. Reproduced with permission from Zhang, J.; Wang, L.; Shao, Y.; Wang, Y.; Gates, B. C.; Xiao, F.-S. A Pd@Zeolite Catalyst for Nitroarene Hydrogenation with High Product Selectivity by Sterically Controlled Adsorption in the Zeolite Micropores. Angew. Chem. Int. Ed. 2017, 56(33), 9747–9751. Copyright 2017, Wiley.

136

Metal containing nanoclusters in zeolites

enhance the conversion of formaldehyde into CO2 in a wet atmosphere.187 In contrast, the hydrophobic nature of zeolite can also be fine-tuned to be hydrophilic by controllable functionalization of silanol groups of the zeolite. This way, the adsorption and diffusion of reactant and intermediates are dramatically influenced and the selectivity to specific products can be achieved.188

6.06.5

Computational modeling

In the previous section various state-of-art and advanced characterization tools for addressing the structural and mechanistic challenges of zeolite catalysis have been discussed. Next to the experimental strategies, computational modeling has been proven indispensable in understanding the physical and chemical properties of nanoclusters confined in zeolites. Nowadays computer simulations, particularly electronic structure calculations based on density functional theory, are widely used to address some of the most challenging fundamental questions such as the structure, electronic properties, and location of the nanoclusters, the interaction mechanism between the active site and the adsorbate, and the complex reaction networks underlying their catalytic properties. The development of operando modeling methodologies further narrows the gap between the experiment and the theoretical model and provides more insights into the nature of the active sites and the reaction mechanism under the realistic conditions.189

6.06.5.1

Structure prediction by operando thermodynamic analysis

The structure of the metal-containing species confined in zeolites and their specific locations related to the framework Al distribution are the most studied question by DFT simulations. At the early stage, various species such as isolated metal ions, small metal clusters, and metal-(hydro)oxo complexes have been evaluated by static DFT calculation. Usually, specific atomic structures were proposed based on the indirect experimental data, and then different local environments of T-sites and spatial distributions were screened and the most stable binding positions were proposed. In some cases, the spectra such as IR, Raman, and UV–Vis were also predicted and correlation with the experimental data was discussed. However, since these DFT calculations were approximated by in vacuum models at 0 K limit, there is a big gap between modeling and experiment. A growing volume of evidence has been accumulated in recent years that such oversimplified models can substantially deviate from the true reactive configurations formed and operating under the conditions of the catalytic reactions. A pragmatic solution to addressing the conditiondependencies of catalyst speciation is provided by the ab initio thermodynamic analysis (aiTA) approach190 that allows expanding the energetics of different catalyst states and configurations computed in the idealistic vacuum/0 K model to the conditiondependent free energies reflecting the stabilities under the experimentally relevant conditions. In this approach, a simplified thermodynamic model is constructed to account for the impact of the finite temperature and pressure as well as the varying composition of the reactive atmosphere. By including as many as possible potential model structures, their formation Gibbs free energies and the associated condition dependencies can be computed providing thus a comprehensive thermodynamic picture of the active site structure. From this, direct information on the most stable and meta-stable configurations under specific reaction conditions can be predicted and related to the equilibrium stabilities of extra-framework species in zeolites. Schneider and co-workers for prediction of the structure and oxidation state of Cu ions exchanged in SSZ-13 zeolite under the standard SCR of NOx with NH3.191 It was found that both 4-fold CuI and 2-fold CuII species coordinated with H2O or OH groups were the most stable complexes under a wide range of SCR conditions. We have extensively applied the aiTA approach to study the nature of active sites and catalytic mechanisms by a wide range of zeolite-based systems such Fe/ZSM-5,192 Cu/MOR,170,193 Al/ FAU,194 Mo/ZSM-5,195 etc. Our most recent computational study on Mo/ZSM-5195 also included the influence of the chemical potential of reactants and products into consideration to understand the evolution of the reactive ensembles in the course of the catalytic transformations. The formation Gibbs free energy of both mononuclear and binuclear Mo-(oxo)carbides were compared and the most stable Mo-containing species under the reaction atmosphere of CH4, H2, and CO were predicted. Besides the thermodynamic stability of these species, the probable formation process was also evaluated and the kinetically most favorable carburization pathways to MoCx and MoCxOy were proposed (Fig. 19A and B).195 Although ab initio thermodynamics analysis is very useful for identifying the most relevant species under the catalyst working condition, the structural and compositional database of such analysis is still biased and heavily relied on the chemical knowledge and intuition of the researchers. To include diverse structures and elementary compositions into evaluation, a combination of aiTA with the methods of global optimization methods represent an attractive practical solution.189,196,197 Very recently, we have explored the utility of genetic algorithm in combination with aiTA to determine the structure of EFAl cations in zeolite mordenite. The global optimization method was used to generate an extensive and comprehensive selection of feasible EFAl configurations with varied compositions, of which the stabilities were directly compared through the aiTA analysis (Fig. 19C).

6.06.5.2

Reactivity scaling relationship and beyond

Electronic structure calculations play nowadays a crucial role in understanding the molecular details and unraveling structureactivity relationships in zeolite catalysis. Computations are routinely used to compute barriers and energetics of the elementary steps in potential reaction pathways taking place over the reactive clusters inside the zeolite pores. The great advances in computational hardware and software of the past decades made it possible to carry out routinely transition state calculations on fully periodic zeolite models using modern computational resources. Nevertheless, the complete mapping of the whole reaction networks on

Metal containing nanoclusters in zeolites

137

Fig. 19 (A) Most stable [Mo2OxCy]2þ species as a function of chemical potentials of mCH4, mCO, and mH2. (B) Gibbs free energy of formation of [Mo2OxCy]2þ as a function of mCO and mH2. (C) Most stable configurations of extraframework Al-containing species identified by genetic algorithm and the corresponding ab initio thermodynamic stability analysis (unpublished results). (B) Reproduced with permission from Li, G.; Vollmer, I.; Liu, C.; Gascon, J.; Pidko, E. A. Structure and Reactivity of the Mo/ZSM-5 Dehydroaromatization Catalyst: An Operando Computational Study. ACS Catal. 2019, 9(9), 8731–8737. Copyright 2019, American Chemical Society.

the global potential energy surface is still outside the reach. Furthermore, the large-scale evaluation of the kinetic parameters for an extended set of possible structures using state-of-the-art electronic structure methods can be a very demanding and challenging task. Substantial efforts have been devoted in the last decades to formulate a unifying framework for predicting the activation barriers

138

Metal containing nanoclusters in zeolites

using a single universal descriptor without the need of the most demanding step of locating transition state structures for each catalyst candidate. Such a reduction of the kinetic problem to a problem of thermodynamics (because the universal descriptor should be associated with the characteristics of the energy minima or the proposed active sites) is key to enabling high-throughput computational screening and reactivity assessment of the intrazeolite reactive ensembles with different structures and chemical compositions. This approach has been proved successful in many studies on catalytic reactions on open surfaces of solid catalysts.198 However, the number of reports utilizing this approach in the field of zeolite catalysis has been quite limited until recently.199–204 Nørskov et al. discovered that the activation of the methane CeH bond within the pore of zeolites can also be correlated with twodimensional descriptors of the hydrogen affinity of the active oxygen site (GH) and the formation energy of such active site (Gf) (Fig. 20A).204 Furthermore, the one-dimensional reaction rate volcano pot as a function of Gf was believed to be able to predict the reactivity of other types of material (Fig. 20B). The structure-property relationships of Cu exchanged ZSM-5 catalyst under reaction-relevant environments were established by Wang et al. using DFT calculation and operando FTIR spectroscopy.205 It was found that the fraction of binuclear Cu species and the compression energy describing the interaction between adjacent adsorbates have a quantitative linear correlation with catalytic activities for NO decomposition and methane oxidation to methanol reactions (Fig. 20C–E). However, Hermans et al. found that for transition metal clusters confined in zeolite, multiple descriptors and reconstruction of the active site induced by the adsorption of the molecule should be taken into explicit consideration to improve the accuracy of prediction.206 It would be very promising if such a scaling relationship is always valid, however, later detailed studies of the CeH bond activation ability of Fe-containing nanocluster in ZSM-5 observed significant deviations due to variation of the active site location in zeolite and the different oxygen active site involved into the CeH bond activation (Fig. 21).207 Detailed structural and electronic analysis demonstrated that both the local confinement effect of zeolite framework on the active site and the steric hindrance around the reaction center have substantial influences on the reactivity and the mechanism, and therefore, break the apparent linear scaling relationship between the reactivity and the simplified descriptor. Such reactivity beyond the linear scaling relationship was also observed by Liu et al.208 They found that the protonation reactivity of the Brønsted acid sites in zeolites deviate from the linear scaling relationships described by either adsorption energy of probing molecule of NH3 or the intrinsic acidity of

Fig. 20 (A) Two-dimensional volcano plot that includes GH and Gf as descriptors for all explored materials. (B) One-dimensional volcano plot for the intrinsic rate of methane activation using Gf as a descriptor. An active site formation temperature of 450
C and a methane activation temperature of 150
C are used. (C) Arrhenius plots for rate constants versus 1/T. (D) Correlation between ln[kapp] at 350 and 550
C and fractions of Cu dimers. (E) Correlation between apparent activation energies at low (red) and high temperatures (blue) and the compression energy derived from Ono–Kondo analysis. (B) Reproduced with permission from Latimer, A. A.; Kulkarni, A. R.; Aljama, H.; Montoya, J. H.; Yoo, J. S.; Tsai, C.; Abild-Pedersen, F.; Studt, F.; Nørskov, J. K. Understanding Trends in C–H Bond Activation in Heterogeneous Catalysis. Nat. Mater. 2016, 16, 225. Copyright 2016, Springer Nature. (E) Reproduced with permission from Xie, P.; Pu, T.; Aranovich, G.; Guo, J.; Donohue, M.; Kulkarni, A.; Wang, C. Bridging Adsorption Analytics and Catalytic Kinetics for Metal-Exchanged Zeolites. Nat. Catal. 2021, 4(2), 144–156. Copyright 2021, Springer Nature.

Metal containing nanoclusters in zeolites

139

Fig. 21 Relationships between the computed kinetic (DEa) and thermodynamic ((A) DER and (B) DEH) parameters of methane activation by binuclear Fe sites in ZSM-5 zeolite (circles) and Fe-MIL-53 MOF (squares). The trend lines shown in the graph provide liner fits for the data sets with 15 kJ/mol shown with shaded areas. Reproduced with permission from Szécsényi, Á.; Khramenkova, E.; Chernyshov, I. Y.; Li, G.; Gascon, J.; Pidko, E. A. Breaking Linear Scaling Relationships with Secondary Interactions in Confined Space: A Case Study of Methane Oxidation by Fe/ZSM-5 Zeolite. ACS Catal. 2019, 9(10), 9276–9284. Copyright 2019, American Chemical Society.

bond order, which implies that the reactions taking place in the micropore of zeolites are much more complicated and are governed by multiple factors beyond a single descriptor or two-dimensional descriptors. It also indicates that it is promising to explore new tailor-made zeolite catalysts for further activity improvement by circumventing the fundamental limitations of scaling relationships.209,210

6.06.5.3

Micro-kinetic modeling and dynamics

The combination of DFT reaction pathway calculation with microkinetic simulations is a widely applied and powerful strategy to investigate the complex reaction mechanism and optimize the catalyst performance. The intrinsic reaction energy and activation barriers calculated by DFT within the in vacuum condition and 0 K approximation can be converted by microkinetic modeling to macroscopic kinetic parameters such as the reaction rate, apparent activation barrier, and the coverage of the surface intermediates in terms of specific reaction conditions. Based on such information, the strategies for further optimization of the catalytic process and improve the catalyst selectivity to desired produced can be delivered.211,212 For example, the previously discussed reactivity volcano plots by DFT coupled with microkinetic modeling have allowed the rational understanding of the correlation between the kinetic terms of activation barriers and the thermodynamic parameters of adsorption energy of one or more intermediates. DFT coupled with microkinetic modeling was used to study the reactivity of the Mo-containing nanoclusters in ZSM-5 zeolite. The location effect on the reactivity of the active site was evaluated (Fig. 22). It was found that the reactivity of Mo nanocluster toward methane dehydroaromatization reaction is very sensitive to the specific interaction motif with zeolite framework and the different confinement environment. The location of the most reactive site is the six-membered ring in the straight channel of ZSM-5 which has the lowest apparent activation barrier compared to Mo at other locations. More interestingly, it was found that the reaction ratedetermining step is also site-dependent. Either the CeC bond coupling or CeH bond dissociation dominants the reaction rate at

140

Metal containing nanoclusters in zeolites

Fig. 22 (A) ZSM-5 framework and selected locations of reactive Mo clusters for MKM investigation. (B) Microkinetic modeling of methane activation over the active site of [Mo2C2]2þ in ZSM-5. The formation rates r (in mol$s 1) of ethylene as a function of temperature are presented. The apparent activation barriers (Eapp) indicated in the figures were calculated using the Arrhenius equation. Dual-site microkinetic models were considered. (C–F) Calculated steady degree of rate control (DRC) analysis for methane activation over the [Mo2C2]2þ site along ethylene pathways. Herein, * and # stand for the C and Mo sites of [Mo2C2]2þ, respectively (unpublished results).

different locations, indicating the heterogeneous nature of the reaction process in zeolite and providing new directions for further catalyst optimization. Rational catalyst development and catalytic property improvement cannot be realized without fundamental insights into the dynamic nature of catalytic systems. Computational modeling has been intensively used to help assign spectra obtained from the experiment and detect the dynamics of the active site under catalyst working conditions by a combination of operando characterization and ab initio molecular dynamics (AIMD) and biased molecular dynamics. For example, the structural assignment of the condition-dependent Cu and Pd speciation in zeolite by XAS, DFT, and AIMD,63,213,214 and the dynamic couple of isolated Cu active site into transient binuclear Cu clusters during the SCR of NH3 discovered by ab initio metadynamics simulations discussed in Section 6.06.3 are good examples in this regards.98,99 It was identified that the mobility of the active site can be elevated

Metal containing nanoclusters in zeolites

141

due to the adsorption of water or NH3 reactant and thus the reaction center is not limited to a specific framework anchoring site anymore. The Gibbs free energy obtained by ab initio molecular dynamics simulations is of great significance by the inclusion of enthalpic and entropic contributions into the total energy of the system, which are key factors influencing the reactivity and selectivity of zeolite confined reactions.215–218 For example, it was found that the different reaction rates within MFI and Beta zeolites are a result of the impacts of the confinement and the intermolecular interactions of reaction intermediates, which dominate the intrinsic activation enthalpy and activation entropy of reactions.219 By the combination of spectroscopic and microscopic characterizations and DFT calculations, an atomic-level quantitative analysis of the hydrothermal stability of isolated Cu2þ and CuOHþ species were demonstrated by Song and coworkers. It is found that the Cu2þ is highly active and selective under a wide range of reaction temperatures, however, due to the low hydrolysis barrier, CuOHþ gradually aggregated and converted into CuOx clusters resulting in primary selectivity deterioration.107

6.06.6

Conclusion and perspective

The field of metal encapsulated zeolite materials is under fast development. Novel synthesis strategies have been developed to introduce nanocluster, subnanometric species, and single-atom sites into zeolite with good size and distribution control. Identification of the nature of the metal species inside zeolite is nowadays possible with an atomistic-level resolution by advanced microscopic and spectroscopic techniques. Especially the tremendous progress in the development and application of in-situ and operando characterization tools enables real-time tracking of the whole process of active site formation, substrate conversion, and catalyst deactivation, providing comprehensive qualitative and/or quantitative data and deep understanding about the structural flexibility and dynamic evolution of the active metal species and the reaction intermediates of a specific reaction. New bimetallic and bi- and multi-functional metal-containing zeolite catalysts have been successfully synthesized and improved to to the levels outperforming catalytic performances for existed and new reactions. Furthermore, the combination of experiment and computational modeling has become a regular approach for structure-performance-mechanistic studies. These remarkable achievements are very thrilling and inspiring for the next stage of generalized rational design of new zeolite-based heterogeneous catalysts with excellent activity, selectivity, and stability. New versatile and generalized synthesis approaches should be continually developed. In-situ and operando characterization tools with improved time and spatial resolutions are desired to deconvolute the structure-property relationships and inherent intracluster heterogeneity coupled with both structural and dynamic perturbations.220–224 On the one hand, the gap between realistic system and theoretical model should be further narrowed by investigating the reaction thermodynamics and kinetics in the microenvironment of zeolite pores under relevant reaction conditions.225 On the other hand, high throughput calculations together with newly emerging machine learning methodologies and data-centered tools will significantly improve the efficiency of new zeolite synthesis,226 reaction network construction and catalyst screening.227,228 Although large effort has been put in studies on the direct catalyst fabrication and optimization, fundamental research of the transient sintering process and catalyst deactivation mechanism should be strengthened to provide insight into the nature of the sintering mechanisms. This knowledge can stimulate from a different angle novel strategies for the development of cutting-edge anti-sintering catalysts or regeneration processes. This is a big challenge that can only be addressed by the effective combination of multiple advanced characterization techniques and advanced computational modeling approaches.229,230

Acknowledgment The authors acknowledge help with editing of selected graphics from the TsyfroCatLab of the University of Tyumen supported by the Tyumen region (Russia) by a grant to non-profit organizations no. 89-don.

References 1. Grommet, A. B.; Feller, M.; Klajn, R. Chemical Reactivity Under Nanoconfinement. Nat. Nanotechnol. 2020, 15 (4), 256–271. 2. Wang, J.; Liu, L.; Dong, X.; Alfilfil, L.; Hsiung, C.-E.; Liu, Z.; Han, Y. Converting Hierarchical to Bulk Structure: A Strategy for Encapsulating Metal Oxides and Noble Metals in Zeolites. Chem. Mater. 2018, 30 (18), 6361–6369. 3. Wang, L.; Xu, S.; He, S.; Xiao, F.-S. Rational Construction of Metal Nanoparticles Fixed in Zeolite Crystals as Highly Efficient Heterogeneous Catalysts. Nano Today 2018, 20, 74–83. 4. Serrano, D. P.; Melero, J. A.; Morales, G.; Iglesias, J.; Pizarro, P. Progress in the Design of Zeolite Catalysts for Biomass Conversion into Biofuels and Bio-Based Chemicals. Catal. Rev. Sci. Eng. 2018, 60 (1), 1–70. 5. Ennaert, T.; Van Aelst, J.; Dijkmans, J.; De Clercq, R.; Schutyser, W.; Dusselier, M.; Verboekend, D.; Sels, B. F. Potential and Challenges of Zeolite Chemistry in the Catalytic Conversion of Biomass. Chem. Soc. Rev. 2016, 45 (3), 584–611. 6. Snyder, B. E. R.; Vanelderen, P.; Bols, M. L.; Hallaert, S. D.; Böttger, L. H.; Ungur, L.; Pierloot, K.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. The Active Site of LowTemperature Methane Hydroxylation in Iron-Containing Zeolites. Nature 2016, 536 (7616), 317–321. 7. Shan, J.; Li, M.; Allard, L. F.; Lee, S.; Flytzani-Stephanopoulos, M. Mild Oxidation of Methane to Methanol or Acetic Acid on Supported Isolated Rhodium Catalysts. Nature 2017, 551 (7682), 605–608.

142

Metal containing nanoclusters in zeolites

8. Morejudo, S. H.; Zanon, R.; Escolastico, S.; Yuste-Tirados, I.; Malerod-Fjeld, H.; Vestre, P. K.; Coors, W. G.; Martinez, A.; Norby, T.; Serra, J. M.; Kjolseth, C. Direct Conversion of Methane to Aromatics in a Catalytic Co-ionic Membrane Reactor. Science 2016, 353 (6299), 563–566. 9. Gao, P.; Li, S.; Bu, X.; Dang, S.; Liu, Z.; Wang, H.; Zhong, L.; Qiu, M.; Yang, C.; Cai, J.; Wei, W.; Sun, Y. Direct Conversion of CO2 into Liquid Fuels with High Selectivity over a Bifunctional Catalyst. Nat. Chem. 2017, 9 (10), 1019–1024. 10. Vogt, E. T. C.; Weckhuysen, B. M. Fluid Catalytic Cracking: Recent Developments on the Grand Old Lady of Zeolite Catalysis. Chem. Soc. Rev. 2015, 44 (20), 7342–7370. 11. Beale, A. M.; Gao, F.; Lezcano-Gonzalez, I.; Peden, C. H. F.; Szanyi, J. Recent Advances in Automotive Catalysis for NOx Emission Control by Small-Pore Microporous Materials. Chem. Soc. Rev. 2015, 44 (20), 7371–7405. 12. Zhang, R.; Liu, N.; Lei, Z.; Chen, B. Selective Transformation of Various Nitrogen-Containing Exhaust Gases Toward N2 over Zeolite Catalysts. Chem. Rev. 2016, 116 (6), 3658–3721. 13. Zeolite-Encapsulated Catalysts: Challenges and Prospects. In Encapsulated Catalysts; Farrusseng, D., Tuel, A., Sadjadi, S., Eds., Academic Press, 2017; pp 335–386. Chapter 11. 14. Valden, M.; Lai, X.; Goodman, D. W. Onset of Catalytic Activity of Gold Clusters on Titania with the Appearance of Nonmetallic Properties. Science 1998, 281 (5383), 1647–1650. 15. Haruta, M. When Gold Is Not Noble: Catalysis by Nanoparticles. Chem. Rec. 2003, 3 (2), 75–87. 16. Chen, M. S.; Goodman, D. W. The Structure of Catalytically Active Gold on Titania. Science 2004, 306 (5694), 252–255. 17. Lee, S.; Molina, L. M.; López, M. J.; Alonso, J. A.; Hammer, B.; Lee, B.; Seifert, S.; Winans, R. E.; Elam, J. W.; Pellin, M. J.; Vajda, S. Selective Propene Epoxidation on Immobilized Au6–10 Clusters: The Effect of Hydrogen and Water on Activity and Selectivity. Angew. Chem. Int. Ed. 2009, 48 (8), 1467–1471. 18. Rivallan, M.; Seguin, E.; Thomas, S.; Lepage, M.; Takagi, N.; Hirata, H.; Thibault-Starzyk, F. Platinum Sintering on H-ZSM-5 Followed by Chemometrics of CO Adsorption and 2D Pressure-Jump IR Spectroscopy of Adsorbed Species. Angew. Chem. Int. Ed. 2010, 49 (4), 785–789. 19. Zecevic, J.; van der Eerden, A. M. J.; Friedrich, H.; de Jongh, P. E.; de Jong, K. P. Heterogeneities of the Nanostructure of Platinum/Zeolite Y Catalysts Revealed by Electron Tomography. ACS Nano 2013, 7 (4), 3698–3705. 20. Chai, Y.; Shang, W.; Li, W.; Wu, G.; Dai, W.; Guan, N.; Li, L. Noble Metal Particles Confined in Zeolites: Synthesis, Characterization, and Applications. Adv. Sci. 2019, 6 (16), 1900299. 21. Wang, N.; Sun, Q.; Yu, J. Ultrasmall Metal Nanoparticles Confined within Crystalline Nanoporous Materials: A Fascinating Class of Nanocatalysts. Adv. Mater. 2019, 31 (1), 1803966. 22. Wang, H.; Wang, L.; Xiao, F.-S. Metal@Zeolite Hybrid Materials for Catalysis. ACS Cent. Sci. 2020, 6 (10), 1685–1697. 23. Kosinov, N.; Liu, C.; Hensen, E. J. M.; Pidko, E. A. Engineering of Transition Metal Catalysts Confined in Zeolites. Chem. Mater. 2018, 30 (10), 3177–3198. 24. Zhao, Z.; Li, Y.; Feyen, M.; McGuire, R.; Mueller, U.; Zhang, W. Pd Nanoparticles Encapsulated in FER Zeolite Through a Layer Reassembling Strategy as Shape-Selective Hydrogenation Catalyst. ChemCatChem 2018, 10 (10), 2254–2259. 25. Iida, T.; Zanchet, D.; Ohara, K.; Wakihara, T.; Roman-Leshkov, Y. Concerted Bimetallic Nanocluster Synthesis and Encapsulation Via Induced Zeolite Framework Demetallation for Shape and Substrate Selective Heterogeneous Catalysis. Angew. Chem. Int. Ed. 2018, 57 (22), 6454–6458. 26. Moliner, M.; Gabay, J.; Kliewer, C.; Serna, P.; Corma, A. Trapping of Metal Atoms and Metal Clusters by Chabazite under Severe Redox Stress. ACS Catal. 2018, 8 (10), 9520–9528. 27. Choi, M.; Wu, Z.; Iglesia, E. Mercaptosilane-Assisted Synthesis of Metal Clusters within Zeolites and Catalytic Consequences of Encapsulation. J. Am. Chem. Soc. 2010, 132 (26), 9129–9137. 28. Goel, S.; Wu, Z.; Zones, S. I.; Iglesia, E. Synthesis and Catalytic Properties of Metal Clusters Encapsulated within Small-Pore (SOD, GIS, ANA) Zeolites. J. Am. Chem. Soc. 2012, 134 (42), 17688–17695. 29. Goel, S.; Zones, S. I.; Iglesia, E. Encapsulation of Metal Clusters within MFI Via Interzeolite Transformations and Direct Hydrothermal Syntheses and Catalytic Consequences of Their Confinement. J. Am. Chem. Soc. 2014, 136 (43), 15280–15290. 30. Liu, L.; Diaz, U.; Arenal, R.; Agostini, G.; Concepcion, P.; Corma, A. Generation of Subnanometric Platinum with High Stability During Transformation of a 2D Zeolite into 3D. Nat. Mater. 2017, 16 (1), 132–138. 31. Liu, L.; Lopez-Haro, M.; Lopes, C. W.; Li, C.; Concepcion, P.; Simonelli, L.; Calvino, J. J.; Corma, A. Regioselective Generation and Reactivity Control of Subnanometric Platinum Clusters in Zeolites for High-Temperature Catalysis. Nat. Mater. 2019, 18 (8), 866–873. 32. Liu, L.; Lopez-Haro, M.; Meira, D. M.; Concepcion, P.; Calvino, J. J.; Corma, A. Regioselective Generation of Single-Site Iridium Atoms and Their Evolution into Stabilized Subnanometric Iridium Clusters in MWW Zeolite. Angew. Chem. Int. Ed. 2020, 59 (36), 15695–15702. 33. Cho, H. J.; Kim, D.; Li, J.; Su, D.; Xu, B. Zeolite-Encapsulated Pt Nanoparticles for Tandem Catalysis. J. Am. Chem. Soc. 2018, 140 (41), 13514–13520. 34. Wang, N.; Sun, Q.; Bai, R.; Li, X.; Guo, G.; Yu, J. In Situ Confinement of Ultrasmall Pd Clusters within Nanosized Silicalite-1 Zeolite for Highly Efficient Catalysis of Hydrogen Generation. J. Am. Chem. Soc. 2016, 138 (24), 7484–7487. 35. Sun, Q.; Wang, N.; Bing, Q.; Si, R.; Liu, J.; Bai, R.; Zhang, P.; Jia, M.; Yu, J. Subnanometric Hybrid Pd-M(OH)2, M ¼ Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts for Hydrogen Generation. Chem 2017, 3 (3), 477–493. 36. Sun, Q. M.; Wang, N.; Fan, Q. Y.; Zeng, L.; Mayoral, A.; Miao, S.; Yang, R. O.; Jiang, Z.; Zhou, W.; Zhang, J. C.; Zhang, T. J.; Xu, J.; Zhang, P.; Cheng, J.; Yang, D. C.; Jia, R.; Li, L.; Zhang, Q. H.; Wang, Y.; Terasaki, O.; Yu, J. H. Subnanometer Bimetallic Platinum-Zinc Clusters in Zeolites for Propane Dehydrogenation. Angew. Chem. Int. Ed. 2020, 59, 19450–19459. 37. Zhu, J.; Osuga, R.; Ishikawa, R.; Shibata, N.; Ikuhara, Y.; Kondo, J. N.; Ogura, M.; Yu, J. H.; Wakihara, T.; Liu, Z. D.; Okubo, T. Ultrafast Encapsulation of Metal Nanoclusters into MFI Zeolite in the Course of Its Crystallization: Catalytic Application for Propane Dehydrogenation. Angew. Chem. Int. Ed. 2020, 59, 19669–19674. 38. Wang, L. X.; Wang, L.; Meng, X. J.; Xiao, F. S. New Strategies for the Preparation of Sinter-Resistant Metal-Nanoparticle-Based Catalysts. Adv. Mater. 2019, 31 (50), 18. 39. Ren, L.; Wu, Q.; Yang, C.; Zhu, L.; Li, C.; Zhang, P.; Zhang, H.; Meng, X.; Xiao, F.-S. Solvent-Free Synthesis of Zeolites from Solid Raw Materials. J. Am. Chem. Soc. 2012, 134 (37), 15173–15176. 40. Wu, Q.; Wang, X.; Qi, G.; Guo, Q.; Pan, S.; Meng, X.; Xu, J.; Deng, F.; Fan, F.; Feng, Z.; Li, C.; Maurer, S.; Müller, U.; Xiao, F.-S. Sustainable Synthesis of Zeolites Without Addition of Both Organotemplates and Solvents. J. Am. Chem. Soc. 2014, 136 (10), 4019–4025. 41. Meng, X.; Xiao, F.-S. Green Routes for Synthesis of Zeolites. Chem. Rev. 2014, 114 (2), 1521–1543. 42. Jin, Y.; Sun, Q.; Qi, G.; Yang, C.; Xu, J.; Chen, F.; Meng, X.; Deng, F.; Xiao, F.-S. Solvent-Free Synthesis of Silicoaluminophosphate Zeolites. Angew. Chem. Int. Ed. 2013, 52 (35), 9172–9175. 43. Zhang, J.; Wang, L.; Zhu, L.; Wu, Q.; Chen, C.; Wang, X.; Ji, Y.; Meng, X.; Xiao, F.-S. Solvent-Free Synthesis of Zeolite Crystals Encapsulating Gold-Palladium Nanoparticles for the Selective Oxidation of Bioethanol. ChemSusChem 2015, 8 (17), 2867–2871. 44. Wang, C.; Wang, L.; Zhang, J.; Wang, H.; Lewis, J. P.; Xiao, F.-S. Product Selectivity Controlled by Zeolite Crystals in Biomass Hydrogenation over a Palladium Catalyst. J. Am. Chem. Soc. 2016, 138 (25), 7880–7883. 45. Xie, B.; Zhang, H.; Yang, C.; Liu, S.; Ren, L.; Zhang, L.; Meng, X.; Yilmaz, B.; Müller, U.; Xiao, F.-S. Seed-Directed Synthesis of Zeolites with Enhanced Performance in the Absence of Organic Templates. Chem. Commun. 2011, 47 (13), 3945–3947. 46. Zhang, J.; Wang, L.; Shao, Y.; Wang, Y.; Gates, B. C.; Xiao, F.-S. A Pd@Zeolite Catalyst for Nitroarene Hydrogenation with High Product Selectivity by Sterically Controlled Adsorption in the Zeolite Micropores. Angew. Chem. Int. Ed. Engl. 2017, 56 (33), 9747–9751. 47. Zhang, J.; Wang, L.; Zhang, B. S.; Zhao, H. S.; Kolb, U.; Zhu, Y. H.; Liu, L. M.; Han, Y.; Wang, G. X.; Wang, C. T.; Su, D. S.; Gates, B. C.; Xiao, F. S. Sinter-Resistant Metal Nanoparticle Catalysts Achieved by Immobilization Within Zeolite Crystals Via Seed-Directed Growth. Nat. Catal. 2018, 1 (7), 540–546.

Metal containing nanoclusters in zeolites

143

48. Wang, L.; Dai, J. J.; Xu, Y.; Hong, Y. L.; Huang, J. L.; Sun, D. H.; Li, Q. B. Titanium Silicalite-1 Zeolite Encapsulating Au Particles as a Catalyst for Vapor Phase Propylene Epoxidation with H2/O2: A Matter of Au-Ti Synergic Interaction. J. Mater. Chem. A 2020, 8 (8), 4428–4436. 49. Qiao, B.; Wang, A.; Yang, X.; Allard, L. F.; Jiang, Z.; Cui, Y.; Liu, J.; Li, J.; Zhang, T. Single-Atom Catalysis of CO Oxidation Using Pt1/FeOx. Nat. Chem. 2011, 3 (8), 634–641. 50. Wang, A.; Li, J.; Zhang, T. Heterogeneous Single-Atom Catalysis. Nat. Rev. Chem. 2018, 2 (6), 65–81. 51. Thomas, J. M. Tens of Thousands of Atoms Replaced by One. Nature 2015, 525 (7569), 325–326. 52. Liu, P.; Qin, R.; Fu, G.; Zheng, N. Surface Coordination Chemistry of Metal Nanomaterials. J. Am. Chem. Soc. 2017, 139 (6), 2122–2131. 53. Lin, L.; Zhou, W.; Gao, R.; Yao, S.; Zhang, X.; Xu, W.; Zheng, S.; Jiang, Z.; Yu, Q.; Li, Y.-W.; Shi, C.; Wen, X.-D.; Ma, D. Low-Temperature Hydrogen Production from Water and Methanol Using Pt/a-MoC Catalysts. Nature 2017, 544 (7648), 80–83. 54. He, X.; He, Q.; Deng, Y.; Peng, M.; Chen, H.; Zhang, Y.; Yao, S.; Zhang, M.; Xiao, D.; Ma, D.; Ge, B.; Ji, H. A Versatile Route to Fabricate Single Atom Catalysts with High Chemoselectivity and Regioselectivity in Hydrogenation. Nat. Commun. 2019, 10 (1), 3663. 55. Liu, J. Catalysis by Supported Single Metal Atoms. ACS Catal. 2017, 7 (1), 34–59. 56. Zhang, T.; Chen, Z.; Walsh, A. G.; Li, Y.; Zhang, P. Single-Atom Catalysts Supported by Crystalline Porous Materials: Views from the Inside. Adv. Mater. 2020;, 2002910. 57. Goellner, J. F.; Gates, B. C.; Vayssilov, G. N.; Rösch, N. Structure and Bonding of a Site-Isolated Transition Metal Complex: Rhodium Dicarbonyl in Highly Dealuminated Zeolite Y. J. Am. Chem. Soc. 2000, 122 (33), 8056–8066. 58. Liang, A. J.; Bhirud, V. A.; Ehresmann, J. O.; Kletnieks, P. W.; Haw, J. F.; Gates, B. C. A Site-Isolated Rhodium–Diethylene Complex Supported on Highly Dealuminated Y Zeolite: Synthesis and Characterization. J. Phys. Chem. B 2005, 109 (51), 24236–24243. 59. Ogino, I.; Gates, B. C. Molecular Chemistry in a Zeolite: Genesis of a Zeolite Y-Supported Ruthenium Complex Catalyst. J. Am. Chem. Soc. 2008, 130 (40), 13338–13346. 60. Uzun, A.; Bhirud, V. A.; Kletnieks, P. W.; Haw, J. F.; Gates, B. C. A Site-Isolated Iridium Diethylene Complex Supported on Highly Dealuminated Y Zeolite: Synthesis and Characterization. J. Phys. Chem. C 2007, 111 (41), 15064–15073. 61. Lu, J.; Serna, P.; Aydin, C.; Browning, N. D.; Gates, B. C. Supported Molecular Iridium Catalysts: Resolving Effects of Metal Nuclearity and Supports as Ligands. J. Am. Chem. Soc. 2011, 133 (40), 16186–16195. 62. Lu, J.; Aydin, C.; Browning, N. D.; Gates, B. C. Imaging Isolated Gold Atom Catalytic Sites in Zeolite NaY. Angew. Chem. Int. Ed. 2012, 51 (24), 5842–5846. 63. Paolucci, C.; Parekh, A. A.; Khurana, I.; Di Iorio, J. R.; Li, H.; Albarracin Caballero, J. D.; Shih, A. J.; Anggara, T.; Delgass, W. N.; Miller, J. T.; Ribeiro, F. H.; Gounder, R.; Schneider, W. F. Catalysis in a Cage: Condition-Dependent Speciation and Dynamics of Exchanged Cu Cations in SSZ-13 Zeolites. J. Am. Chem. Soc. 2016, 138 (18), 6028–6048. 64. Oda, A.; Torigoe, H.; Itadani, A.; Ohkubo, T.; Yumura, T.; Kobayashi, H.; Kuroda, Y. Unprecedented Reversible Redox Process in the ZnMFIdH2 System Involving Formation of Stable Atomic ZnO. Angew. Chem. Int. Ed. 2012, 51 (31), 7719–7723. 65. Phadke, N. M.; Van der Mynsbrugge, J.; Mansoor, E.; Getsoian, A. B.; Head-Gordon, M.; Bell, A. T. Characterization of Isolated Ga3þ Cations in Ga/H-MFI Prepared by VaporPhase Exchange of H-MFI Zeolite with GaCl3. ACS Catal. 2018, 8 (7), 6106–6126. 66. Chai, Y.; Wu, G.; Liu, X.; Ren, Y.; Dai, W.; Wang, C.; Xie, Z.; Guan, N.; Li, L. Acetylene-Selective Hydrogenation Catalyzed by Cationic Nickel Confined in Zeolite. J. Am. Chem. Soc. 2019, 141 (25), 9920–9927. 67. Kistler, J. D.; Chotigkrai, N.; Xu, P.; Enderle, B.; Praserthdam, P.; Chen, C.-Y.; Browning, N. D.; Gates, B. C. A Single-Site Platinum CO Oxidation Catalyst in Zeolite KLTL: Microscopic and Spectroscopic Determination of the Locations of the Platinum Atoms. Angew. Chem. Int. Ed. 2014, 53 (34), 8904–8907. 68. Tang, Y.; Li, Y.; Fung, V.; Jiang, D.-E.; Huang, W.; Zhang, S.; Iwasawa, Y.; Sakata, T.; Nguyen, L.; Zhang, X.; Frenkel, A. I.; Tao, F. Single Rhodium Atoms Anchored in Micropores for Efficient Transformation of Methane under Mild Conditions. Nat. Commun. 2018, 9 (1), 1231. 69. Maeno, Z.; Yasumura, S.; Wu, X.; Huang, M.; Liu, C.; Toyao, T.; Shimizu, K.-I. Isolated Indium Hydrides in CHA Zeolites: Speciation and Catalysis for Nonoxidative Dehydrogenation of Ethane. J. Am. Chem. Soc. 2020, 142 (10), 4820–4832. 70. Zhao, P.; Ye, L.; Sun, Z.; Lo, B. T. W.; Woodcock, H.; Huang, C.; Tang, C.; Kirkland, A. I.; Mei, D.; Edman Tsang, S. C. Entrapped Single Tungstate Site in Zeolite for Cooperative Catalysis of Olefin Metathesis with Brønsted Acid Site. J. Am. Chem. Soc. 2018, 140 (21), 6661–6667. 71. Yang, M.; Liu, J.; Lee, S.; Zugic, B.; Huang, J.; Allard, L. F.; Flytzani-Stephanopoulos, M. A Common Single-Site Pt(II)–O(OH)x–Species Stabilized by Sodium on “Active” and “Inert” Supports Catalyzes the Water-Gas Shift Reaction. J. Am. Chem. Soc. 2015, 137 (10), 3470–3473. 72. Yang, M.; Li, S.; Wang, Y.; Herron, J. A.; Xu, Y.; Allard, L. F.; Lee, S.; Huang, J.; Mavrikakis, M.; Flytzani-Stephanopoulos, M. Catalytically Active Au-O(OH)x-Species Stabilized by Alkali Ions on Zeolites and Mesoporous Oxides. Science 2014, 346 (6216), 1498–1501. 73. Zhai, Y.; Pierre, D.; Si, R.; Deng, W.; Ferrin, P.; Nilekar, A. U.; Peng, G.; Herron, J. A.; Bell, D. C.; Saltsburg, H.; Mavrikakis, M.; Flytzani-Stephanopoulos, M. Alkali-Stabilized Pt-OHx Species Catalyze Low-Temperature Water-Gas Shift Reactions. Science 2010, 329 (5999), 1633–1636. 74. Zugic, B.; Zhang, S.; Bell, D. C.; Tao, F.; Flytzani-Stephanopoulos, M. Probing the Low-Temperature Water–Gas Shift Activity of Alkali-Promoted Platinum Catalysts Stabilized on Carbon Supports. J. Am. Chem. Soc. 2014, 136 (8), 3238–3245. 75. Liu, Y.; Li, Z.; Yu, Q.; Chen, Y.; Chai, Z.; Zhao, G.; Liu, S.; Cheong, W.-C.; Pan, Y.; Zhang, Q.; Gu, L.; Zheng, L.; Wang, Y.; Lu, Y.; Wang, D.; Chen, C.; Peng, Q.; Liu, Y.; Liu, L.; Chen, J.; Li, Y. A General Strategy for Fabricating Isolated Single Metal Atomic Site Catalysts in Y Zeolite. J. Am. Chem. Soc. 2019, 141 (23), 9305–9311. 76. Sun, Q.; Wang, N.; Zhang, T.; Bai, R.; Mayoral, A.; Zhang, P.; Zhang, Q.; Terasaki, O.; Yu, J. Zeolite-Encaged Single-Atom Rhodium Catalysts: Highly-Efficient Hydrogen Generation and Shape-Selective Tandem Hydrogenation of Nitroarenes. Angew. Chem. Int. Ed. 2019, 58 (51), 18570–18576. 77. Qiu, J. Z.; Hu, J. B.; Lan, J. G.; Wang, L. F.; Fu, G. Y.; Xiao, R. J.; Ge, B. H.; Jiang, J. X. Pure Siliceous Zeolite-Supported Ru Single-Atom Active Sites for Ammonia Synthesis. Chem. Mater. 2019, 31 (22), 9413–9421. 78. Juneau, M.; Liu, R.; Peng, Y.; Malge, A.; Ma, Z.; Porosoff, M. D. Characterization of Metal-Zeolite Composite Catalysts: Determining the Environment of the Active Phase. ChemCatChem 2020, 12 (7), 1826–1852. 79. Ortalan, V.; Uzun, A.; Gates, B. C.; Browning, N. D. Direct Imaging of Single Metal Atoms and Clusters in the Pores of Dealuminated HY Zeolite. Nat. Nanotechnol. 2010, 5 (7), 506–510. 80. Browning, N. D.; Aydin, C.; Lu, J.; Kulkarni, A.; Okamoto, N. L.; Ortalan, V.; Reed, B. W.; Uzun, A.; Gates, B. C. Quantitative Z-Contrast Imaging of Supported Metal Complexes and ClustersA Gateway to Understanding Catalysis on the Atomic Scale. ChemCatChem 2013, 5 (9), 2673–2683. 81. Xu, P.; Lu, J.; Aydin, C.; Debefve, L. M.; Browning, N. D.; Chen, C.-Y.; Gates, B. C. Imaging Individual Lanthanum Atoms in Zeolite Y by Scanning Transmission Electron Microscopy: Evidence of Lanthanum Pair Sites. Microporous Mesoporous Mater. 2015, 213, 95–99. 82. Aydin, C.; Lu, J.; Shirai, M.; Browning, N. D.; Gates, B. C. Ir-6 Clusters Compartmentalized in the Supercages of Zeolite NaY: Direct Imaging of a Catalyst with AberrationCorrected Scanning Transmission Electron Microscopy. ACS Catal. 2011, 1 (11), 1613–1620. 83. Aydin, C.; Lu, J.; Liang, A. J.; Chen, C. Y.; Browning, N. D.; Gates, B. C. Tracking Iridium Atoms with Electron Microscopy: First Steps of Metal Nanocluster Formation in OneDimensional Zeolite Channels. Nano Lett. 2011, 11 (12), 5537–5541. 84. Schüßler, F.; Pidko, E. A.; Kolvenbach, R.; Sievers, C.; Hensen, E. J. M.; van Santen, R. A.; Lercher, J. A. Nature and Location of Cationic Lanthanum Species in High Alumina Containing Faujasite Type Zeolites. J. Phys. Chem. C 2011, 115 (44), 21763–21776. 85. Mayoral, A.; Readman, J. E.; Anderson, P. A. Aberration-Corrected STEM Analysis of a Cubic Cd Array Encapsulated in Zeolite A. J. Phys. Chem. C 2013, 117 (46), 24485– 24489. 86. Mayoral, A.; Carey, T.; Anderson, P. A.; Diaz, I. Atomic Resolution Analysis of Porous Solids: A Detailed Study of Silver Ion-Exchanged Zeolite A. Microporous Mesoporous Mater. 2013, 166, 117–122. 87. Mayoral, A.; Carey, T.; Anderson, P. A.; Lubk, A.; Diaz, I. Atomic Resolution Analysis of Silver Ion-Exchanged Zeolite A. Angew. Chem. Int. Ed. 2011, 50 (47), 11230–11233.

144

Metal containing nanoclusters in zeolites

88. Filippousi, M.; Turner, S.; Katsikini, M.; Pinakidou, F.; Zamboulis, D.; Pavlidou, E.; Van Tendeloo, G. Direct Observation and Structural Characterization of Natural and Metal Ion-Exchanged HEU-Type Zeolites. Microporous Mesoporous Mater. 2015, 210, 185–193. 89. Altantzis, T.; Coutino-Gonzalez, E.; Baekelant, W.; Martinez, G. T.; Abakumov, A. M.; Van Tendeloo, G.; Roeffaers, M. B. J.; Bals, S.; Hofkens, J. Direct Observation of Luminescent Silver Clusters Confined in Faujasite Zeolites. ACS Nano 2016, 10 (8), 7604–7611. 90. Mayoral, A.; Hall, R. M.; Jackowska, R.; Readman, J. E. Imaging the Atomic Position of Light Cations in a Porous Network and the Europium(III) Ion Exchange Capability by Aberration-Corrected Electron Microscopy. Angew. Chem. Int. Ed. 2016, 55 (52), 16127–16131. 91. Mayoral, A.; Zhang, Q.; Zhou, Y.; Chen, P.; Ma, Y.; Monji, T.; Losch, P.; Schmidt, W.; Schüth, F.; Hirao, H.; Yu, J.; Terasaki, O. Direct Atomic-Level Imaging of Zeolites: Oxygen, Sodium in Na-LTA and Iron in Fe-MFI. Angew. Chem. Int. Ed. 2020, 59 (44), 19510–19517. 92. Liu, L.; Lopez-Haro, M.; Calvino, J. J.; Corma, A. Tutorial: Structural Characterization of Isolated Metal Atoms and Subnanometric Metal Clusters in Zeolites. Nat. Protoc. 2020, 16 (4), 1871–1906. 93. Liu, L. M.; Wang, N.; Zhu, C. Z.; Liu, X. N.; Zhu, Y. H.; Guo, P.; Alfilfil, L.; Dong, X. L.; Zhang, D. L.; Han, Y. Direct Imaging of Atomically Dispersed Molybdenum that Enables Location of Aluminum in the Framework of Zeolite ZSM-5. Angew. Chem. Int. Ed. 2020, 59 (2), 819–825. 94. Shen, B. Y.; Chen, X.; Cai, D. L.; Xiong, H.; Liu, X.; Meng, C. G.; Han, Y.; Wei, F. Atomic Spatial and Temporal Imaging of Local Structures and Light Elements Inside Zeolite Frameworks. Adv. Mater. 2019, 32 (4), 1906103. 95. Fodor, D.; Ishikawa, T.; Krumeich, F.; van Bokhoven, J. A. Synthesis of Single Crystal Nanoreactor Materials with Multiple Catalytic Functions by Incipient Wetness Impregnation and Ion Exchange. Adv. Mater. 2015, 27 (11), 1919–1923. 96. Friedrich, H.; de Jongh, P. E.; Verkleij, A. J.; de Jong, K. P. Electron Tomography for Heterogeneous Catalysts and Related Nanostructured Materials. Chem. Rev. 2009, 109 (5), 1613–1629. 97. Sushkevich, V. L.; Palagin, D.; Ranocchiari, M.; van Bokhoven, J. A. Selective Anaerobic Oxidation of Methane Enables Direct Synthesis of Methanol. Science 2017, 356 (6337), 523–527. 98. Paolucci, C.; Khurana, I.; Parekh, A. A.; Li, S.; Shih, A. J.; Li, H.; Di Iorio, J. R.; Albarracin-Caballero, J. D.; Yezerets, A.; Miller, J. T.; Delgass, W. N.; Ribeiro, F. H.; Schneider, W. F.; Gounder, R. Dynamic Multinuclear Sites Formed by Mobilized Copper Ions in NOx Selective Catalytic Reduction. Science 2017, 357 (6354), 898–903. 99. Kerkeni, B.; Berthout, D.; Berthomieu, D.; Doronkin, D. E.; Casapu, M.; Grunwaldt, J. D.; Chizallet, C. Copper Coordination to Water and Ammonia in CuII-Exchanged SSZ-13: Atomistic Insights from DFT Calculations and In Situ XAS Experiments. J. Phys. Chem. C 2018, 122 (29), 16741–16755. 100. Negri, C.; Selleri, T.; Borfecchia, E.; Martini, A.; Lomachenko, K. A.; Janssens, T. V. W.; Cutini, M.; Bordiga, S.; Berlier, G. Structure and Reactivity of Oxygen-Bridged Diamino Dicopper(II) Complexes in Cu-Ion-Exchanged Chabazite Catalyst for NH3-Mediated Selective Catalytic Reduction. J. Am. Chem. Soc. 2020, 142 (37), 15884–15896. 101. Becher, J.; Sanchez, D. F.; Doronkin, D. E.; Zengel, D.; Meira, D. M.; Pascarelli, S.; Grunwaldt, J.-D.; Sheppard, T. L. Chemical Gradients in Automotive Cu-SSZ-13 Catalysts for NOx Removal Revealed by Operando X-Ray Spectrotomography. Nat. Catal. 2021, 4 (1), 46–53. 102. Borfecchia, E.; Beato, P.; Svelle, S.; Olsbye, U.; Lamberti, C.; Bordiga, S. Cu-CHAdA Model System for Applied Selective Redox Catalysis. Chem. Soc. Rev. 2018, 47 (22), 8097–8133. 103. Shan, Y.; Du, J.; Zhang, Y.; Shan, W.; Shi, X.; Yu, Y.; Zhang, R.; Meng, X.; Xiao, F.-S.; He, H. Selective Catalytic Reduction of NOx with NH3: Opportunities and Challenges of Cu-Based Small-Pore Zeolites. Natl. Sci. Rev. 2021;, nwab010. 104. Paolucci, C.; Di Iorio, J. R.; Schneider, W. F.; Gounder, R. Solvation and Mobilization of Copper Active Sites in Zeolites by Ammonia: Consequences for the Catalytic Reduction of Nitrogen Oxides. Acc. Chem. Res. 2020, 53 (9), 1881–1892. 105. Pidko, E. A.; Hensen, E. J. M.; van Santen, R. A. Self-Organization of Extraframework Cations in Zeolites. Proc. R. Soc. A Math. Phys. Eng. Sci. 2012, 468 (2143), 2070–2086. 106. Dinh, K. T.; Sullivan, M. M.; Narsimhan, K.; Serna, P.; Meyer, R. J.; Dinca, M.; Román-Leshkov, Y. Continuous Partial Oxidation of Methane to Methanol Catalyzed by Diffusion-Paired Copper Dimers in Copper-Exchanged Zeolites. J. Am. Chem. Soc. 2019, 141 (29), 11641–11650. 107. Song, J.; Wang, Y.; Walter, E. D.; Washton, N. M.; Mei, D.; Kovarik, L.; Engelhard, M. H.; Prodinger, S.; Wang, Y.; Peden, C. H. F.; Gao, F. Toward Rational Design of Cu/SSZ13 Selective Catalytic Reduction Catalysts: Implications from Atomic-Level Understanding of Hydrothermal Stability. ACS Catal. 2017, 7 (12), 8214–8227. 108. Kiani, D.; Sourav, S.; Tang, Y.; Baltrusaitis, J.; Wachs, I. E. Methane Activation by ZSM-5-Supported Transition Metal Centers. Chem. Soc. Rev. 2021, 50 (2), 1251–1268. 109. Kosinov, N.; Hensen, E. J. M. Reactivity, Selectivity, and Stability of Zeolite-Based Catalysts for Methane Dehydroaromatization. Adv. Mater. 2020, 32 (44), 2002565. 110. Lezcano-González, I.; Oord, R.; Rovezzi, M.; Glatzel, P.; Botchway, S. W.; Weckhuysen, B. M.; Beale, A. M. Molybdenum Speciation and Its Impact on Catalytic Activity During Methane Dehydroaromatization in Zeolite ZSM-5 as Revealed by Operando X-Ray Methods. Angew. Chem. Int. Ed. 2016, 55 (17), 5215–5219. 111. Vollmer, I.; Yarulina, I.; Kapteijn, F.; Gascon, J. Progress in Developing a Structure-Activity Relationship for the Direct Aromatization of Methane. ChemCatChem 2019, 11 (1), 39–52. 112. Agote-Arán, M.; Kroner, A. B.; Islam, H. U.; Sławinski, W. A.; Wragg, D. S.; Lezcano-González, I.; Beale, A. M. Determination of Molybdenum Species Evolution During NonOxidative Dehydroaromatization of Methane and Its Implications for Catalytic Performance. ChemCatChem 2019, 11 (1), 473–480. 113. Kosinov, N.; Coumans, F.; Uslamin, E.; Kapteijn, F.; Hensen, E. J. M. Selective Coke Combustion by Oxygen Pulsing During Mo/ZSM-5-Catalyzed Methane Dehydroaromatization. Angew. Chem. Int. Ed. 2016, 55 (48), 15086–15090. 114. Kosinov, N.; Coumans, F.; Li, G.; Uslamin, E.; Mezari, B.; Wijpkema, A. S. G.; Pidko, E. A.; Hensen, E. J. M. Stable Mo/HZSM-5 Methane Dehydroaromatization Catalysts Optimized for High-Temperature Calcination-Regeneration. J. Catal. 2017, 346, 125–133. 115. Kosinov, N.; Coumans, F. J. A. G.; Uslamin, E. A.; Wijpkema, A. S. G.; Mezari, B.; Hensen, E. J. M. Methane Dehydroaromatization by Mo/HZSM-5: Mono- or Bifunctional Catalysis? ACS Catal. 2017, 7 (1), 520–529. 116. Kosinov, N.; Wijpkema, A. S. G.; Uslamin, E.; Rohling, R.; Coumans, F. J. A. G.; Mezari, B.; Parastaev, A.; Poryvaev, A. S.; Fedin, M. V.; Pidko, E. A.; Hensen, E. J. M. Confined Carbon Mediating Dehydroaromatization of Methane over Mo/ZSM-5. Angew. Chem. Int. Ed. 2018, 57 (4), 1016–1020. 117. Kosinov, N.; Uslamin, E. A.; Meng, L.; Parastaev, A.; Liu, Y.; Hensen, E. J. M. Reversible Nature of Coke Formation on Mo/ZSM-5 Methane Dehydroaromatization Catalysts. Angew. Chem. Int. Ed. 2019, 58 (21), 7068–7072. 118. Agote-Arán, M.; Kroner, A. B.; Wragg, D. S.; Sławinski, W. A.; Briceno, M.; Islam, H. U.; Sazanovich, I. V.; Rivas, M. E.; Smith, A. W. J.; Collier, P.; Lezcano-González, I.; Beale, A. M. Understanding the Deactivation Phenomena of Small-Pore Mo/H-SSZ-13 During Methane Dehydroaromatisation. Molecules 2020, 25 (21), 5048. 119. Vollmer, I.; van der Linden, B.; Ould-Chikh, S.; Aguilar-Tapia, A.; Yarulina, I.; Abou-Hamad, E.; Sneider, Y. G.; Olivos Suarez, A. I.; Hazemann, J.-L.; Kapteijn, F.; Gascon, J. On the Dynamic Nature of Mo Sites for Methane Dehydroaromatization. Chem. Sci. 2018, 9 (21), 4801–4807. 120. Moliner, M.; Gabay, J. E.; Kliewer, C. E.; Carr, R. T.; Guzman, J.; Casty, G. L.; Serna, P.; Corma, A. Reversible Transformation of Pt Nanoparticles into Single Atoms Inside High-Silica Chabazite Zeolite. J. Am. Chem. Soc. 2016, 138 (48), 15743–15750. 121. Liu, L.; Zakharov, D. N.; Arenal, R.; Concepcion, P.; Stach, E. A.; Corma, A. Evolution and Stabilization of Subnanometric Metal Species in Confined Space by In Situ TEM. Nat. Commun. 2018, 9, 574. 122. Eremin, D. B.; Ananikov, V. P. Understanding Active Species in Catalytic Transformations: From Molecular Catalysis to Nanoparticles, Leaching, “Cocktails” of Catalysts and Dynamic Systems. Coord. Chem. Rev. 2017, 346, 2–19. 123. Bordiga, S.; Lamberti, C.; Bonino, F.; Travert, A.; Thibault-Starzyk, F. Probing Zeolites by Vibrational Spectroscopies. Chem. Soc. Rev. 2015, 44 (20), 7262–7341. 124. Woertink, J. S.; Smeets, P. J.; Groothaert, M. H.; Vance, M. A.; Sels, B. F.; Schoonheydt, R. A.; Solomon, E. I. A [Cu2O]2þ Core in Cu-ZSM-5, the Active Site in the Oxidation of Methane to Methanol. Proc. Natl. Acad. Sci. U. S. A. 2009, 106 (45), 18908–18913. 125. Groothaert, M. H.; Smeets, P. J.; Sels, B. F.; Jacobs, P. A.; Schoonheydt, R. A. Selective Oxidation of Methane by the Bis(mu-Oxo)Dicopper Core Stabilized on ZSM-5 and Mordenite Zeolites. J. Am. Chem. Soc. 2005, 127 (5), 1394–1395.

Metal containing nanoclusters in zeolites

145

126. Vanelderen, P.; Snyder, B. E. R.; Tsai, M.-L.; Hadt, R. G.; Vancauwenbergh, J.; Coussens, O.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. Spectroscopic Definition of the Copper Active Sites in Mordenite: Selective Methane Oxidation. J. Am. Chem. Soc. 2015, 137 (19), 6383–6392. 127. Snyder, B. E. R.; Vanelderen, P.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. Second-Sphere Effects on Methane Hydroxylation in Cu-Zeolites. J. Am. Chem. Soc. 2018, 140 (29), 9236–9243. 128. Smeets, P. J.; Hadt, R. G.; Woertink, J. S.; Vanelderen, P.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. Oxygen Precursor to the Reactive Intermediate in Methanol Synthesis by Cu-ZSM-5. J. Am. Chem. Soc. 2010, 132 (42), 14736–14738. 129. Pappas, D. K.; Borfecchia, E.; Dyballa, M.; Pankin, I. A.; Lomachenko, K. A.; Martini, A.; Signorile, M.; Teketel, S.; Arstad, B.; Berlier, G.; Lamberti, C.; Bordiga, S.; Olsbye, U.; Lillerud, K. P.; Svelle, S.; Beato, P. Methane to Methanol: Structure Activity Relationships for Cu-CHA. J. Am. Chem. Soc. 2017, 139 (42), 14961–14975. 130. Ipek, B.; Wulfers, M. J.; Kim, H.; Göltl, F.; Hermans, I.; Smith, J. P.; Booksh, K. S.; Brown, C. M.; Lobo, R. F. Formation of [Cu2O2]2þ and [Cu2O]2þ toward C–H Bond Activation in Cu-SSZ-13 and Cu-SSZ-39. ACS Catal. 2017, 7 (7), 4291–4303. 131. Gao, J.; Zheng, Y. T.; Tang, Y. D.; Jehng, J. M.; Grybos, R.; Handzlik, J.; Wachs, I. E.; Podkolzin, S. G. Spectroscopic and Computational Study of Cr Oxide Structures and Their Anchoring Sites on ZSM-5 Zeolites. ACS Catal. 2015, 5 (5), 3078–3092. 132. Gao, J.; Zheng, Y. T.; Jehng, J. M.; Tang, Y. D.; Wachs, I. E.; Podkolzin, S. G. Identification of Molybdenum Oxide Nanostructures on Zeolites for Natural Gas Conversion. Science 2015, 348 (6235), 686–690. 133. Göltl, F.; Conrad, S.; Wolf, P.; Müller, P.; Love, A. M.; Burt, S. P.; Wheeler, J. N.; Hamers, R. J.; Hummer, K.; Kresse, G.; Mavrikakis, M.; Hermans, I. UV–Vis and Photoluminescence Spectroscopy to Understand the Coordination of Cu Cations in the Zeolite SSZ-13. Chem. Mater. 2019, 31 (23), 9582–9592. 134. Li, H.; Paolucci, C.; Khurana, I.; Wilcox, L. N.; Göltl, F.; Albarracin-Caballero, J. D.; Shih, A. J.; Ribeiro, F. H.; Gounder, R.; Schneider, W. F. Consequences of Exchange-Site Heterogeneity and Dynamics on the UV-Visible Spectrum of Cu-Exchanged SSZ-13. Chem. Sci. 2019, 10 (8), 2373–2384. 135. Vimont, A.; Thibault-Starzyk, F.; Daturi, M. Analysing and Understanding the Active Site by IR Spectroscopy. Chem. Soc. Rev. 2010, 39 (12), 4928–4950. 136. Khivantsev, K.; Jaegers, N. R.; Kovarik, L.; Hanson, J. C.; Tao, F.; Tang, Y.; Zhang, X.; Koleva, I. Z.; Aleksandrov, H. A.; Vayssilov, G. N.; Wang, Y.; Gao, F.; Szanyi, J. Achieving Atomic Dispersion of Highly Loaded Transition Metals in Small-Pore Zeolite SSZ-13: High-Capacity and High-Efficiency Low-Temperature CO and Passive NOx Adsorbers. Angew. Chem. Int. Ed. 2018, 57 (51), 16672–16677. 137. Kwak, J. H.; Varga, T.; Peden, C. H. F.; Gao, F.; Hanson, J. C.; Szanyi, J. Following the Movement of Cu Ions in a SSZ-13 Zeolite during Dehydration, Reduction and Adsorption: A Combined In Situ TP-XRD, XANES/DRIFTS Study. J. Catal. 2014, 314, 83–93. 138. Zaera, F. New Advances in the Use of Infrared Absorption Spectroscopy for the Characterization of Heterogeneous Catalytic Reactions. Chem. Soc. Rev. 2014, 43 (22), 7624–7663. 139. Xu, J.; Wang, Q.; Deng, F. Metal Active Sites and Their Catalytic Functions in Zeolites: Insights from Solid-State NMR Spectroscopy. Acc. Chem. Res. 2019, 52 (8), 2179–2189. 140. Li, S.; Lafon, O.; Wang, W.; Wang, Q.; Wang, X.; Li, Y.; Xu, J.; Deng, F. Recent Advances of Solid-State NMR Spectroscopy for Microporous Materials. Adv. Mater. 2020, 32 (44), 2002879. 141. Yu, Z.; Zheng, A.; Wang, Q.; Chen, L.; Xu, J.; Amoureux, J.-P.; Deng, F. Insights into the Dealumination of Zeolite HY Revealed by Sensitivity-Enhanced 27Al DQ-MAS NMR Spectroscopy at High Field. Angew. Chem. Int. Ed. 2010, 49 (46), 8657–8661. 142. Wang, X.; Qi, G.; Xu, J.; Li, B.; Wang, C.; Deng, F. NMR-Spectroscopic Evidence of Intermediate-Dependent Pathways for Acetic Acid Formation from Methane and Carbon Monoxide over a ZnZSM-5 Zeolite Catalyst. Angew. Chem. Int. Ed. 2012, 51 (16), 3850–3853. 143. Wu, J.-F.; Yu, S.-M.; Wang, W. D.; Fan, Y.-X.; Bai, S.; Zhang, C.-W.; Gao, Q.; Huang, J.; Wang, W. Mechanistic Insight into the Formation of Acetic Acid from the Direct Conversion of Methane and Carbon Dioxide on Zinc-Modified H–ZSM-5 Zeolite. J. Am. Chem. Soc. 2013, 135 (36), 13567–13573. 144. Wu, X.; Xu, S.; Zhang, W.; Huang, J.; Li, J.; Yu, B.; Wei, Y.; Liu, Z. Direct Mechanism of the First Carbon–Carbon Bond Formation in the Methanol-to-Hydrocarbons Process. Angew. Chem. Int. Ed. 2017, 56 (31), 9039–9043. 145. Zhao, Z.; Shi, H.; Wan, C.; Hu, M. Y.; Liu, Y.; Mei, D.; Camaioni, D. M.; Hu, J. Z.; Lercher, J. A. Mechanism of Phenol Alkylation in Zeolite H-BEA Using In Situ Solid-State NMR Spectroscopy. J. Am. Chem. Soc. 2017, 139 (27), 9178–9185. 146. Jaegers, N. R.; Mueller, K. T.; Wang, Y.; Hu, J. Z. Variable Temperature and Pressure Operando MAS NMR for Catalysis Science and Related Materials. Acc. Chem. Res. 2020, 53 (3), 611–619. 147. Zheng, H.; Ma, D.; Bao, X.; Hu, J. Z.; Kwak, J. H.; Wang, Y.; Peden, C. H. F. Direct Observation of the Active Center for Methane Dehydroaromatization Using an Ultrahigh Field 95Mo NMR Spectroscopy. J. Am. Chem. Soc. 2008, 130 (12), 3722–3723. 148. Hu, J. Z.; Kwak, J. H.; Wang, Y.; Peden, C. H. F.; Zheng, H.; Ma, D.; Bao, X. Studies of the Active Sites for Methane Dehydroaromatization Using Ultrahigh-Field Solid-State 95 Mo NMR Spectroscopy. J. Phys. Chem. C 2009, 113 (7), 2936–2942. 149. Qi, G.; Wang, Q.; Xu, J.; Trébosc, J.; Lafon, O.; Wang, C.; Amoureux, J.-P.; Deng, F. Synergic Effect of Active Sites in Zinc-Modified ZSM-5 Zeolites as Revealed by High-Field Solid-State NMR Spectroscopy. Angew. Chem. Int. Ed. 2016, 55 (51), 15826–15830. 150. Gao, P.; Wang, Q.; Xu, J.; Qi, G.; Wang, C.; Zhou, X.; Zhao, X.; Feng, N.; Liu, X.; Deng, F. Brønsted/Lewis Acid Synergy in Methanol-to-Aromatics Conversion on Ga-Modified ZSM-5 Zeolites, as Studied by Solid-State NMR Spectroscopy. ACS Catal. 2018, 8 (1), 69–74. 151. Zhao, X.; Chu, Y.; Qi, G.; Wang, Q.; Gao, W.; Wang, X.; Li, S.; Xu, J.; Deng, F. Probing the Active Sites for Methane Activation on Ga/ZSM-5 Zeolites with Solid-State NMR Spectroscopy. Chem. Commun. 2020, 56 (80), 12029–12032. 152. Cai, Y.; Kumar, R.; Huang, W.; Trewyn, B. G.; Wiench, J. W.; Pruski, M.; Lin, V. S. Y. Mesoporous Aluminum Silicate Catalyst with Single-Type Active Sites: Characterization by Solid-State NMR and Studies of Reactivity for Claisen Rearrangement Reactions. J. Phys. Chem. C 2007, 111 (3), 1480–1486. 153. Sklenak, S.; Dedecek, J.; Li, C.; Wichterlová, B.; Gábová, V.; Sierka, M.; Sauer, J. Aluminum Siting in Silicon-Rich Zeolite Frameworks: A Combined High-Resolution 27Al NMR Spectroscopy and Quantum Mechanics/Molecular Mechanics Study of ZSM-5. Angew. Chem. Int. Ed. 2007, 46 (38), 7286–7289. 154. Brown, S. P.; Spiess, H. W. Advanced Solid-State NMR Methods for the Elucidation of Structure and Dynamics of Molecular, Macromolecular, and Supramolecular Systems. Chem. Rev. 2001, 101 (12), 4125–4156. 155. Yi, X.; Liu, K.; Chen, W.; Li, J.; Xu, S.; Li, C.; Xiao, Y.; Liu, H.; Guo, X.; Liu, S.-B.; Zheng, A. Origin and Structural Characteristics of Tri-coordinated Extra-framework Aluminum Species in Dealuminated Zeolites. J. Am. Chem. Soc. 2018, 140 (34), 10764–10774. 156. Wang, Z.; O’Dell, L. A.; Zeng, X.; Liu, C.; Zhao, S.; Zhang, W.; Gaborieau, M.; Jiang, Y.; Huang, J. Insight into Three-Coordinate Aluminum Species on Ethanol-to-Olefin Conversion over ZSM-5 Zeolites. Angew. Chem. Int. Ed. 2019, 58 (50), 18061–18068. 157. Malicki, N.; Mali, G.; Quoineaud, A.-A.; Bourges, P.; Simon, L. J.; Thibault-Starzyk, F.; Fernandez, C. Aluminium Triplets in Dealuminated Zeolites Detected by 27Al NMR Correlation Spectroscopy. Microporous Mesoporous Mater. 2010, 129 (1), 100–105. 158. Popovych, N.; Kyriienko, P.; Soloviev, S.; Baran, R.; Millot, Y.; Dzwigaj, S. Identification of the Silver State in the Framework of Ag-Containing Zeolite by XRD, FTIR, Photoluminescence, 109Ag NMR, EPR, DR UV-Vis, TEM and XPS Investigations. Phys. Chem. Chem. Phys. 2016, 18 (42), 29458–29465. 159. Cruz, P.; Fajardo, M.; del Hierro, I.; Pérez, Y. Selective Oxidation of Thioanisole by Titanium Complexes Immobilized on Mesoporous Silica Nanoparticles: Elucidating the Environment of Titanium(iv) Species. Cat. Sci. Technol. 2019, 9 (3), 620–633. 160. Gordon, C. P.; Engler, H.; Tragl, A. S.; Plodinec, M.; Lunkenbein, T.; Berkessel, A.; Teles, J. H.; Parvulescu, A.-N.; Copéret, C. Efficient Epoxidation over Dinuclear Sites in Titanium Silicalite-1. Nature 2020, 586 (7831), 708–713. 161. Pei, S.; Zajac, G. W.; Kaduk, J. A.; Faber, J.; Boyanov, B. I.; Duck, D.; Fazzini, D.; Morrison, T. I.; Yang, D. S. Re-investigation of Titanium Silicalite by X-Ray Absorption Spectroscopy: Are the Novel Titanium Sites Real? Catal. Lett. 1993, 21 (3), 333–344. 162. Weckhuysen, B. M. Fresh Evidence Challenges the Consensus View of Active Sites in an Industrial Catalyst. Nature 2020, 586 (7831), 678–679.

146

Metal containing nanoclusters in zeolites

163. Peng, B.; Yao, Y.; Zhao, C.; Lercher, J. A. Towards Quantitative Conversion of Microalgae Oil to Diesel-Range Alkanes with Bifunctional Catalysts. Angew. Chem. Int. Ed. 2012, 51 (9), 2072–2075. 164. Kim, J.; Kim, W.; Seo, Y.; Kim, J.-C.; Ryoo, R. n-Heptane Hydroisomerization over Pt/MFI Zeolite Nanosheets: Effects of Zeolite Crystal Thickness and Platinum Location. J. Catal. 2013, 301, 187–197. 165. Zhang, Q.; Yu, J.; Corma, A. Applications of Zeolites to C1 Chemistry: Recent Advances, Challenges, and Opportunities. Adv. Mater. 2020, 32 (44), 2002927. 166. Jovanovic, Z. R.; Lange, J.-P.; Ravi, M.; Knorpp, A. J.; Sushkevich, V. L.; Newton, M. A.; Palagin, D.; van Bokhoven, J. A. Oxidation of Methane to Methanol over CuExchanged Zeolites: Scientia Gratia Scientiae or Paradigm Shift in Natural Gas Valorization? J. Catal. 2020, 385, 238–245. 167. Burnett, L.; Rysakova, M.; Wang, K.; González-Carballo, J.; Tooze, R. P.; García-García, F. R. Isothermal Cyclic Conversion of Methane to Methanol Using Copper-Exchanged ZSM-5 Zeolite Materials under Mild Conditions. Appl. Catal. A 2019, 587, 117272. 168. Zhao, G.; Benhelal, E.; Adesina, A.; Kennedy, E.; Stockenhuber, M. Comparison of Direct, Selective Oxidation of Methane by N2O over Fe-ZSM-5, Fe-Beta, and Fe-FER Catalysts. J. Phys. Chem. C 2019, 123 (45), 27436–27447. 169. Newton, M. A.; Knorpp, A. J.; Sushkevich, V. L.; Palagin, D.; van Bokhoven, J. A. Active Sites and Mechanisms in the Direct Conversion of Methane to Methanol Using Cu in Zeolitic Hosts: A Critical Examination. Chem. Soc. Rev. 2020, 49 (5), 1449–1486. 170. Grundner, S.; Markovits, M. A. C.; Li, G.; Tromp, M.; Pidko, E. A.; Hensen, E. J. M.; Jentys, A.; Sanchez-Sanchez, M.; Lercher, J. A. Single-Site Trinuclear Copper Oxygen Clusters in Mordenite for Selective Conversion of Methane to Methanol. Nat. Commun. 2015, 6, 7546. 171. Sushkevich, V. L.; van Bokhoven, J. A. Kinetic Study and Effect of Water on Methane Oxidation to Methanol Over Copper-Exchanged Mordenite. Cat. Sci. Technol. 2020, 10 (2), 382–390. 172. Tomkins, P.; Mansouri, A.; Bozbag, S. E.; Krumeich, F.; Park, M. B.; Alayon, E. M. C.; Ranocchiari, M.; van Bokhoven, J. A. Isothermal Cyclic Conversion of Methane into Methanol over Copper-Exchanged Zeolite at Low Temperature. Angew. Chem. Int. Ed. 2016, 55, 5467–5471. 173. Koishybay, A.; Shantz, D. F. Water Is the Oxygen Source for Methanol Produced in Partial Oxidation of Methane in a Flow Reactor over Cu-SSZ-13. J. Am. Chem. Soc. 2020, 142 (28), 11962–11966. 174. Jin, Z.; Wang, L.; Zuidema, E.; Mondal, K.; Zhang, M.; Zhang, J.; Wang, C.; Meng, X.; Yang, H.; Mesters, C.; Xiao, F.-S. Hydrophobic Zeolite Modification for In Situ Peroxide Formation in Methane Oxidation to Methanol. Science 2020, 367 (6474), 193–197. 175. Martinez Galeano, Y.; Negri, F.; Sergio Moreno, M.; Munera, J.; Cornaglia, L.; Tarditi, A. M. Pt Encapsulated into NaA Zeolite as Catalyst for the WGS Reaction. Appl. Catal. A 2019, 572, 176–184. 176. Wang, C. T.; Guan, E.; Wang, L.; Chu, X. F.; Wu, Z. Y.; Zhang, J.; Yang, Z. Y.; Jiang, Y. W.; Zhang, L.; Meng, X. J.; Gates, B. C.; Xiao, F. S. Product Selectivity Controlled by Nanoporous Environments in Zeolite Crystals Enveloping Rhodium Nanoparticle Catalysts for CO2 Hydrogenation. J. Am. Chem. Soc. 2019, 141 (21), 8482–8488. 177. Wang, C.; Zhang, J.; Qin, G.; Wang, L.; Zuidema, E.; Yang, Q.; Dang, S.; Yang, C.; Xiao, J.; Meng, X.; Mesters, C.; Xiao, F.-S. Direct Conversion of Syngas to Ethanol within Zeolite Crystals. Chem 2020, 6 (3), 646–657. 178. Wang, Z.; Kim, K.-D.; Zhou, C.; Chen, M.; Maeda, N.; Liu, Z.; Shi, J.; Baiker, A.; Hunger, M.; Huang, J. Influence of Support Acidity on the Performance of Size-Confined Pt Nanoparticles in the Chemoselective Hydrogenation of Acetophenone. Cat. Sci. Technol. 2015, 5 (5), 2788–2797. 179. Sun, Q.; Wang, N.; Bai, R.; Hui, Y.; Zhang, T.; Do, D. A.; Zhang, P.; Song, L.; Miao, S.; Yu, J. Synergetic Effect of Ultrasmall Metal Clusters and Zeolites Promoting Hydrogen Generation. Adv. Sci. 2019, 6 (10), 1802350. 180. Cho, H. J.; Kim, D.; Li, S.; Su, D.; Ma, D.; Xu, B. Molecular-Level Proximity of Metal and Acid Sites in Zeolite-Encapsulated Pt Nanoparticles for Selective Multistep Tandem Catalysis. ACS Catal. 2020, 10 (5), 3340–3348. 181. Zhang, B.; Tian, Y.; Chen, D.; Li, L.; Li, G.; Wang, L.; Zhang, X.; Liu, G. Selective Steam Reforming of n-Dodecane over Stable Subnanometric NiPt Clusters Encapsulated in Silicalite-1 Zeolite. AIChE J. 2020, 66 (5), e16917. 182. Wang, Y.; Hu, Z.-P.; Lv, X.; Chen, L.; Yuan, Z.-Y. Ultrasmall PtZn Bimetallic Nanoclusters Encapsulated in Silicalite-1 Zeolite with Superior Performance for Propane Dehydrogenation. J. Catal. 2020, 385, 61–69. 183. Chen, Q.; Wang, M.; Zhang, C.; Ren, K.; Xin, Y.; Zhao, M.; Xing, E. Selectivity Control on Hydrogenation of Substituted Nitroarenes through End-on Adsorption of Reactants in Zeolite-Encapsulated Platinum Nanoparticles. Chem. Asian J. 2018, 13 (16), 2077–2084. 184. Chen, Q.; Kang, H. Z.; Liu, X.; Jiang, K.; Bi, Y. F.; Zhou, Y. M.; Wang, M. Y.; Zhang, M.; Liu, L.; Xing, E. H. Selective Hydrogenation of Aromatic Ketone over Pt@Y Zeolite Through Restricted Adsorption Conformation of Reactants by Zeolitic Micropores. ChemCatChem 2020, 12 (7), 1948–1952. 185. Vercammen, J.; Bocus, M.; Neale, S.; Bugaev, A.; Tomkins, P.; Hajek, J.; Van Minnebruggen, S.; Soldatov, A.; Krajnc, A.; Mali, G.; Van Speybroeck, V.; De Vos, D. ShapeSelective C–H Activation of Aromatics to Biarylic Compounds Using Molecular Palladium in Zeolites. Nat. Catal. 2020, 3 (12), 1002–1009. 186. Chai, Y.; Liu, S.; Zhao, Z.-J.; Gong, J.; Dai, W.; Wu, G.; Guan, N.; Li, L. Selectivity Modulation of Encapsulated Palladium Nanoparticles by Zeolite Microenvironment for Biomass Catalytic Upgrading. ACS Catal. 2018, 8 (9), 8578–8589. 187. Jin, Z.; Wang, L.; Hu, Q.; Zhang, L.; Xu, S.; Dong, X.; Gao, X.; Ma, R.; Meng, X.; Xiao, F.-S. Hydrophobic Zeolite Containing Titania Particles as Wettability-Selective Catalyst for Formaldehyde Removal. ACS Catal. 2018, 8 (6), 5250–5254. 188. Wang, C.; Liu, Z.; Wang, L.; Dong, X.; Zhang, J.; Wang, G.; Han, S.; Meng, X.; Zheng, A.; Xiao, F.-S. Importance of Zeolite Wettability for Selective Hydrogenation of Furfural over Pd@Zeolite Catalysts. ACS Catal. 2018, 8 (1), 474–481. 189. Grajciar, L.; Heard, C. J.; Bondarenko, A. A.; Polynski, M. V.; Meeprasert, J.; Pidko, E. A.; Nachtigall, P. Towards Operando Computational Modeling in Heterogeneous Catalysis. Chem. Soc. Rev. 2018, 47, 8307–8348. 190. Reuter, K.; Scheffler, M. First-Principles Atomistic Thermodynamics for Oxidation Catalysis: Surface Phase Diagrams and Catalytically Interesting Regions. Phys. Rev. Lett. 2003, 90 (4), 046103. 191. McEwen, J. S.; Anggara, T.; Schneider, W. F.; Kispersky, V. F.; Miller, J. T.; Delgass, W. N.; Ribeiro, F. H. Integrated Operando X-Ray Absorption and DFT Characterization of Cu–SSZ-13 Exchange Sites during the Selective Catalytic Reduction of NOx with NH3. Catal. Today 2012, 184 (1), 129–144. 192. Li, G.; Pidko, E. A.; van Santen, R. A.; Li, C.; Hensen, E. J. M. Stability of Extraframework Iron-Containing Complexes in ZSM-5 Zeolite. J. Phys. Chem. C 2013, 117 (1), 413–426. 193. Zheng, J.; Lee, I.; Khramenkova, E.; Wang, M.; Peng, B.; Gutiérrez, O. Y.; Fulton, J. L.; Camaioni, D. M.; Khare, R.; Jentys, A.; Haller, G. L.; Pidko, E. A.; SanchezSanchez, M.; Lercher, J. A. Importance of Methane Chemical Potential for Its Conversion to Methanol on Cu-Exchanged Mordenite. Chem. A Eur. J. 2020, 26 (34), 7563–7567. 194. Liu, C.; Li, G.; Hensen, E. J. M.; Pidko, E. A. Nature and Catalytic Role of Extraframework Aluminum in Faujasite Zeolite: A Theoretical Perspective. ACS Catal. 2015, 5 (11), 7024–7033. 195. Li, G.; Vollmer, I.; Liu, C.; Gascon, J.; Pidko, E. A. Structure and Reactivity of the Mo/ZSM-5 Dehydroaromatization Catalyst: An Operando Computational Study. ACS Catal. 2019, 9 (9), 8731–8737. 196. Heiles, S.; Johnston, R. L. Global Optimization of Clusters Using Electronic Structure Methods. Int. J. Quantum Chem. 2013, 113 (18), 2091–2109. 197. Jørgensen, M. S.; Groves, M. N.; Hammer, B. Combining Evolutionary Algorithms with Clustering toward Rational Global Structure Optimization at the Atomic Scale. J. Chem. Theory Comput. 2017, 13 (3), 1486–1493. 198. Zhao, Z.-J.; Liu, S.; Zha, S.; Cheng, D.; Studt, F.; Henkelman, G.; Gong, J. Theory-Guided Design of Catalytic Materials Using Scaling Relationships and Reactivity Descriptors. Nat. Rev. Mater. 2019, 4 (12), 792–804. 199. Brogaard, R. Y.; Wang, C.-M.; Studt, F. Methanol–Alkene Reactions in Zeotype Acid Catalysts: Insights from a Descriptor-Based Approach and Microkinetic Modeling. ACS Catal. 2014, 4 (12), 4504–4509.

Metal containing nanoclusters in zeolites

147

200. Wang, C.-M.; Brogaard, R. Y.; Weckhuysen, B. M.; Nørskov, J. K.; Studt, F. Reactivity Descriptor in Solid Acid Catalysis: Predicting Turnover Frequencies for Propene Methylation in Zeotypes. J. Phys. Chem. Lett. 2014, 5 (9), 1516–1521. 201. Wang, C.-M.; Brogaard, R. Y.; Xie, Z.-K.; Studt, F. Transition-State Scaling Relations in Zeolite Catalysis: Influence of Framework Topology and Acid-Site Reactivity. Cat. Sci. Technol. 2015, 5 (5), 2814–2820. 202. Siahrostami, S.; Falsig, H.; Beato, P.; Moses, P. G.; Nørskov, J. K.; Studt, F. Exploring Scaling Relations for Chemisorption Energies on Transition-Metal-Exchanged Zeolites ZSM-22 and ZSM-5. ChemCatChem 2016, 8 (4), 767–772. 203. Bukowski, B. C.; Greeley, J. Scaling Relationships for Molecular Adsorption and Dissociation in Lewis Acid Zeolites. J. Phys. Chem. C 2016, 120 (12), 6714–6722. 204. Latimer, A. A.; Kulkarni, A. R.; Aljama, H.; Montoya, J. H.; Yoo, J. S.; Tsai, C.; Abild-Pedersen, F.; Studt, F.; Nørskov, J. K. Understanding Trends in C–H Bond Activation in Heterogeneous Catalysis. Nat. Mater. 2016, 16, 225. 205. Xie, P.; Pu, T.; Aranovich, G.; Guo, J.; Donohue, M.; Kulkarni, A.; Wang, C. Bridging Adsorption Analytics and Catalytic Kinetics for Metal-Exchanged Zeolites. Nat. Catal. 2021, 4 (2), 144–156. 206. Göltl, F.; Müller, P.; Uchupalanun, P.; Sautet, P.; Hermans, I. Developing a Descriptor-Based Approach for CO and NO Adsorption Strength to Transition Metal Sites in Zeolites. Chem. Mater. 2017, 29 (15), 6434–6444. 207. Szécsényi, Á.; Khramenkova, E.; Chernyshov, I. Y.; Li, G.; Gascon, J.; Pidko, E. A. Breaking Linear Scaling Relationships with Secondary Interactions in Confined Space: A Case Study of Methane Oxidation by Fe/ZSM-5 Zeolite. ACS Catal. 2019, 9 (10), 9276–9284. 208. Liu, C.; Tranca, I.; van Santen, R. A.; Hensen, E. J. M.; Pidko, E. A. Scaling Relations for Acidity and Reactivity of Zeolites. J. Phys. Chem. C 2017, 121 (42), 23520–23530. 209. Gani, T. Z. H.; Kulik, H. J. Understanding and Breaking Scaling Relations in Single-Site Catalysis: Methane to Methanol Conversion by FeIV]O. ACS Catal. 2018, 8 (2), 975–986. 210. Pérez-Ramírez, J.; López, N. Strategies to Break Linear Scaling Relationships. Nat. Catal. 2019, 2 (11), 971–976. 211. Liu, C.; van Santen, R. A.; Poursaeidesfahani, A.; Vlugt, T. J. H.; Pidko, E. A.; Hensen, E. J. M. Hydride Transfer Versus Deprotonation Kinetics in the Isobutane–Propene Alkylation Reaction: A Computational Study. ACS Catal. 2017, 7 (12), 8613–8627. 212. Sengar, A.; van Santen, R. A.; Kuipers, J. A. M. Deactivation Kinetics of the Catalytic Alkylation Reaction. ACS Catal. 2020, 10 (13), 6988–7006. 213. Martini, A.; Borfecchia, E.; Lomachenko, K. A.; Pankin, I. A.; Negri, C.; Berlier, G.; Beato, P.; Falsig, H.; Bordiga, S.; Lamberti, C. Composition-Driven Cu-Speciation and Reducibility in Cu-CHA Zeolite Catalysts: A Multivariate XAS/FTIR Approach to Complexity. Chem. Sci. 2017, 8 (10), 6836–6851. 214. Mandal, K.; Gu, Y.; Westendorff, K. S.; Li, S.; Pihl, J. A.; Grabow, L. C.; Epling, W. S.; Paolucci, C. Condition-Dependent Pd Speciation and NO Adsorption in Pd/Zeolites. ACS Catal. 2020, 10 (21), 12801–12818. 215. Gounder, R.; Iglesia, E. The Roles of Entropy and Enthalpy in Stabilizing Ion-Pairs at Transition States in Zeolite Acid Catalysis. Acc. Chem. Res. 2012, 45 (2), 229–238. 216. Van der Mynsbrugge, J.; Janda, A.; Mallikarjun Sharada, S.; Lin, L.-C.; Van Speybroeck, V.; Head-Gordon, M.; Bell, A. T. Theoretical Analysis of the Influence of Pore Geometry on Monomolecular Cracking and Dehydrogenation of n-Butane in Brønsted Acidic Zeolites. ACS Catal. 2017, 7 (4), 2685–2697. 217. Jones, A. J.; Iglesia, E. Kinetic, Spectroscopic, and Theoretical Assessment of Associative and Dissociative Methanol Dehydration Routes in Zeolites. Angew. Chem. Int. Ed. 2014, 53 (45), 12177–12181. 218. Janda, A.; Vlaisavljevich, B.; Lin, L.-C.; Smit, B.; Bell, A. T. Effects of Zeolite Structural Confinement on Adsorption Thermodynamics and Reaction Kinetics for Monomolecular Cracking and Dehydrogenation of n-Butane. J. Am. Chem. Soc. 2016, 138 (14), 4739–4756. 219. Shetty, M.; Wang, H.; Chen, F.; Jaegers, N.; Liu, Y.; Camaioni, D. M.; Gutiérrez, O. Y.; Lercher, J. A. Directing the Rate-Enhancement for Hydronium Ion Catalyzed Dehydration Via Organization of Alkanols in Nanoscopic Confinements. Angew. Chem. Int. Ed. 2021, 60 (5), 2304–2311. 220. Kalz, K. F.; Kraehnert, R.; Dvoyashkin, M.; Dittmeyer, R.; Gläser, R.; Krewer, U.; Reuter, K.; Grunwaldt, J.-D. Future Challenges in Heterogeneous Catalysis: Understanding Catalysts under Dynamic Reaction Conditions. ChemCatChem 2017, 9 (1), 17–29. 221. Liu, A.; Liu, L.; Cao, Y.; Wang, J.; Si, R.; Gao, F.; Dong, L. Controlling Dynamic Structural Transformation of Atomically Dispersed CuOx Species and Influence on Their Catalytic Performances. ACS Catal. 2019, 9 (11), 9840–9851. 222. Shamzhy, M.; Opanasenko, M.; Concepcion, P.; Martinez, A. New Trends in Tailoring Active Sites in Zeolite-Based Catalysts. Chem. Soc. Rev. 2019, 48 (4), 1095–1149. 223. Tan, S. F.; Chee, S. W.; Baraissov, Z.; Jin, H.; Tan, T. L.; Mirsaidov, U. Real-Time Imaging of Nanoscale Redox Reactions over Bimetallic Nanoparticles. Adv. Funct. Mater. 2019, 29 (37), 1903242. 224. Kang, J.; Carnis, J.; Kim, D.; Chung, M.; Kim, J.; Yun, K.; An, G.; Cha, W.; Harder, R.; Song, S.; Sikorski, M.; Robert, A.; Thanh, N. H.; Lee, H.; Choi, Y. N.; Huang, X.; Chu, Y. S.; Clark, J. N.; Song, M. K.; Yoon, K. B.; Robinson, I. K.; Kim, H. Time-Resolved In Situ Visualization of the Structural Response of Zeolites During Catalysis. Nat. Commun. 2020, 11 (1), 5901. 225. Collinge, G.; Yuk, S. F.; Nguyen, M.-T.; Lee, M.-S.; Glezakou, V.-A.; Rousseau, R. Effect of Collective Dynamics and Anharmonicity on Entropy in Heterogenous Catalysis: Building the Case for Advanced Molecular Simulations. ACS Catal. 2020, 10 (16), 9236–9260. 226. Moliner, M.; Roman-Leshkov, Y.; Corma, A. Machine Learning Applied to Zeolite Synthesis: The Missing Link for Realizing High-Throughput Discovery. Acc. Chem. Res. 2019, 52 (10), 2971–2980. 227. Ulissi, Z. W.; Medford, A. J.; Bligaard, T.; Nørskov, J. K. To Address Surface Reaction Network Complexity Using Scaling Relations Machine Learning and DFT Calculations. Nat. Commun. 2017, 8 (1), 14621. 228. Back, S.; Yoon, J.; Tian, N.; Zhong, W.; Tran, K.; Ulissi, Z. W. Convolutional Neural Network of Atomic Surface Structures to Predict Binding Energies for High-Throughput Screening of Catalysts. J. Phys. Chem. Lett. 2019, 10 (15), 4401–4408. 229. Goodman, E. D.; Schwalbe, J. A.; Cargnello, M. Mechanistic Understanding and the Rational Design of Sinter-Resistant Heterogeneous Catalysts. ACS Catal. 2017, 7 (10), 7156–7173. 230. Gao, M.; Li, H.; Liu, W.; Xu, Z.; Peng, S.; Yang, M.; Ye, M.; Liu, Z. Imaging Spatiotemporal Evolution of Molecules and Active Sites in Zeolite Catalyst During Methanol-toOlefins Reaction. Nat. Commun. 2020, 11 (1), 3641.

6.07 Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites Dieter Plessersa,*, Max L. Bolsa,*, Hannah M. Rhodab, Alexander J. Heyerb, Edward I. Solomonb,c, Bert F. Selsa, and Robert A. Schoonheydta, a Department of Microbial and Molecular Systems, Centre for Sustainable Catalysis and Engineering, KU Leuven, Leuven, Belgium; b Department of Chemistry, Stanford University, Stanford, CA, United States; and c Photon Science, SLAC National Accelerator Laboratory, Menlo Park, CA, United States © 2023 Elsevier Ltd. All rights reserved.

6.07.1 6.07.2 6.07.3 6.07.3.1 6.07.3.1.1 6.07.3.1.2 6.07.3.1.3 6.07.3.2 6.07.3.2.1 6.07.3.2.2 6.07.3.2.3 6.07.4 6.07.5 References

Introduction Transition metal ions in zeolites Achieving site selective spectroscopy Bare mononuclear transition metal ions UV-Vis-NIR and EPR spectroscopy on Cu2þ Site selective spectroscopy on Fe2þ and Fe3þ UV-Vis-NIR spectroscopy on Co2þ Complexes of metal ions in zeolites with extraframework oxygen atoms [CuOCu]2þ and [CuOOCu]2þ [Fe]O]2þ Zn-, Ga-, Co- and Ni-zeolites Site selective spectroscopy and oxo/oxyl-catalysis Conclusions and outlook

148 149 150 151 151 153 155 156 156 157 159 161 162 163

Abstract Transition metal ions (TMIs) are commonly introduced in zeolites in the form of, or as part of, cations that form a coordination complex with the zeolite framework. Their introduction enables the use of zeolites as redox catalysts. Of particular interest are TMI containing reactive oxygen complexes that may be used for mild selective partial oxidations. Various spectroscopic techniques are suited to study TMIs, but for TMIs in zeolites the interpretation of the spectroscopic data is particularly complicated because of their heterogeneous occurrence. The TMIs may simultaneously occur in different oxidation states (Cu2þ, Cuþ, Fe2þ, Fe3þ, .). They may agglomerate into di- and trinuclear clusters, or nanoparticles of the (xn  2y) type [Mxþ . Moreover various extraframework ligands (NO, O, OH, .) can be part of the first coordination sphere n Oy] in various binding modes besides the zeolite framework. Disentangling spectroscopic data obtained on heterogeneous TMI zeolites requires spectroscopic features that can unambiguously be linked to specific TMI species in the zeolite, or ‘site selective’ spectroscopic data. In addition, individual spectroscopic techniques rarely allow access to sufficient chemical properties to achieve a full model of the TMI species and a combination of techniques is required. In this article the background on zeolite-TMI interactions is given in combination with the so far successfully determined TMI sites in zeolites to demonstrate how site selective spectroscopy is achieved. Reactive surface oxygen sites that are part of, or derived from, these sites are given extra attention. The treatment of spectroscopically identified reactive oxygen is expanded with examples on non-transition metal ions (such as Ga and Zn) in zeolites. These examples demonstrate the use of site selective spectroscopy and spectroscopic handles to study reactions site selectively in the last section.

6.07.1

Introduction

Transition metal ions (TMIs) are commonly introduced in zeolites in the form of, or as part of, cations that form a coordination complex with the zeolite framework. Their introduction enables the use of zeolites as redox catalysts.1–3 Of particular interest are TMI containing reactive oxygen complexes that may be used for mild selective partial oxidations.4 The oxidation state and both the short range and long range coordination environment of these zeolite hosted TMIs are important for their reactive and adsorptive properties, and therefore desirable to know. Various spectroscopic techniques are suited to study TMIs, but for TMIs in zeolites the interpretation of the spectroscopic data is particularly complicated because of their heterogeneous occurrence.5 The TMIs may simultaneously occur in different oxidation states (Cu2þ, Cuþ, Fe2þ, Fe3þ, .). They may agglomerate into diand trinuclear clusters, or nanoparticles of the type [MnxþOy](xn  2y). Moreover various extraframework ligands (NO, O, OH, .)

*

Equal contribution.

148

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00008-X

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

149

can be part of the first coordination sphere in various binding modes besides the zeolite framework. Disentangling spectroscopic data obtained on heterogeneous TMI zeolites requires spectroscopic features that can unambiguously be linked to specific TMI species in the zeolite, or “site selective” spectroscopic data. In addition, individual spectroscopic techniques rarely allow access to sufficient chemical properties to achieve a full model of the TMI species and a combination of techniques is required. In this article the background on zeolite-TMI interactions is given in combination with the so far successfully determined TMI sites in zeolites to demonstrate how site selective spectroscopy is achieved. Reactive surface oxygen sites that are part of, or derived from, these sites are given extra attention. The treatment of spectroscopically identified reactive oxygen is expanded with examples on non-transition metal ions (such as Ga and Zn) in zeolites. These examples demonstrate the use of site selective spectroscopy and spectroscopic handles to study reactions site selectively in the last section.

6.07.2

Transition metal ions in zeolites

Zeolites are 3-dimensional silicates (or tectosilicates) in which part of Si4þ is substituted by Al3þ. These Si4þ and Al3þ, also denoted “T-atoms,” are each coordinated by four oxygen atoms that connect them to a next T-atom.6,7 In AleOeSi sequences the AleO bond is typically 0.185 nm, while the SieO bond in SieOeSi is typically 0.161 nm. The basicity of the oxygen atoms is affected by Al3þ substitution. Between the oxygen atoms in SieOeSi and AleOeSi, the latter is more basic due to the higher electronegativity of Si4þ. Tracing back to the same T-atom through the shortest sequence of T-O-T linkages that connects a group of T-atoms defines rings that are typically in the Ångström scale. The number of T-atoms (or equivalently the number of oxygen atoms) determines the free diameters of the rings. Only molecules smaller than the free diameter can pass through the ring, while larger molecules are excluded. This property controls access to the inner regions of the three dimensional crystal structure (described later) and gives zeolites their function as molecular sieves.8,9 The most common rings are shown in Fig. 1 with their free dimensions. Their symbol is depicted as nMR: n is the number of T-atoms (or equivalently oxygen atoms) forming the ring and MR means “membered ring.” Rings sizes can vary from 3MR to 24MR, and can also vary in shape from round to elliptical.10,11 Each particular three-dimensional configuration of interconnected T-atoms gives rise to a crystallographic unit cell that defines the zeolite’s topology, denoted by a three letter code. All known zeolite topologies can be found on the website of the International Zeolite Association (IZA).10 From the interconnected T-atoms nanometer-scale cages and channels arise in the long range, three dimensional structure, and the access to pores and cages is determined by the smallest ring that has to be passed through to enter from outside the crystal. The full network of intracrystalline voids is the pore system, and the relatively large specific void volume combined with its nanometer dimensions makes the zeolite microporous.8,9,12 The substitution of tetravalent Si4þ by trivalent Al3þ creates a negative lattice charge which can be charge compensated by transition metal ions (TMI).4 The TMI is then said to be ion exchanged onto the zeolite. In dehydrated zeolites the exchangeable cations commonly take well-defined positions, or exchange sites, in the pores and channels. The TMI then forms a coordination complex with the zeolite through covalent bonds with zeolite lattice oxygen atoms. These were first identified and compiled for a large number of zeolite topologies by Mortier13 and later extended to pentasil zeolites by Wichterlova et al.14 Examples of zeolite structures relevant to this article are shown in Fig. 2. To indicate the cation exchange positions in pentasil frameworks MFI, MOR, FER and *BEA, Wichterlova et al. used lowercase Greek letters.14 For the LTA, CHA and FAU topologies we will refer to the six-membered rings (6MRs), eight-membered rings (8MRs), and double six-membered rings (d6rs) as indicated in the figure. On the bottom of Fig. 2 the detailed structures of the exchange sites are shown. The a, b and g sites of Wichterlova14 are all 6MRs, except for the 8MR b-site in MOR. Besides by their structure defined by the interconnection of the T-atoms, exchange sites are also defined by the number of Al3þ substitutions they contain. The substitutions are restricted by Loewenstein’s rule which prohibits Al-O-Al sequences.9,12 There are two consequences: (1) there cannot be more than three Al ions in a 6MR, with the limiting case a 6MR with alternating Si- and Altetrahedra; (2) in the case of two Al-tetrahedra, two types of rings can be distinguished: a 6MRpara with two Al tetrahedra opposite each other and separated on each side by two Si tetrahedra, and a 6MRmeta with an Al-O-Si-O-Al sequence. The full identity of the

Fig. 1 Different ring sizes in zeolites. Adapted with permission from Snyder, B.E.R.; Bols, M.L.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. Chem. Rev. 2018, 118, 2718–2768.

150

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

Fig. 2 Framework structures and exchange sites of zeolite topologies used in this article. Color codes link the exchange sites indicated to their naming in the legend. Lower right: detail of the exchange site structures for each topology. Adapted with permission from Snyder, B.E.R.; Bols, M.L.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. Chem. Rev. 2018, 118, 2718–2768.

exchange site determines the strength with which it holds particular transition metal ions and their complexes, and determines the coordination geometry adopted by the coordination complex.

6.07.3

Achieving site selective spectroscopy

Three levels of complexity are now apparent in the analysis of TMI containing cation sites in zeolites and their spectroscopy: (1) the number of TMIs of the same element per site or “nuclearity” and their chemical interconnection; (2) the symmetry of the first-sphere coordination, including the zeolite lattice and extraframework ligands; (3) the structure, T-atom configuration, and crystallographic location of the exchange site. A fourth level worth mentioning is the relative positioning to different TMI sites, but it will not be treated in this article. The goal is to achieve site selective spectroscopy. That is, we want spectroscopic features that we can use as unambiguous handles on particular TMI containing cations coordinated in a particular way to the zeolite framework with a particular configuration of Al3þ substitutions. Such handles may be achieved by carefully combining spectroscopic techniques and sample preparation techniques, along with ab initio quantum chemical modeling. The ideal combination will be sample and conditions dependent. In general, site selective spectroscopy involves triangulation that allows the definitive assignment of observed TMI spectroscopic features. Once this assignment is achieved, the site selective spectroscopic handle (s) from (an) individual spectroscopic technique(s) may be used independently, although with caution when conditions or preparation techniques are changed. Site selective spectroscopic handles have been obtained on only a handful of TMI containing cations on zeolites. In this section we will first discuss spectroscopic data on bare mononuclear Cu2þ, Co2þ and Fe2þ cations coordinated (only) to lattice oxygen

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

151

atoms. Subsequently, the coordination of these TMIs (or their higher oxidation states) and Zn, Ga, and Ni with extralattice oxygen atoms is discussed. For Cu2þ typically UV-Vis-NIR, EPR and resonance Raman (rRaman) are applied. For Co2þ UV-Vis-NIR is discussed, and for Fe2þ we combine UV-Vis-NIR with magnetic circular dichroism (MCD), nuclear resonance vibrational spectroscopy (NRVS), and Mössbauer spectroscopy. With these spectroscopic techniques the first coordination spheres are probed. Within the boundaries of mononuclear, bare TMIs coordinated to lattice oxygen only, the spectroscopic properties of the TMIs are determined by the configuration of Al3þ- and Si4þ-tetrahedra in the coordination site and by the zeolite topology (geometry of the exchange sites). When extralattice O atoms are in the first coordination sphere, they will be dominant and strongly determine the spectroscopic properties of the TMI with lattice oxygen atoms and topology in second place.

6.07.3.1 6.07.3.1.1

Bare mononuclear transition metal ions UV-Vis-NIR and EPR spectroscopy on Cu2þ

The first site selective spectroscopic handles on TMI-zeolites were achieved by Packet and Schoonheydt in the 1980s on Cu2þ exchanged zeolites.15,16 The use of a cell that allowed in situ measurement on the same sample of both diffuse reflectance (DR) UV-Vis-NIR and electron paramagnetic resonance (EPR) spectra enabled triangulation using multiple spectroscopic techniques. In addition, low Cu2þ loading (typically less than one Cu2þ per unit cell) were used to yield highly resolved EPR spectra. Fig. 3A shows the typical UV-Vis-NIR spectrum of Cu2þ in zeolite Y (CBV-100, Si/Al ¼ 2.5). Cu2þ in zeolite A (LTA) gives a similar spectrum to the Cu-Y one shown.15 An intense lattice oxygen-to-metal charge transfer band in the 35,000–45,000 cm 1 region is present combined with a weak d-d band system between 10,000 and 15,000 cm 1. The latter consists of three bands: the most intense band lies in the range 10,500–11,000 cm 1. The two other bands are only partially resolved with maxima around 12,500 and 14,500 cm 1.15,16 The d-d spectrum of Cu2þ in MOR is shown in Fig. 3B. It has a maximum around 13,500 cm 1 with ill-defined shoulders both at lower and higher energy. An additional band around 16,750 cm 1 is present when the copper loading exceeds an undetermined threshold (at least Cu/Al > 0.06).17 An unexplained band around 20,800 cm 1 is also present. This spectrum is also typical for Cu2þ in MFI, but the band system is less resolved.18 Finally, Cu2þ in CHA is characterized by a d-d spectrum with four bands: 11,000, 13,500, 16,000 and 20,000 cm 1 (Fig. 3C).3,19,20 Based on the Cu2þ d-d spectra, zeolite topologies can be divided in three groups: (1) LTA and FAU with a d-d band maximum in the range 10,500–11,000 cm 1; (2) MOR and MFI with the d-d band maximum around 13,500 cm 1; and (3) CHA with a spectrum of four clearly defined bands. The d-d band maximum of Cu2þ in MOR and MFI lies about 2500 cm 1 higher in energy than in LTA and FAU, indicating a stronger ligand field in the former group. In all these zeolites the spectra have been interpreted as aggregates of spectral components derived from differently coordinated Cu2þ with four lattice oxygen atoms from 6MRs in its ligand field. The stronger ligand field in MOR and MFI is reflected in shorter CueO bond distances, calculated with B3LYP-DFT (Table 1).21,22 For Cu2þ in LTA and FAU, the room temperature DR-UV-Vis-NIR spectra cannot be used to resolve the configuration of Al3þ tetrahedra in the Cu2þ hosting 6MRs. The d-d transitions as calculated by B3LYP-DFT in combination with CASPT2 fall in the experimental range, but the differences between theoretical d-d spectra of Cu2þ in 6MRs with one, two or three Al are too small.21 Also for Cu2þ in MFI a distinction of Cu2þ in different 6MRs on the basis of the d-d bands has not yet been published. For Cu2þ in MOR on the other hand, the overall d-d spectrum could be resolved into 2 components, one with a maximum around 13,500 cm 1 and one with a maximum around 16,750 cm 1.17 The 16,750 cm 1 band is assigned to Cu2þ in a 6MR or a 5MR with one Al3þ on the basis of DFT-CASPT2 calculations. The 13,500 cm 1 band is assigned to Cu2þ in 6MRs with an Al-O-Si-O-Al sequence.17 The assignment of the four d-d bands in the spectrum of Cu2þ-CHA is still debated in the literature, with two competing interpretations. Godiksen et al.23 interpreted the four bands as a superposition of d-d bands of two Cu2þ atoms. The first Cu2þ with d-d bands at 11,000, 13,500 and 16,000 cm 1 was assigned to Cu2þ in 6MRmeta and the second Cu2þ with d-d bands at 13,500, 16,000 and 20,000 cm 1 was assigned to Cu2þ in 6MRpara. Packet et al. studied Cu-CHA with 0.15 Cu2þ per unit cell.24 At these low Cu2þ loadings they found a spectrum with three bands: 11,000, 13,500 and 16,000 cm 1, which is consistent to the bands grouped

Fig. 3

DR-UV-Vis-NIR spectra of (A) Cu-Y, (B) Cu-MOR, and (C) Cu-CHA.3,16,17

152

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites Table 1

Average Cu-O distances and ligand fields strengths obtained with B3LYP-DFT.17,21,22

LTA & FAU MOR & MFI

Number of Al3þ in 6MR

(nm)

10 Dq (cm1) a

1 3 1 2 (Al-Si-Al)

0.204–0.208

9476 7750 11,020–11,623 10,388–10,593

0.198–0.204

10 Dq is taken as the dxy / dx2y2 transition, assuming square planar symmetry (D4h point group).

a

together for Cu2þ in the 6MRmeta by Godiksen et al.23 In the competing interpretation, Li et al.25 proposed that a mixture of multinuclear copper-oxo and copper-hydroxo species contribute to the quadruplet. They observed that the four d-d bands disappear upon reduction with CO, leaving behind a weak two-bands spectrum with maxima around 11,350 and 13,000 cm 1. These two d-d bands were assigned to [CuOH]þ. The first interpretation by Godiksen et al.23 claims to resolve Cu2þ in 6MRpara and 6MRmeta exchange sites. However, CASPT2 calculations on 6MR in LTA and FAU suggested that Cu2þ d-d bands in these similar environments should not be easy to distinguish.21 Whether and why a distinction can be made between 6MRpara and 6MRmeta based on Cu2þ d-d bands on CHA therefore needs further confirmation. The second interpretation by Li et al.25 agrees with earlier computational results of Delabie et al. in Cu-MOR,22 where two of the d-d transitions for a three coordinate [CuOH]þ species (coordinated to two framework oxygen atoms and one extraframework OH ligand) fall in the range 11,500–14,000 cm 1. Additionally, three coordinate Cu2þ that are coordinated to three framework oxygen atoms of a 6MR are shown to have an energy below 9000 cm 1, significantly beneath the bands attributed to [CuOH]þ.21 illustrating the dominance of the extraframework oxygenligand in determining spectroscopic properties. In any case, to reach consensus on the interpretation of the d-d spectra of Cu2þ in CHA, a systematic study of Cu-CHA is needed with different Cu2þ levels from very low to full exchange. This should be coupled to experimental variations in Al distribution by modifying the zeolite synthesis.26 Further assignment of the Cu2þ d-d bands requires support from EPR, a sensitive technique that gives highly resolved spectra at very low copper loadings. On Cu2þ exchanged zeolites two axially symmetric EPR spectra of Cu2þ are invariably obtained on every topology studied so far.15–18,21–23,27 The characteristic parameters of these signals are summarized in Table 2. The g|| and A|| parameters of the EPR signals suggest the three groups already proposed earlier based on the DR-UV-Vis-NIR spectra: LTA/FAU, MOR/MFI and CHA, although the parameters of Cu2þ in CHA are very close to those of Cu2þ in LTA/FAU. For LTA/FAU we have one d-d spectrum with 3 d-d transitions, and two EPR spectra. It suggests that EPR might be able to distinguish among 6MRs with one, two or three Al3þ substitutions in LTA/FAU. DFT-CASPT2 calculations established that 6MRs with one Al3þ substitution are the most distorted.21,22 Cu2þ fourfold coordinated in these rings gives the lowest g|| values. 6MRs with three Al3þ substitutions are the least distorted and give rise to the highest g|| values. For 6MRs with two Al3þ substitutions, 6MRmeta can fall in either category while 6MRpara falls in the category with lower g|| (Table 2). For Cu2þ in MOR the spectral resolution is complete. The 16,750 cm 1 has been linked to the g|| ¼ 2.27 signal and the d-d maximum at 13,500 cm 1 to g|| ¼ 2.32. The former is due to Cu2þ in 6MRs or 5MRs with one Al and the latter is due to Cu2þ in 6MRs with two Al3þ substitutions.17 As for Cu2þ in LTA/ FAU the most distorted 6MR gives rise to the lowest g|| value. As mentioned before, the interpretation of the EPR and UV–Vis-NIR spectra of Cu2þ in CHA awaits confirmation with a systematic study of different Cu2þ loadings in chabazite zeolites with different Si/Al ratios and Al distributions. In conclusion, there is no doubt that certain groups of zeolite frameworks can be distinguished on the basis of UV-Vis-NIR and EPR spectra of Cu2þ. With EPR one can go one step further: 6MRs with one Al are the most distorted 6MRs and give the lowest g||

Table 2

EPR parameters of Cu2þ in different locations in zeolites.15–18,21–23,27

Topology

Exchange site

g||

A|| (MHz)

LTA & FAU

6MRmeta, 6MRpara, 6MR(3Al) 6MR (1Al), 6MRmeta a(2Al) a(1Al), g(1Al) a(2Al) a,b,g(1Al) 6MRmeta 6MRpara

2.36–2.41 2.30–2.34 2.30–2.32 2.26–2.28 2.31–2.32 2.26–2.28 2.358 2.325

360–435 500–565 507 573 520 549 464 487

MOR MFI CHA

gꓕ is not given as it falls in the same range for all zeolites: 2.05–2.08. Aꓕ is not resolved in most of the spectra.

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

153

and highest A|| values. 6MRs with 3 Al are the least distorted and lead to the highest g|| and lowest A|| values. Only in the case of MOR a unique set of EPR and d-d bands have been obtained and interpreted as summarized in Table 3.

6.07.3.1.2

Site selective spectroscopy on Fe2þ and Fe3þ

Fe commonly occurs as Fe3þ or Fe2þ cations in zeolites. These cations have respectively a d5 and d6 electron configuration. In the dehydrated zeolite environment, the oxygen ligands tend to induce the high spin state. In its 2S þ 1 ¼ 6 state, Fe3þ has every d orbital singly occupied and the excitation of one electron within the d orbital manifold results in a change of spin state to 2S þ 1 ¼ 4. As a consequence, d-d transitions of high spin Fe3þ are spin forbidden and therefore weak. Fully identified Fe3þ in zeolites is rare. The only example is high spin Fe3þ isomorphously substituted into the framework with tetrahedral coordination (analogous to Al3þ substitution), which is easily synthesized in a single site form (i.e., 100% of iron) and has been characterized in the 1990s. The spectroscopic features of framework substituted Fe3þ are similar across different zeolite topologies28 and from EXAFS, Fe3þ–O bond lengths are 1.87 Å.29 Four weak d-d bands are observed in the UV-Vis-NIR spectrum at 21,300, 22,900, 24,500 and 26,700 cm 1 that were assigned to the transitions from the 6A1 ground state to the 4T2(4G), 4A1(4G) & 4E(4G), 4 T2(4D) and 4E(4D) excited states respectively (note that the lowest energy 4T1 was not observed experimentally). The crystal splitting was calculated to be 10Dq ¼ 8500 cm 1.29 Later studies added further spectroscopic data. A g-value of 2.0 was found with EPR,28 and characteristic rRaman and Mössbauer spectra were measured.30,31 Just like Al3þ substitution for Si4þ, Fe3þ substitution creates a positive charge deficit that must be compensated. When compensated by a proton, a bridging OH is formed between framework Fe3þ and Si4þ with an OH stretch at 3630 cm 1 in the IR spectrum.32 This would have to fit with the experimental reported Mössbauer parameters, particularly the small QS which is thought to derive from the tetrahedral symmetry of the Fe ligand field.33 Fe2þ does have spin allowed d-d transitions, but they are in most cases difficult to resolve due to the heterogeneity of Fe2þ in the sample and they are crowded together in the near infrared. D-d transitions and binding energies for different bare mononuclear Fe2þ sites in different exchange sites with two Al3þ substitutions on MFI, *BEA and FER were calculated with CASPT2 by Hallaert et al. and compared to experimental UV-Vis-NIR spectra (Fig. 4A).34 Only one type of Fe2þ, with high spin (S ¼ 2) and square planar coordination to a 6MRpara exchange site could be experimentally resolved from other Fe2þ sites by their uniquely high energy dz2  dx2y2 transitions. This square planar Fe2þ is named a-Fe2þ and it has interesting reactivity properties (vide infra). It has been extensively studied with various spectroscopic techniques by Snyder et al., thanks to the preparation of nearly single site (> 90% of iron) Fe2þ *BEA zeolites with 0.30 wt% Fe by auto-reduction of Fe3þ zeolites in He at 900  C.35 The UV-Vis-NIR spectrum of the Fe2þ-*BEA is shown in Fig. 4A. It consists of three bands linked to the a-Fe2þ site: one with maximum below 5000 cm 1, a strong d-d band at 15,900 cm 1, and a LMCT band with maximum around 40,000 cm 1.35 The sharp band at 7000 cm 1 is the second overtone of zeolitic OH groups and the weak broad band around 9000 cm 1 is due to residual octahedral Fe2þ. The position and the intensity of the high energy d-d band in the visible depends on the zeolite topology as shown in Fig. 4A. The intensity of the 15,900 cm 1 follows the order BEA* > MFI > CHA > FER > MOR. In the latter case it is almost absent. In Fe2þ-CHA a-Fe2þ has its d-d bands shifted to 13,000 cm 1 and at 5400 cm 1, and their relative intensity is inverted.36 The shift to lower energy indicates a weaker ligand field in CHA, while the inverted relative intensity reflects a different distortion in coordination symmetry. These observations are well explained by the models placing a-Fe2þ in a 6MRpara exchange site (Fig. 4E).36 In *BEA this is the b exchange site where Fe2þ adopts a square planar coordination with a distortion to C2 symmetry. In CHA the occupied site is the wider, more symmetric 6MR of the d6r exchange site where the Fe2þ adopts a square planar coordination with a weaker distortion to CS symmetry. Hence, zeolite exchange sites can be discerned from the position and relative intensity of the dz2 / dx2y2 electron transition of quasi-square planar Fe2þ in 6MRs, which occurs in the 13,000–16,000 cm 1 range. The electronic properties of a-Fe2þ were further investigated with variable-temperature-variable-field (VTVH) magnetic circular dichroism (MCD).35 This technique involves the induction of circular dichroism by the presence of a magnetic field parallel to the propagation of the circularly polarized light. MCD has properties that make the technique suitable for the study of these materials: (1) there is no diamagnetic contribution to the signal so that small amounts of a TMI (< 1 wt%) can be measured in a diamagnetic matrix such as a zeolite. (2) It is possible to study one component of a mixture when it can be identified by a unique d-d band. (3) The spectra contain information on electronic transitions and their polarization that is complementary to UV-Vis-NIR spectroscopy.4 The d-d absorption bands of Fe2þ seen in the UV-Vis-NIR spectra are reproduced in the MCD spectrum (Fig. 4B). The intensity of the 15,900 cm 1 band in MCD is dependent on the temperature and on the magnetic field from which the zero field splitting parameters (E and D) and the effective g value can be derived (Fig. 4B–D). These data confirm that a-Fe2þ (the precursor of a-O) is a high spin, square planar Fe2þ, corroborating the assignment of the 15,900 cm 1 band to the dz2 / dx2y2 transition.35

Table 3

d-d bands and EPR spectra of Cu2þ in MOR.17

d-d (cm 1)

g||

A|| (MHz)

Assigned exchange site

16,750 13,500

2.27 2.32

573 507

6MR(1Al), 5MR(1Al) 6MR(2Al)

154 Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

Fig. 4 (A) UV-Vis-NIR spectra of Fe2þ in zeolites *BEA, MFI, CHA, FER and FAU. (B–D) VTVH-MCD spectroscopy on Fe2þ in *BEA. (E) Structural models of a-Fe2þ in *BEA and CHA. (F) Mössbauer spectrum of a-Fe2þ in *BEA. (G) Experimental and modeled NRVS on a-Fe2þ in *BEA. (A–D) Adapted with permission from Snyder, B.E.R.; Vanelderen, P.; Bols, M.L.; Hallaert, S.D.; Böttger, L.H.; Ungur, L.; Pierloot, K.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. Nature 2016, 536, 317–321, (E) adapted with permission from Bols, M.L.; Hallaert, S.D.; Snyder, B.E.R; Devos, J.; Plessers, D.; Rhoda H.M.; Dusselier, M.; Schoonheydt, R.A.; Pierloot, K.; Solomon, E.I.; Sels, B.F. J. Am. Chem. Soc. 2018, 140, 12021–12032, (F) adapted with permission from Snyder, B.E.R.; Vanelderen, P.; Bols, M.L.; Hallaert, S.D.; Böttger, L.H.; Ungur, L.; Pierloot, K.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. Nature 2016, 536, 317–321, (G) adapted with permission from Snyder, B.E.R.; Böttger, L.H.; Bols, M.L.; Yan, J.J.; Rhoda, H.M.; Jacobs, A.B.; Hu, M.Y.; Zhao, J.; Ercan Alp, E.; Hedman, B.; Hodgson, K.O.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. Proc. Natl. Acad. Sci. U. S. A. 2018, 115, 4565–4570.

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

155

The specific 6MRpara Al3þ configuration was confirmed by nuclear resonance vibrational spectroscopy (NRVS) on 57Fe isotopically labeled a-Fe2þ sites, which selectively observes vibrational modes that involve the motion of 57Fe isotope. The observed and modeled spectra are shown in Fig. 4G.37 Also the room temperature Mössbauer spectrum was collected, which element- (and isotope-) specifically probes the resonant absorption of g-photons by the nuclear excitation of 57Fe from I ¼ 1/2 to I ¼ 3/2. The result for Fe2þ is a doublet, shown for a-Fe2þ in Fig. 4F.35,36 The two bands in the spectrum correspond to the excited states I ¼  1/2 and I ¼  3/2. The center between the two peaks is the isomer shift (d) and the separation between the two components of the doublet is the quadrupole splitting |QS |. Both are sensitive to the oxidation state and the spin state of the Fe, as well as the ligand field. The very low |QS | of 0.55 mm s 1 is quite unusual as high spin Fe2þ usually shows a large |QS | and reflects the square planar nature of the ligand field at the Fe2þ site. This is illustrated for a-Fe2þ in *BEA and CHA in Table 4. The |QS | for a-Fe2þ is especially low for mononuclear high spin Fe2þ, and this connects to the electronic structure with a low lying, doubly occupied dz2 combined with the square planar ligand field generating a low electric field gradient around the Fe2þ nucleus.

6.07.3.1.3

UV-Vis-NIR spectroscopy on Co2þ

Co2þ has a d7 shell with a 4F ground state and a 4P excited state. In a ligand field 4F and 4P will split into a number of components, depending on the symmetry of the ligand field. The coordination of Co2þ to lattice oxygen atoms in zeolites has been systematically studied by UV–Vis-NIR, and this has been proposed as an identifier for exchange sites by Dedecek et al.38–40 Typical d-d spectra of Co2þ in LTA, FAU and MFI are shown in Fig. 5. The spectrum of dehydrated Co2þ in LTA (Co-A) consists of three band systems (I, II and III): I has its maximum around 6500 cm 1 and encompasses at least three components; II is a band system of three to four bands in the visible range (15,000–20,000 cm 1), and III is a weak doublet in the 22,000–25,000 cm 1 region (Fig. 5A).41,42 Band system II is the most intense one and it gives the Co-zeolites their characteristic blue color. The spectra of Co2þ in FAU (zeolites X and Y) are very similar. The relative intensities of the components in band systems I and II are somewhat different, while band system III is not visible (Fig. 5B).41,42 It may be submerged in the background. Klier used a ligand field model of Co2þ coordinated to three O atoms in the 6MRs of LTA and assigned band systems I and II to transitions within the 4F manifold and band system III to transitions within the 4P manifold.43 Pierloot et al. used multiconfigurational perturbation theory (CASPT2) and extended Klier’s model for zeolite A to the 6MRs in zeolites X and Y with one, two or three Al3þ substitutions.42 They agreed that band system I is due to transitions within the 4F manifold. However, they assigned band systems II and III to transitions from the ground state into the energy levels of the 4P manifold. In addition, they found that the 6MR with one Al is the most distorted one and it gives the strongest ligand field. This conclusion was also reached for Cu2þ.21,22 The UV-Vis-NIR spectra of Co2þ in MOR, MFI and FER have the same d-d band system I in the NIR, but band system II encompasses the region 15,000–22,000 cm 1 and band system III is absent (Fig. 5C for Co-FER).38–40 Possibly band systems II and III are fused together in the 15,000–22,000 cm 1 region. The group of Wichterlova has studied the coordination of Co2þ in MOR, FER, MFI and *BEA as a function of Co loading and of the type of co-cation with special attention to the 15,000–22,000 cm 1 region (band system II).44 They concluded that the 15,000–22,000 cm 1 region contains bands due to Co2þ coordinated in sites a, b and g (Fig. 2). Each coordination site has its typical bands and absorption coefficients, as shown in Table 5. However, there is not yet a quantum-chemical foundation of the spectra assigned to Co2þ in a, b and g sites. If we adopt the multiconfigurational model developed by Pierloot et al.42 for the 6MRs of LTA and FAU, the 5000–6200 cm 1 manifold is due to transitions within the 4F manifold and the bands of Table 5 are due to transitions from the ground state into the 4P manifold of Co2þ. 4P can be split into two or three levels depending on the symmetry of the ligand field. Each site (a, b, g) can then give rise to two or three transitions, so totaling six to nine transitions for a zeolite in which all three exchange sites are occupied. For the reasons above, one band for Co2þ in site a is unlikely. Probably other bands are not resolved under the other bands, or one of the bands assigned to Co2þ in site b is due to Co2þ in site a. The next question is why Co2þ in sites a, b and g have such different transitions within the 4P manifold. Is it due to differences in the ligand field strength and do we have to invoke spin-orbit coupling here? The conclusions of the analysis of the d-d spectra of Co2þ coordinated to lattice oxygen atoms in zeolites are very similar to those of Cu2þ. Again, two groups of zeolites can be distinguished: LTA and FAU vs. MOR, MFI, FER and BEA. Similarly to Cu, the 6MRs are the preferred Co2þ coordination sites in all zeolites and 6MRs with one Al give the most distorted coordination environment and the strongest ligand field. However, when coordination of Co2þ in 6MRs with one, two and three Al3þ substitutions occur simultaneously, it is hardly possible, if not impossible, to distinguish them on the basis of the UV-Vis-NIR spectra alone. Therefore Wichterlova’s original proposal44 of Co2þ coordination in a, b and g sites would benefit from a quantum-chemical foundation or better future spectroscopic definition. Table 4

a-Fe a-O

Isomer shifts (mm/s) and quadrupole splittings (mm/s) of a-Fe and a-O in zeolites.35,36

d |QS|

D

|QS|

Fe-*BEA

Fe-CHA

0.89 0.55 0.30 0.50

0.93 0.63 0.28 0.72

156

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

(A)

Kubelka-Munk (a.u.)

(B)

II

2.5

(C)

II

1.5

X

0.8

2.0 0.6 1.5

I

0.4

1.0

1.0 Y

I Y X

0.5

IlI 0.2

0.5 0.0 5000

10000

15000

20000

25000

30000

0.0

I

II

0.0 5000

Wavenumber (cm–1)

10000

15000

20000

25000

30000

10000

Wavenumber (cm–1)

20000

30000

40000

50000

Wavenumber (cm–1)

Fig. 5 (A) d-d spectra of Co-A, (B) Co-X and Co-Y zeolites, (C) Co-FER. (A and B) Adapted with permission from Pierloot K.; Delabie A.; Ribbing, C.; Verberckmoes, A.A.; Schoonheydt, R.A. J. Phys. Chem. B 1998, 102, 10789–10798. (C) adapted with permission from Kaucký, D.; Dedecek, J.; Wichterlová, B. Microporous Mesoporous Mater. 1999, 31, 75–87.

6.07.3.2

Complexes of metal ions in zeolites with extraframework oxygen atoms

At the end of the 20th century Iwamoto et al. discovered that Cu- zeolites were excellent catalysts for deNOx.45 At about the same time Panov’s group found that Fe-zeolites are excellent catalysts for selective oxidations, such as methane to methanol and benzene to phenol, and they introduced the terms a-Fe and a-O.46–50 Groothaert et al.51 also reported the selective oxidation of methane into methanol over Cu-MFI zeolites. In these examples, the catalytic cycle goes through Cuþ/Cu2þ or Fe2þ/Fe4þ oxidation states, and in their oxidized forms, Cu2þ and Fe4þ sites coordinate a reactive extraframework oxygen atom in their ligand field. In the case of Fe zeolites, a-Fe2þ of the previous section is formed by auto-reduction, and it can abstract an oxygen from N2O to form the reactive oxidation site, so-called a-O. For Cu-zeolites, autoreduction generates Cuþ cations that are spectroscopically elusive, but they form reactive oxidation sites upon activation in O2 or N2O. These reactive intermediates with extraframework oxygen, through their close link to catalytic applications, prompted an extensive research effort into understanding Cu- and Fe-zeolites, and into comparing them to Cu- and Fe-enzymes.4 Also non-transition metal ions ligated to zeolite exchange sites form reactive oxygen intermediates and will be discussed.52,53 The alkyl pathway of methane activation through C-H cleavage on isolated Zn sites is well known.54 Recently a surprisingly reactive Zn2þ bound oxygen site capable of rapid methane C-H activation and CO oxidation below room temperature was identified as well.55,56

6.07.3.2.1

[CuOCu]2þ and [CuOOCu]2þ

Upon exposure to O2 at room temperature, an absorption band at 29,000 cm 1 is formed on autoreduced Cu-MFI. When heated to > 150  C, this converts to a 22,700 cm 1 band. Alternatively the 22,700 cm 1 feature can be derived from N2O at  150  C (Fig. 6A).51 Laser excitation into the 22,700 cm 1 band revealed a resonance Raman spectrum which was assigned with the aid of 18O2 and 16O18O experiments to a bent mono-m-oxo dicopper(II) cluster, [CuOCu]2þ. The precursor with absorption band at 29,000 cm 1 is assigned to a m-(h2:h2) peroxo dicopper(II) species, [CuOOCu]2þ (Fig. 6B).57,58 The assignment of the [CuOCu]2þ vibrations is given in panel 6C. Three [CuOCu]2þ species have been characterized so far, one in Cu-MFI57 and two in Cu-MOR,59 and they have similar UV-Vis-NIR and rRaman spectroscopic features (panel 6D). To accommodate a [CuOCu]2þ cluster two Al3þ substitutions are required to neutralize the divalent cationic cluster charge and the separation of the Al3þ substitutions in the zeolite rings must fit the dimensions of the [CuOCu]2þ unit. This is the case for the 10MRs of MFI in which two Al3þ substitutions are separated by two Si4þ tetrahedra.57 For MOR the intersections of the 12MR with 8MR channels are suitable sites in addition to the elliptical 8MRs. However, the latter site is only accessible via the 8MR channels so that a [CuOCu]2þ in that position can only interact with sufficiently small molecules such as methane.59,60

Table 5

Band signatures of Co2þ in sites a, b and g of different zeolites.44 Energy (cm1)

Absorption coefficient (mol cm g1)

Topology

site a

site b

site g

site a

site b

site g

MOR FER MFI BEA

14,800 15,000 15,100 14,600

15,900, 17,500, 19,200, 21,100 16,000, 17,100, 18,700, 20,600 16,000, 17,150, 18,600, 21,200 15,500, 16,300, 17,570, 21,700

20,150, 22,050 20,300, 22,000 20,100, 22,000 18,900, 20,600

7.3 2.7 3.7 10.8

2.7 2.5 2.7 6.7

1.9 1.1 0.9 5.1

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

157

Fig. 6 (A) UV-Vis spectra of Cu-MFI showing the 29,000 cm 1 absorption band of the [CuOOCu]2þ precursor and the 22,700 cm 1 band related to the [CuOCu]2þ active site.57,58 (B) Resonance Raman spectrum of the [CuOCu]2þ in active site in Cu-MFI upon laser excitation into the 22,700 cm 1 band: red for 16O and blue for 18O.57 insert in D: Resonance Raman spectrum of the [CuOOCu]2þ precursor in Cu-MFI upon laser excitation into the 29,000 cm 1 band: black for 16O and blue for 18O. (C) Resonance Raman vibrations of [CuOCu]2þ in Cu-MFI and their assignment.57 (D) Comparison of the spectroscopy of [CuOCu]2þ in MFI and in MOR.57,59 (B) Adapted with permission from Smeets, P.J.; Hadt, R.G.; Woertink, J.S.; Vanelderen, P.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. J. Am. Chem. Soc. 2010, 132, 14736–14738.

A trinuclear cluster, [Cu3(m-O)3]2þ, with a characteristic absorption around 31,000 cm1 has also been proposed in Cu-MFI and Cu-MOR based on titration experiments, but up to now a confirmation by complementary spectroscopic techniques such as rRaman is missing.61–64

6.07.3.2.2

[Fe]O]2þ

Activating a-Fe2þ discussed in an earlier section with N2O results in a reactive site named a-O.46 Its UV-Vis-NIR, MCD, Mössbauer and NRVS spectroscopic data are given in Fig. 7.35,37 The UV-Vis-NIR spectrum contains a band at  6000 cm 1 and a band at 16,900 cm 1 with a high energy shoulder (Fig. 7B). In the MCD spectrum the NIR band around 6000 cm 1 has three components at 5500, 6700 and 7700 cm 1 (Fig. 7B–D). The separation by 1000–1200 cm 1 is reminiscent of a vibrational progression. The UVVis band at 16,900 cm 1 with shoulder at 20,100 cm 1 is seen in MCD with reversed intensity. Although a full interpretation of the UV-Vis-NIR and MCD spectra has not yet been obtained, we do know that these spectra are due to a high spin square pyramidal ferryl oxo complex with four coordination bonds to four lattice oxygen atoms (Fig. 7A).35 The Fe(IV)]O stretch has been observed by Nuclear Resonance Vibrational Spectroscopy (NRVS) at 885 cm 1 and with NRVS the exchange site was pinpointed to the 6MRpara site.37 The room temperature Mössbauer spectrum of a-O is a doublet with isomer shift ¼ 0.30 mm s 1 and quadrupole splitting ¼ 0.50 mm s 1.35 In Fe2þ-CHA the d-d spectrum of a-Fe2þ is converted into the spectrum of a-O, [Fe]O]2þ, with bands at 17,500, 27,000 and 37,000 cm1.36 [Fe]O]2þ in *BEA and in CHA have very similar spectroscopic signatures, in contrast to the clear differences in the spectroscopy of a-Fe2þ. This suggests that the oxo ligand dominates the spectroscopy and not the exchange site’s geometry.

158 Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

Fig. 7 (A) Structural model of a-O in *BEA. (B) UV-Vis-NIR (top) and MCD (bottom) spectra on a-O in *BEA. (C and D) VTVH-MCD data on a-O in *BEA. (E) Mössbauer spectrum of a-O in *BEA.35 (F and G) Experimental and modeled NRVS on a-O in *BEA.37 (A-E) Adapted with permission from Snyder, B.E.R.; Vanelderen, P.; Bols, M.L.; Hallaert, S.D.; Böttger, L.H.; Ungur, L.; Pierloot, K.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. Nature 2016, 536, 317–321, (F and G) Adapted with permission from Snyder, B.E.R.; Böttger, L.H.; Bols, M.L.; Yan, J.J.; Rhoda, H.M.; Jacobs, A.B.; Hu, M.Y.; Zhao, J.; Ercan Alp, E.; Hedman, B.; Hodgson, K.O.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. Proc. Natl. Acad. Sci. U. S. A. 2018, 115, 4565–4570.

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

159

Fig. 8 (A) Routes to Zn2þ- and Ga3þ-oxyls in MFI zeolites. (B) UV-VIS-NIR spectra of the ozonides and (C) of the oxyl complexes of Zn2þ and Ga3þ in MFI.65 Adapted with permission from Oda, A.; Tanaka, T.; Sawabe, K.; Satsuma, A. J. Phys. Chem. Lett. 2020, 9426–9431.

6.07.3.2.3

Zn-, Ga-, Co- and Ni-zeolites

Oda et al. prepared oxyl complexes in Zn- and Ga-MFI and identified site selective spectroscopic handles for these in UV-Vis-NIR and EPR spectroscopy.53,65 Reacting Zn-H and Ga-H precursors with O2 forms ozonide complexes that transform into [Zn2þ-O]þ and [Ga3þ-O]2þ upon room temperature evacuation. The ozonide to oxyl transformation is reversible upon reintroduction of O2 (Fig. 8A). The electronic spectra of the ozonide and oxyl complexes are shown in Fig. 8B. In the UV-Vis-NIR spectra of the ozonides as well as the oxyls, electronic bands with vibronic progression are visible, which are remarkably clear for the ozonides. The ozonides have a band with maximum around 23,000 cm 1 and a vibronic progression with a frequency of 900 cm 1. [Ga3þ-O]2þ and [Zn2þ-O]þdMFI have an absorption in the 13,000–14,000 cm 1 range with a 600 cm 1 vibronic progression frequency. The 13,000 cm 1 band is interpreted as a transition from the s orbital of GaeO or ZneO bonds to the nonbonding orbitals on the oxygen. The 600 cm 1 corresponds to the GaeO or ZneO vibrational stretch. The clear vibronic progression is indicative of unique sharply defined sites with GaeO and ZneO freely vibrating in the 10MR channels of MFI or at the channel intersections. As exchange sites, an isolated Al3þ substitution was suggested for the Zn-oxyl while a two Al3þ substituted exchange site is suggested for the Ga-oxyl, based on DFT calculations.53,65 The exact crystallographic position is uncertain, but may not matter much. The requirement of only a single Al3þ substitution, with little restrictions otherwise is a promising feature to develop these materials with high densities of active Zn2þ sites. So far, from XANES, an estimated 10–15% of Zn2þ formed the reactive oxyl.65

Fig. 9 (A) UV-Vis-NIR spectra of Co-MFI with the increase of intensity of the 11,400 cm 1 band upon exposure to O2, enlarged in the insert. (B) Similar spectra for Ni-MFI. (A) Adapted with permission from Oda, A.; Mamenari, Y.; Ohkubo, T.; Kuroda, Y. J. Phys. Chem. C 2019, 123, 17842– 17854, (B) adapted with permission from Oda, A.; Nanjo T.; Ohkubo, T.; Kuroda, Y. J. Phys. Chem. C 2020, 124, 11544–11557.

160

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

Fig. 10 (A) The formation of [CuOOCu]2þ (top) and its transformation into [CuOCu]2þ (bottom) followed in time with operando UV–Vis-NIR spectroscopy on Cu-MFI. (B) The decay of the 22,700 cm 1 band of [CuOCu]2þ on Cu-MFI during methane reaction. (C) The decay of the two [CuOCu]2þ bands of Cu-MOR during methane reaction. (D) Decay of the 6000 and 16,900 cm 1 a-O bands on Fe-*BEA with methane reaction. (E) Transformation of the Mössbauer spectrum of Fe-*BEA after N2O activation of a-Fe(II) and subsequent exposure to benzene. (A) Adapted with permission from Smeets, P.J.; Hadt, R.G.; Woertink, J.S.; Vanelderen, P.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. J. Am. Chem. Soc. 2010, 132, 14736–14738, (B) adapted with permission from Groothaert, M.H.; Smeets P.J;, Sels, B.F; Jacobs, P.A; Schoonheydt, R.A. J. Am. Chem. Soc. 2005, 127, 1394–1395, (C) adapted with permission from Vanelderen, P.; Snyder, B.E.R.; Tsai, M.L.; Hadt, R.G.; Vancauwenbergh, J.; Coussens, O.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I.; J. Am. Chem. Soc. 2015, 137, 6383–6392, (D) adapted with permission from Snyder, B.E.R.; Vanelderen, P.; Bols, M.L.; Hallaert, S.D.; Böttger, L.H.; Ungur, L.; Pierloot, K.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. Nature 2016, 536, 317– 321, (E) adapted with permission from Snyder, B.E.R; Bols, M.L.; Rhoda, H.M.; Vanelderen, P.; Böttger, L.H.; Braun, A.; Yan, J.J.; Hadt, R.G.; Babicz, J.T.; Hu, M.Y.; Zhao, J.; Ercan Alp, E.; Hedman, B.; Hodgson, K.O.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. Proc. Natl. Acad. Sci. U. S. A. 2018, 115, 12124–12129.

Fig. 11 (A) Decay of the 22,700 cm 1 band of [CuOCu]2þ on Cu-MFI during methane reaction followed in time. (B) Arrhenius plots of conversion data as shown in (A) for CH4 and CD4 reaction. (C) Transition state enthalpies and entropies on the reaction coordinate reacting [CuOCu]2þ with CH4 derived from Eyring plots on Cu-MFI and Cu-MOR.59 (A and B) Adapted with permission from Woertink, J.S.; Smeets, P.J.; Groothaert, M.H.; Vance, M.A.; Sels, B.F.; Schoonheydt, R.A.; Solomon, E.I. Proc. Natl. Acad. Sci. 2009, 106, 18908–18913.

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

161

Co2þ- and Ni2þ-MFI were reduced with CO to Niþ and Coþ carbonyl complexes, which decompose to [Ni(CO)]þ and [Co(CO)2]þ upon gentle heating. Reaction with O2 gives complexes described as side-on peroxo of Co3þ and side-on superoxo in the case of Ni2þ.66,67 The characteristic electronic spectra of the complexes are shown in Fig. 9. The bands of Co-peroxo and Ni-superoxo are both centered around 11,400 cm 1. Both bands have a clear vibronic progression with frequency of 560 cm 1 in the case of Co-peroxo and 1000 cm 1 for Ni-superoxo (the latter defining it as a superoxo). The electronic bands are assigned as d / d* transitions, referring to the bonding/antibonding interactions of the O2 p* and a d orbital both perpendicular to the MO2 plane. The Ni-superoxo is an S ¼ 1/2 complex with electronic structure of the ground state (p)2 (d)2 (3dz2)2 (3dx2y2)2 (3dzy)2 (d*)1 (p*)0. The unpaired electron resides in the out of plane p* orbital of O2 that has some metal character as evidenced by the EPR spectrum with gz value that is significantly greater than 2.00 (gx ¼ 2.064, gy ¼ 2.089, and gz ¼ 2.169).67 For the side-on Co-O2 complex Oda et al.66 arrived at the electronic configuration of the ground state as (p)2 (d)2 (3dz2)2 (3dx2y2)2 (3dzy)1 (d*)1 (p*)0. As for Ga- and Zn-oxyls, the sites of these Co-O2 and Ni-O2 complexes are isolated from additional low energy vibrational progressions by the zeolite pore. Otherwise, the vibronic progression would not be resolved. The number of these sites is presently unknown.

6.07.4

Site selective spectroscopy and oxo/oxyl-catalysis

In the previous sections spectroscopic features on TMI-, Ga3þ, and Zn2þ exchanged zeolites were linked to specific metal ion species hosted in the exchange sites of the framework. Unique spectroscopic handles that allow their distinction from other sites were shown. These metal sites participate in catalytically relevant redox chemistry involving the formation of reactive oxo or oxyl species, and for both the oxo/oxyl-species and their precursors site selective spectroscopic handles are available.4,56,65 The spectroscopic handles enabled unique insight into the geometric and electronic structure of these reactive intermediates, backed by experiment. In a next step we use these handles to connect the knowledge of the structures to reactivity, obtaining “structure-function relationships.” Such relationships are essential tools for the understanding and improvement of catalysis. Structure-function relationships are easily hypothesized from ab initio models,68–71 but they are challenging to verify experimentally. Experimental access is enabled by spectroscopic handles on active sites and intermediates. In this section we discuss the formation of the copper- and iron bound reactive oxo groups, the zinc bound oxyl group, and their reactions with methane and benzene. Fig. 10A shows the formation and decay of the spectroscopic handle in UV-Vis-NIR of the [CuOOCu]2þ peroxo precursor site at 29,000 cm 1.58 The band grows in when reduced Cu-MFI is exposed to O2, and it decays with heating. Heating transforms the peroxo into the reactive [CuOCu]2þ as indicated by its UV-Vis-NIR handle at 22,700 cm 1 and its associated rRaman spectra.57 The simultaneous growth of the 22,700 cm 1 band and decay of the 29,000 cm 1 band is a convincing argument that the former is

Fig. 12 Catalytic reaction cycle of benzene to phenol using N2O on Fe-*BEA. In red the deactivating mechanism is shown. Spectroscopic techniques used to characterize the structures are written in purple. Adapted with permission from Snyder, B.E.R; Bols, M.L.; Rhoda, H.M.; Vanelderen, P.; Böttger, L.H.; Braun, A.; Yan, J.J.; Hadt, R.G.; Babicz, J.T.; Hu, M.Y.; Zhao, J.; Ercan Alp, E.; Hedman, B.; Hodgson, K.O.; Schoonheydt, R.A.; Sels, B.F.; Solomon, E.I. Proc. Natl. Acad. Sci. U. S. A. 2018, 115, 12124–12129.

162

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

the precursor to the latter. Similarly, the decay of the 22,700 cm 1 band upon exposure to CH4, shown in Fig. 10B, correlates the [CuOCu]2þ to the activation of methane.57 After the removal of the [CuOCu]2þ with methane, methanol can be desorbed from the zeolite. The [CuOCu]2þ UV-Vis-NIR bands are sensitive to small changes in the active site, enabling their site selective monitoring. This is illustrated in Fig. 10C for Cu-MOR, where two [CuOCu]2þ sites were identified with bands at 21,900 and 23,100 cm 1 and assigned by rRaman spectra. Both bands can be observed to decay during methane reaction.59 Also the UV-Vis-NIR handle on a-O, the reactive oxo formed from N2O on iron zeolites, correlates with methane activation and the formation of desorbable methanol (Fig. 10D).35 In contrast with the copper zeolites, where [CuOCu]2þ only represents a minority of copper,17 the a-O represents a majority of iron on the zeolite. From quantitative analysis it follows that a-O must be responsible for the formation of methanol, while a similar quantitative link between lost absorbance at  22,500 cm 1 with activated methane of produced methanol has not yet been published. The a-O reaction with benzene to yield extractable phenol has also been studied. In this reaction the product spectrum clouds the loss of the 16,900 cm 1 UV-Vis-NIR feature of a-O, but the role of a-O is clearly visible using its Mössbauer spectroscopy handle (Fig. 10E).72 On Zn-MFI, the appearance of the [Zn2þ-O]þ features on UV-Vis-NIR and EPR spectroscopy could be correlated to the length of the UV irradiation step required to form it from its ozonide precursor (vide supra).55 Reaction of the oxyl containing Zn-MFI with methane at room temperature produced Zn bound hydroxyls and methoxides, both visible in IR spectroscopy. The methoxides could quantitatively be recovered as methanol through water extraction, and the signals and time of irradiation linearly correlate with the amount of recovered methanol.55 The Zn2þ oxyl EPR feature was also found to disappear in the presence of CO at 150 K, although the observation of the expected CO2 product was not reported. The disappearance of the site-selective UV-VisNIR feature of Zn2þ-oxyl correlated linearly with the appearance of a new EPR signal that was attributed to Znþ, which would be the product from oxyl transfer to CO.56 Zn2þ superoxide was also observed in EPR, but the spectral features did not show the correlations with reactivity observed for the oxyl.56 Similarities in reactivity between the oxyls on Zn-MFI and Ga-MFI have not yet been explored. Provided the time resolution of the spectroscopic measurement is sufficient relative to the reaction rate, kinetic data can be extracted from the growth and decay of the spectroscopic handles. This was achieved for the reaction of the [CuOCu]2þ sites in Cu-MFI and Cu-MOR with methane (Fig. 11).57,59 Arrhenius or Eyring plots were derived from the decay of the  22,500 cm 1 handles on the different [CuOCu]2þ sites, and therefore the obtained kinetic data is specific for the reaction on the active sites represented by the spectroscopic feature. Also here, the two [CuOCu]2þ sites identified on Cu-MOR (see Fig. 10C) could be disentangled to reveal their site specific activities.59 Such data cannot be obtained from bulk conversion data obtained through product or reagent monitoring and present the unique advantage of spectroscopic site selective measurement to obtain genuine kinetic data. This is especially the case for copper zeolites where multiple types of active site may simultaneously contribute to reactivity with similar activities.64 These measurements enable the establishment of structure-function relations. As mentioned in Section 6.07.3, MOR1 is located on the intersection of the 12MR and the 8MR side pocket. MOR2 is situated on the other side of the 8MR side pocket, across the elliptical 8MR channel.59 Fig. 11C shows that MOR2 has a lower O Hz (favorable for high reactivity) and a lower (more negative) O Sz (unfavorable for high reactivity) than both the [CuOCu]2þ in Cu-MFI and MOR1. The lower O Sz could readily be explained

by the more confined reaction environment of MOR2 deep inside the side pocket of MOR. The difference in O Hz, however, was too large to be explained by the small differences in the geometric/electronic structure between the [Cu-O-Cu]2þ species. Further theoretical investigations showed that the formation of the reactant complex in the confined reaction environment of MOR2 leads to an enthalpy decrease due to close van der Waals contact between CH4 and the walls of the site.60 The van der Waals stabilization of substrate adsorbed in the zeolite side pocket driving the reaction is reminiscent of active-site pocket chemistry of metalloenzymes. Besides active site selective reaction kinetics, the availability of site selective spectroscopic handles enables reliable links between intermediates of a reaction cycle. This was demonstrated for the conversion of benzene to phenol with a-O on Fe-*BEA and Fe-ZSM5. Earlier studies had identified a bis-Fe(III) phenolate as the reaction product using rRaman spectroscopy, without the availability of spectroscopic handles on the a-Fe(II) and a-O sites.73 Instead, a dinuclear Fe(III) peroxo active site was presumed to be the active oxidation site. Snyder et al. reproduced the rRaman data formerly assigned to the bis-Fe(III) phenolate on Fe-*BEA samples with predominantly mononuclear iron and reassigned it as an Fe(III)-phenolate (Fig. 12).72 Moreover, the Fe(III)-phenolate was not the main product. Instead, a cycle occurred regenerating Fe(II), susceptible to reactivation with N2O to form a-O. The Fe(II) was identified as a-C6H6, i.e., a-Fe(II) coordinated to benzene. The regeneration of a-Fe(II) can restart the cycle, opening the door to catalysis. The Fe(III)-phenolate on the other hand is a poison to the catalyst, and its accumulation may lead to cokes. Full characterization of the Fe(III)-phenolate and the benzene bound a-Fe(II) required a combination of techniques (added in purple to Fig. 12) for full identification. This is as described in Section 6.07.3; full identification of the active sites required triangulation of several techniques. This extends to their reaction intermediates and products.

6.07.5

Conclusions and outlook

The preceding examples illustrate how site-selective spectroscopy allowed to identify TMI species and their reactive oxygen complexes in zeolites. In most cases this requires a clever combination of spectroscopic techniques and is facilitated by using optimized materials with little spectator sites. Mononuclear and dinuclear complexes with one, two and three oxygen atoms have been

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

163

spectroscopically identified. These complexes are very reactive. Operando spectroscopy has been used to obtain kinetic parameters of methane activation in Cu- and Fe-zeolites. Recently, mobility of Cu ions has been demonstrated in Cu-CHA during selective catalytic reduction (SCR) of NOx with NH3,74,75 and during catalytic partial methane oxidation.76 We expect that the presented approach will stay valuable in the future, with more focus on operando spectroscopy in an attempt to monitor this dynamic speciation of the TMI ions in the catalyst under transient operating conditions.77

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54.

Smeets, P. J.; Woertink, J. S.; Sels, B. F.; Solomon, E. I.; Schoonheydt, R. A. Inorg. Chem. 2010, 49, 3573–3583. Kosinov, N.; Liu, C.; Hensen, E. J. M.; Pidko, E. A. Chem. Mater. 2018, 30, 3177–3198. Borfecchia, E.; Beato, P.; Svelle, S.; Olsbye, U.; Lamberti, C.; Bordiga, S. Chem. Soc. Rev. 2018, 47, 8097–8133. Snyder, B. E. R.; Bols, M. L.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. Chem. Rev. 2018, 118, 2718–2768. Bols, M. L.; Rhoda, H. M.; Snyder, B. E. R.; Solomon, E. I.; Pierloot, K.; Schoonheydt, R. A.; Sels, B. F. Dalton Trans. 2020, 49, 14749–14757. Huheey, J. E.; Keiter, E. A.; Keiter, R. L. In Inorganic ChemistrydPrinciples of Structure and Reactivity; Huheey, J. E., Keiter, E. A., Keiter, R. L., Eds., 4th edn.; HarperCollins Publishers, 1993; pp 738–823. Guth, J. L.; Kessler, H. In Catalysis and ZeolitesdFundamentals and Applications; Weitkamp, J., Puppe, L., Eds., Springer, 1999; pp 1–52. Flanigen, E. M. In Studies in Surface Science and Catalysis; Van Bekkum, H., Flanigen, E. M., Jacobs, P. A., Jansen, J. C., Eds.; vol. 137; Elsevier Science B.V., 2001; pp 11–35. Davis, M. E. Ind. Eng. Chem. Res. 1991, 30, 1675–1683. International Zeolite Association, (n.d.) http://www.iza-online.org/. Baerlocher, C.; McCusker, L. B.; Olson, D. H. Atlas of Zeolite Framework Types, Elsevier, 2007. Dusselier, M.; Davis, M. E. Chem. Rev. 2018, 118, 5265–5329. Mortier, W.; Structure Commission of the International Zeolite Association. Compilation of Extra Framework Sites in Zeolites, Butterworth Scientific Limited: Leuven, 1982. Wichterlová, B.; Dedecek, J.; Sobalik, Z. In Catalysis by Unique Metal Ion Structures in Solid Matrices: From Science to Application; Centi, G., Wichterlová, B., Bell, A. T., Eds.; NATO Science Series A, Kluwer Academic Publishers, 2001; pp 31–53. Packet, D.; Schoonheydt, R. A. In Structure and Reactivity of Modified Zeolites; Jacobs, P. A., Jaeger, N. I., JírŮ, P., Kazansky, V. B., Schulz-Ekloff, G., Eds., Elsevier Science Publishing Company, 1984; pp 41–48. Packet, D.; Schoonheydt, R. A. In New Developments in Zeolite Science and Technology; Murakami, Y., Iijima, A., Ward, J. W., Eds., Elsevier, 1986; pp 385–392. Vanelderen, P.; Vancauwenbergh, J.; Tsai, M.-L.; Hadt, R. G.; Solomon, E. I.; Schoonheydt, R. A.; Sels, B. F. ChemPhysChem 2014, 15, 91–99. Groothaert, M. H.; Van Bokhoven, J. A.; Battiston, A. A.; Weckhuysen, B. M.; Schoonheydt, R. A. J. Am. Chem. Soc. 2003, 125, 7629–7640. Ipek, B.; Wulfers, M. J.; Kim, H.; Göltl, F.; Hermans, I.; Smith, J. P.; Booksh, K. S.; Brown, C. M.; Lobo, R. F. ACS Catal. 2017, 7, 4291–4303. Giordanino, F.; Vennestrøm, P. N. R.; Lundegaard, L. F.; Stappen, F. N.; Mossin, S.; Beato, P.; Bordiga, S.; Lamberti, C. Dalton Trans. 2013, 42, 12741. Pierloot, K.; Delabie, A.; Groothaert, M. H.; Schoonheydt, R. A. Phys. Chem. Chem. Phys. 2001, 3, 2174–2183. Delabie, A.; Pierloot, K.; Groothaert, M. H.; Weckhuysen, B. M.; Schoonheydt, R. A. Phys. Chem. Chem. Phys. 2002, 4, 134–145. Godiksen, A.; Stappen, F. N.; Vennestrøm, P. N. R.; Giordanino, F.; Rasmussen, S. B.; Lundegaard, L. F.; Mossin, S. J. Phys. Chem. C 2014, 118, 23126–23138. Packet, D.; Dehertogh, W.; Schoonheydt, R. A. In Zeolites Synthesis, Structure Technology & Application; Drzaj, B., Hocevar, S., Pejovnik, S., Eds., Elsevier, 1985; pp 351–358. Li, H.; Paolucci, C.; Khurana, I.; Wilcox, L. N.; Göltl, F.; Albarracin-Caballero, J. D.; Shih, A. J.; Ribeiro, F. H.; Gounder, R.; Schneider, W. F. Chem. Sci. 2019, 10, 2373–2384. Di Iorio, J. R.; Gounder, R. Chem. Mater. 2016, 28, 2236–2247. De Tavernier, S.; Schoonheydt, R. A. Zeolites 1991, 11, 155–163. Goldfarb, D.; Bernardo, M.; Strohmaier, K. G.; Vaughan, D. E. W.; Thomann, H. J. Am. Chem. Soc. 1994, 116, 6344–6353. Patarin, J.; Tuilier, J. D.; Kessler, H. Zeolites 1992, 12, 70–75. Lázár, K.; Pozdnyakova, O.; Wootsch, A.; Fejes, P. Hyperfine Interact. 2006, 167, 779–784. Sun, K.; Fan, F.; Xia, H.; Feng, Z.; Li, W. X.; Li, C. J. Phys. Chem. C 2008, 112, 16036–16041. Bordiga, S.; Buzzoni, R.; Geobaldo, F.; Lamberti, C.; Giamello, E.; Zecchina, A.; Leofanti, G.; Petrini, G.; Tozzola, G.; Vlaic, G. J. Catal. 1996, 158, 486–501. Taboada, J. B.; Overweg, A. R.; Kooyman, P. J.; Arends, I. W. C. E.; Mul, G. J. Catal. 2005, 231, 56–66. Hallaert, S. D.; Bols, M. L.; Vanelderen, P.; Schoonheydt, R. A.; Sels, B. F.; Pierloot, K. Inorg. Chem. 2017, 56, 10681–10690. Snyder, B. E. R.; Vanelderen, P.; Bols, M. L.; Hallaert, S. D.; Böttger, L. H.; Ungur, L.; Pierloot, K.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. Nature 2016, 536, 317–321. Bols, M. L.; Hallaert, S. D.; Snyder, B. E. R.; Devos, J.; Plessers, D.; Rhoda, H. M.; Dusselier, M.; Schoonheydt, R. A.; Pierloot, K.; Solomon, E. I.; Sels, B. F. J. Am. Chem. Soc. 2018, 140, 12021–12032. Snyder, B. E. R.; Böttger, L. H.; Bols, M. L.; Yan, J. J.; Rhoda, H. M.; Jacobs, A. B.; Hu, M. Y.; Zhao, J.; Ercan Alp, E.; Hedman, B.; Hodgson, K. O.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. Proc. Natl. Acad. Sci. U. S. A. 2018, 115, 4565–4570. Dedecek, J.; Kaucký, D.; Wichterlová, B. Microporous Mesoporous Mater. 2000, 35, 483–494. Dedecek, J.; Wichterlová, B. J. Phys. Chem. B 1999, 103, 1462–1476. Kaucký, D.; Dedecek, J.; Wichterlová, B. Microporous Mesoporous Mater. 1999, 31, 75–87. Verberckmoes, A. A.; Weckhuysen, B. M.; Schoonheydt, R. A. Stud. Surf. Sci. Catal. 1997, 105, 623–630. Pierloot, K.; Delabie, A.; Ribbing, C.; Verberckmoes, A. A.; Schoonheydt, R. A. J. Phys. Chem. B 1998, 102, 10789–10798. Klier, K.; Hutta, P. J.; Kellerman, R. Molecular SievesdII, Webster: New York, 1977; pp 108–119. Wichterlová, B.; Dedecek, J.; Sobalík, Z. In Catalysis by Unique Metal Ion Structures in Solid Matrices From Science to Application; Centi, G., Wichterlova, B., Bell, A. T., Eds.; NATO Science Series, Kluwer Acadamic, 2001; pp 31–54. Iwamoto, M.; Yahiro, H.; Tanda, K.; Mizuno, N.; Mine, Y.; Kagawa, S. J. Phys. Chem. 1991, 95, 3727–3730. Panov, G. I.; Sheveleva, G. A.; Kharitonov, A. S.; Romannikov, V. N.; Vostrikova, L. A. Appl. Catal. A. Gen. 1992, 82, 31–36. Sobolev, V. I.; Panov, G. I.; Kharitonov, A. S.; Romannikov, V. N.; Volodin, A. M.; Ione, K. G. J. Catal. 1993, 139, 435–443. Kharitonov, A. S.; Sheveleva, G. A.; Panov, G. I.; Sobolev, V. I.; Paukshtis, Y. A.; Romannikov, V. N. Appl. Catal. A. Gen. 1993, 98, 33–43. Panov, G. I.; Uriarte, A. K.; Rodkin, M. A.; Sobolev, V. I. Catal. Today 1998, 41, 365–385. Panov, G. I.; Sobolev, V. I.; Kharitonov, A. S. J. Mol. Catal. 1990, 61, 85–97. Groothaert, M. H.; Smeets, P. J.; Sels, B. F.; Jacobs, P. A.; Schoonheydt, R. A. J. Am. Chem. Soc. 2005, 127, 1394–1395. Oda, A.; Torigoe, H.; Itadani, A.; Ohkubo, T.; Yumura, T.; Kobayashi, H.; Kuroda, Y. Angew. Chem. Int. Ed. 2012, 51, 7719–7723. Oda, A.; Tanaka, T.; Sawabe, K.; Satsuma, A. J. Phys. Chem. Lett. 2020, 9426–9431. Gabrienko, A. A.; Arzumanov, S. S.; Luzgin, M. V.; Stepanov, A. G.; Parmon, V. N. J. Phys. Chem. C 2015, 119, 24910–24918.

164 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77.

Single site spectroscopy of transition metal ions and reactive oxygen complexes in zeolites

Oda, A.; Ohkubo, T.; Yumura, T.; Kobayashi, H.; Kuroda, Y. Inorg. Chem. 2019, 58, 327–338. Oda, A.; Kumagai, J.; Ohkubo, T.; Kuroda, Y. Inorg. Chem. Front. 2021, 8 (2), 319–328. Woertink, J. S.; Smeets, P. J.; Groothaert, M. H.; Vance, M. A.; Sels, B. F.; Schoonheydt, R. A.; Solomon, E. I. Proc. Natl. Acad. Sci. 2009, 106, 18908–18913. Smeets, P. J.; Hadt, R. G.; Woertink, J. S.; Vanelderen, P.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. J. Am. Chem. Soc. 2010, 132, 14736–14738. Vanelderen, P.; Snyder, B. E. R.; Tsai, M. L.; Hadt, R. G.; Vancauwenbergh, J.; Coussens, O.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. J. Am. Chem. Soc. 2015, 137, 6383–6392. Snyder, B. E. R.; Vanelderen, P.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. J. Am. Chem. Soc. 2018, 140, 9236–9243. Grundner, S.; Markovits, M. A. C.; Li, G.; Tromp, M.; Pidko, E. A.; Hensen, E. J. M.; Jentys, A.; Sanchez-Sanchez, M.; Lercher, J. A. Nat. Commun. 2015, 6, 1–9. Markovits, M. A. C.; Jentys, A.; Tromp, M.; Sanchez-Sanchez, M.; Lercher, J. A. Top. Catal. 2016, 59, 1554–1563. Zheng, J.; Lee, I.; Khramenkova, E.; Wang, M.; Peng, B.; Gutiérrez, O. Y.; Fulton, J. L.; Camaioni, D. M.; Khare, R.; Jentys, A.; Haller, G. L.; Pidko, E. A.; Sanchez-Sanchez, M.; Lercher, J. A. Chem. Eur. J. 2020, 26, 7563–7567. Newton, M. A.; Knorpp, A. J.; Sushkevich, V. L.; Palagin, D.; Van Bokhoven, J. A. Chem. Soc. Rev. 2020, 49, 1449–1486. Oda, A.; Ohkubo, T.; Yumura, T.; Kobayashi, H.; Kuroda, Y. Angew. Chem. Int. Ed. 2017, 56, 9715–9718. Oda, A.; Mamenari, Y.; Ohkubo, T.; Kuroda, Y. J. Phys. Chem. C 2019, 123, 17842–17854. Oda, A.; Nanjo, T.; Ohkubo, T.; Kuroda, Y. J. Phys. Chem. C 2020, 124, 11544–11557. Wang, G.; Chen, W.; Huang, L.; Liu, Z.; Sun, X.; Zheng, A. Catal. Today 2019, 338, 108–116. Mahyuddin, M. H.; Shiota, Y.; Yoshizawa, K. Cat. Sci. Technol. 2019, 9, 1744–1768. Latimer, A. A.; Kulkarni, A. R.; Aljama, H.; Montoya, J. H.; Yoo, J. S.; Tsai, C.; Abild-Pedersen, F.; Studt, F.; Nørskov, J. K. Nat. Mater. 2017, 16, 225–229. Kulkarni, A. R.; Zhao, Z. J.; Siahrostami, S.; Nørskov, J. K.; Studt, F. Cat. Sci. Technol. 2018, 8, 114–123. Snyder, B. E. R.; Bols, M. L.; Rhoda, H. M.; Vanelderen, P.; Böttger, L. H.; Braun, A.; Yan, J. J.; Hadt, R. G.; Babicz, J. T.; Hu, M. Y.; Zhao, J.; Ercan Alp, E.; Hedman, B.; Hodgson, K. O.; Schoonheydt, R. A.; Sels, B. F.; Solomon, E. I. Proc. Natl. Acad. Sci. U. S. A. 2018, 115, 12124–12129. Xia, H.; Sun, K.; Sun, K.; Feng, Z.; Li, W. X.; Li, C. J. Phys. Chem. C 2008, 112, 9001–9005. Paolucci, C.; Parekh, A. A.; Khurana, I.; Di Iorio, J. R.; Li, H.; Albarracin Caballero, J. D.; Shih, A. J.; Anggara, T.; Delgass, W. N.; Miller, J. T.; Ribeiro, F. H.; Gounder, R.; Schneider, W. F. J. Am. Chem. Soc. 2016, 138, 6028–6048. Paolucci, C.; Khurana, I.; Parekh, A. A.; Li, S.; Shih, A. J.; Li, H.; Di Iorio, J. R.; Albarracin-Caballero, J. D.; Yezerets, A.; Miller, J. T.; Delgass, W. N.; Ribeiro, F. H.; Schneider, W. F.; Gounder, R. Science 2017, 357 (80), 898–903. Dinh, K. T.; Sullivan, M. M.; Narsimhan, K.; Serna, P.; Meyer, R. J.; Dincǎ, M.; Román-Leshkov, Y. J. Am. Chem. Soc. 2019, 141, 11641–11650. Krishna, S. H.; Jones, C. B.; Gounder, R. Annu. Rev. Chem. Biomol. Eng. 2021, 12, 16.1–16.22.

6.08

Dynamic evolution of catalytic active sites within zeolite catalysis

Massimo Bocus, Samuel E. Neale, Pieter Cnudde, and Ve´ronique Van Speybroeck, Center for Molecular Modeling, Ghent University, Tech Lane Ghent Science Park Campus A, Zwijnaarde, Belgium © 2023 Elsevier Ltd. All rights reserved.

6.08.1 6.08.2 6.08.2.1 6.08.2.1.1 6.08.2.1.2 6.08.2.2 6.08.2.3 6.08.2.3.1 6.08.2.3.2 6.08.2.3.3 6.08.2.3.4 6.08.3 6.08.3.1 6.08.3.2 6.08.3.2.1 6.08.3.2.2 6.08.3.2.3 6.08.3.2.4 6.08.4 References

Introduction Experimental and theoretical evidence for active site mobility in zeolites Proton mobility in Brønsted-acidic zeolites BAS mobility in the pristine zeolite framework Protic molecules mediated hopping and solvation of the BAS Framework-associated and extra framework aluminum (EFAL) Mobility of active sites in TM-exchanged zeolites Mobility of copper sites in Cu-CHA during low-temperature NH3-SCR-NOx Solvation and mobility of Pd in SSZ-13 Mobility of Rh in zeolite Y and consequences for ethene hydrogenation & oligomerization Mobility of Ag sites in MFI during C3H8-SCR reactivity Computational assessment of active site mobility in zeolites Overview of enhanced sampling methods over static methods Case studies Proton mobility in zeolites Ni-SSZ-24 for ethene oligomerization Mobility of active sites in H-SSZ13 during fast NH3-SCR-NOx Mobility of multinuclear Cu sites in chabazites for the selective catalytic reduction (SCR) of nitrogen oxides Conclusions and perspectives

165 167 167 167 169 170 172 173 181 181 182 183 183 186 186 187 187 192 194 194

Abstract Zeolites with Brønsted acid and cation-exchanged sites are extremely effective single-site heterogeneous catalysts and are routinely employed in various industrially important processes such as MTH and SCR. However it is now generally accepted that these sites of catalytic activity are not necessarily statically fixed to the zeolite architecture, but are more dynamic in nature and can mobilize within the pores & cages of the respective zeolites at operando conditions. Herein, an overview of the state of understanding of the mobility of (i) protons in pristine BAS-zeolites, (ii) aluminum ions in the process of framework decomposition, and (iii) metal sites in TM-exchanged zeolites is presented. In turn, the state of play of computationally probing active-site mobility will then be discussed, whereby enhanced sampling techniques is highlighted in particular as an emerging and promising method for probing active site mobility. Finally, a selection of case studies are highlighted, in which enhanced sampling techniques were employed to elucidate the mobility of catalytic sites in zeolites.

6.08.1

Introduction

Over the last decades, zeolites have become the workhorse for acid-catalyzed transformations in the chemical industry with applications in (hydro)cracking, methane reforming, methanol-to-hydrocarbons (MTH), NOx reduction, biomass upgrading, etc.1–5 Owing to their large surface area and characteristic porous network which exhibits a unique type of shape selectivity, zeolites compose a versatile class of materials for catalyzing chemical transformations which would otherwise be too highly activated to be feasible. Archetypical zeolites are crystalline microporous aluminosilicate frameworks built up by SiO4 or AlO4 tetrahedra. The incorporation of trivalent Al3þ as substituent for the tetravalent Si4þ ions in the framework results in the creation of catalytically active sites in the framework, shown in Fig. 1.2 The overall negative charge on the framework can be compensated by the incorporation of cationic species. If protons (Hþ) are introduced as charge-compensating cations, strong Brønsted acid sites (BAS) are formed as bridging hydroxyl groups. Alternatively, charge balancing can be accomplished by other cationic species such as metallic cations or (transition) metal complexes, giving rise to redox sites or Lewis acid sites (LAS). Albeit less common, lattice modifications through isomorphous substitution of framework Si sites by other trivalent (B3þ,Fe3þ,Ga3þ, .) or tetravalent (Sn4þ,Ti4þ, Zr4þ, .) heteroatoms also lead to the introduction of Brønsted and Lewis acid sites respectively in the catalyst. Finally, depending on the specific conditions during the zeolite synthesis and hydrothermal treatment, these heteroatoms and bridging hydroxyl groups

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00012-1

165

166

Dynamic evolution of catalytic active sites within zeolite catalysis

Fig. 1 Schematic representation of the steps occurring in acid zeolite catalyzed reaction. The active sites: (A) Brønsted acid sites (BAS), (B) framework-bound Lewis acid sites (LAS) and (C) mobilized LAS, can dynamically change by interacting with the reacting species, as highlighted by the different colors.

can to some extent be extracted from the framework, leading to the formation of LASs in the form of extra-framework species like extra-framework aluminum (EFAl).6–9 Tailoring the catalytic function of the formed Bronsted and Lewis acid sites is a critical factor in achieving zeolite catalysts with a high activity.10,11 While the activity of BASs stems from the (partial) proton transfer to basic guest molecules, LASs are characterized by a vacant orbital of the metal complex which can coordinate with electron-rich groups of the guest species. The catalytic behavior of the zeolite catalyst is often not solely determined by intrinsic reaction rates, but also by diffusion phenomena as well as the location, density and strength of the acid sites.12–15 The latter are key characteristics influencing the activity and selectivity of the catalyst which can be tuned by altering the type of the charge-compensating cations and their locations in the framework. Recently, dedicated studies showed progress towards controlling the targeted acid site distribution in the zeolite micropores.16–19 Computational modeling has become ubiquitous for obtaining insight into the nature of the active sites and their interaction with crucial reaction intermediates and products. The ultimate goal is to unravel the catalytic mechanism and reactivity of the zeolite catalyst, which may eventually lead to dedicated control of the product selectivity and mitigating catalyst deactivation. One of the main challenges in modeling catalytic processes on heterogeneous nanoporous catalysts is to accurately describe the interactions of adsorbates with the catalytically active sites at operating conditions. The state and the function of the catalytic material is critically dependent on the external conditions such as temperature, pressure, acidity, presence of moisture, . Specifically, zeolites were long believed to behave as rather static materials with a well-defined pore topology and a very stable lattice consisting of relatively strong SieO and AleO bonds. However, recent awareness has grown that the catalytic function may dynamically reorganize upon exposing the material to varying external process conditions.10,20–23 In this sense, the active sites do not remain unaltered at fixed framework positions but can instead evolve in the course of the reaction. Depending on several factors such as the synthesis conditions, reaction temperature and presence of guest species, the active sites can actually detach from the framework and transform into mobile sites inside the catalyst micropores. For example, the intrinsic mobility of Brønsted acid protons, a phenomenon labeled proton hopping, is a function of the local structure and composition of the zeolite as well as the temperature.24–26 At low temperature, liberated protons tend to jump solely between the four framework oxygens adjacent to the Al substitution, while at high temperature, protons may travel further to nearby framework positions.27 The presence of protic molecules like ammonia, water or methanol around the BAS is also known to induce proton mobility by formation of ion-paired hydronium or methanol clusters.28–34 Furthermore, even unsaturated apolar hydrocarbons such as alkenes or aromatics might readily deprotonate the framework, facilitating proton transfer through the formation of carbocation intermediates.35–37 Lewis acid sites are characterized by their dynamic nature as well due to the existence of both framework-bound complexes and extra-framework complexes detached from the lattice.38–42 Framework-modified zeolites may thus contain a heterogeneous distribution of active sites in the micropores, ranging from isolated cationic species over mono- or binuclear metal-oxo complexes to metal oxide nanoclusters and the nature of these sites can be significantly affected by the specific reaction conditions.11 Interestingly, the nature and stability of the catalytically active sites is sensitive to the zeolite topology and chemical composition, i.e., the distribution and/or proximity of the Al substituents in the framework.39 Accurate zeolite models capable of capturing this complex active site behavior are therefore essential to understand and predict the outcome of relevant catalytic processes. To achieve this goal, a close synergy between theoretical simulations and experimental characterization to identify the precise nature of the active sites is indispensable.43 Assessing the precise location and dynamic behavior of the active sites is a topical subject in zeolite catalysis. Fundamental insight into catalytic materials is only possible by a strong connection of theoretical and experimental data. In this context, the

Dynamic evolution of catalytic active sites within zeolite catalysis

167

recent developments of in situ and operando spectroscopy techniques to map the function of the material with high spatial and temporal resolution at operating conditions is very promising.44,45 In this article, we will demonstrate how theoretical models can aid in characterizing the active site mobility in acid zeolite catalysts and gaining a proper atomic level understanding on the interactions of guest molecules with the active sites. First, the state-of-the-art on the role of the different active sites in zeolite catalysis, i.e., Brønsted acid sites, (partially) hydrolyzed Al sites and transition metal-exchanged sites, will be introduced. A clear connection to both experimental and theoretical methods is given to identify the dynamic evolution of active sites at operating conditions. Subsequently, various computational methodologies to account for the operando and in situ catalytic conditions will be briefly reviewed. Finally, four case studies will be discussed demonstrating the value of operando modeling tools in characterizing the nature of the active sites. The first case study describes the Brønsted acid proton mobility in the context of methanol-to-hydrocarbons (MTH) conversion and ethanol dehydration. The second case study is devoted to Nickel active sites in ethylene oligomerization. The third and fourth example concern the investigation of H-SSZ-13 and Cu exchanged zeolites respectively for the selective catalytic reduction (SCR) of NOx. Each case study highlights how the acid function in zeolites can typically not be considered as a static site on the lattice but will rather change dynamically depending on the complexity of the operating environment and process conditions. Capturing this behavior correctly is a prerequisite in order to predict the catalytic properties and design catalysts with improved selectivities as well as enhanced lifetimes.

6.08.2

Experimental and theoretical evidence for active site mobility in zeolites

6.08.2.1

Proton mobility in Brønsted-acidic zeolites

If the charge compensation of the zeolite framework occurs through incorporation of a proton, bound to a bridging Si-O-Al oxygen atom, Brønsted acid sites (BAS) are formed. Despite its apparent simplicity, the BAS already possesses dynamic behavior that can significantly affect its catalytic properties. Herein, a survey of evidences collected for BAS mobility is presented for two situations of increasing complexity, namely proton mobility in the pristine zeolite framework (Fig. 2A) and proton mobility induced by the presence of adsorbed protic guest molecules (Fig. 2B and C).

6.08.2.1.1

BAS mobility in the pristine zeolite framework

In zeolites or zeotype materials, the BAS is commonly depicted as a static moiety, strongly anchored to a framework oxygen and dissociation is only considered upon active participation of the proton in the reaction under investigation, for instance by protonating an organic substrate. Proton transfer and exchange is indeed a key initial step in multiple important industrial processes, like aromatic electrophilic substitution46 and alkene cracking.36,37 The static description of the BAS is, however, contradicted by experimental observations above a certain temperature. 1H solid state Magic-Angle-Spin Nuclear Magnetic Resonance (MAS NMR) represents one of the major techniques to investigate the nature of protons in zeolites.47 Measurements conducted on various zeolites have proven that, above a temperature of  400 K, the BAS is no longer localized on a specific site but rather hops between the 4 Albound oxygen atoms (Fig. 2A).48–55 Despite a general agreement on the temperature induced free proton hopping, the barriers obtained for this process are seldom in agreement (Table 1).49–51 In H-ZSM-5, for instance, barriers ranging between 11 and 45 kJ mol 1 have been reported. For the first value, the variation in the BAS signal width with temperature was considered and it has then later been suggested that partial overlap with a broad peak at higher ppmdinitially not visible because of technical limitationsdcould be responsible for the values discrepancy (vide infra and Fig. 3) but a definitive conclusion is, to the best of our knowledge, not yet reported.62 Proton hopping can also be investigated using InfraRed (IR) spectroscopy,56,63 with the advantage that even higher temperatures can be reached with respect to NMR. The decrease in intensity of the O-H stretching mode peak with increasing temperature is normally monitored. Such decrease has been attributed to the dissociation of the OeH bond, consistent with proton hopping. By deducing the equilibrium constant of the hopping from the integrated intensity at various temperatures, Osuga et al.56 were able to use the Van’t Hoff equation to retrieve the activation energy of the process. Surprisingly, they found two different temperature regions with two different activation energies: for medium-high temperatures (398–548 K, 37–28 kJ mol 1 in H-ZSM-5) they attribute the decrease in intensity to a localized hopping on the first coordination sphere around the Al defect, while at higher

Fig. 2 Schematic depiction of the BAS mobilization in the framework in different circumstances. (A) Proton hopping in the pristine framework at high temperatures. (B) Proton hopping mediated by an adsorbed protic molecule (methanol for example). (C) Fully solvated BAS at low temperature. Color code: H white, C gray, O red, Si yellow and Al purple.

168

Dynamic evolution of catalytic active sites within zeolite catalysis Table 1

Zeolite H-ZSM-5

Apparent activation energies for the proton hopping in various zeolite frameworks, with different Si/Al ratios and different methodologies, as extracted from the available literature. Si/Al

21 38 12–53 35–90 15–500 – – H-MOR 7 39 53 H-Y 1.2–2.6 3 3 – H-SSZ-13 39 11 11

Methodology 1

H NMRa H NMRb 1 H NMRa IR Impedance spectroscopy DFT (cluster) QM-pot 1 H NMRb IR 1 H NMRa 1 H NMR 1 H NMRb 1 H NMRa QM-pot IR QM-pot QM-pot 1

act (kJ mol 1) Eapp

References

11 45 17–20 (37–28)-(2322)c 89–126 117 52–98 54 23–24c 28 21–42 61 50 68–106 23–18c 58–97 70–102

49 50 51 56 57,58 59 60 50 56 52 48 50 55 60 56 61 60

a

Based on the linewidth of isotropic resonance. Based on the intensity loss of spin sidebands with temperature. c Different values for low (398–548 K) and high (573–773 K) temperatures, respectively. b

Fig. 3 1H MAS NMR spectrum of the H-ZSM-5 zeolite (Si/Al ¼ 15). The inset shows a magnification of the 10–17 ppm region. The most common attributions for each peak are shown. Adapted with permission from Chen, K.; Abdolrahmani, M.; Horstmeier, S.; Pham, T.N.; Nguyen, V.T.; Zeets, M.; Wang, B.; Crossley, S.; White, J.L., Brønsted–Brønsted Synergies Between Framework and Noncrystalline Protons in Zeolite H-ZSM-5. ACS Catal. 2019, 9(7), 6124–6136. Copyright 2019, American Chemical Society.

temperatures (573–773 K, 23–22 kJ mol 1) they invoked a more delocalized hopping, in which the proton can almost freely move in the zeolite pores (Fig. 2A). Such an inter site hopping has also been previously proposed based on impedance spectroscopy experiments,57,58 in which higher activation energies were however found (89–126 kJ mol 1, decreasing with decreasing Si/Al ratio). Static calculations64 provided even higher barriers for the inter Al proton hopping with similar Si/Al ratio as the impedance spectroscopy, although the correlation with the Si/Al ratio was maintained (> 200 kJ mol 1). This disagreement between computational and experimental proton hopping barriers is quite general in the available literature. Many computational investigations of proton hopping around the first coordination sphere of Al have been performed with static methodologies, both on cluster59,65,66 and periodic60,61,64,67 models. It was found that the activation energy strongly depends on the two specific oxygens involved in the hopping. As an example, Sierka & Sauer61 reported zero-point corrected barriers for proton hopping in a fully periodic H-SSZ-13 model varying between 58 and 97 kJ mol 1, as a function of the considered O atoms (B3LYP/T(O)DZP). Again, the computed activation energies are significantly higher than the experimentally measured ones (Table 1). As the adsorption of protic molecules on the BAS has been proposed to significantly reduce the hopping barriers (vide infra), it has been proposed that the lower experimental values

Dynamic evolution of catalytic active sites within zeolite catalysis

169

could be justified by the presence of residual amounts of water in the catalyst. A broad band centered at  6 ppm, observed in the 1H NMR spectrum of H-ZSM-5 (Fig. 3), was indeed later attributed to residual adsorbed water sticking to the framework even after high temperature drying,62 despite previous reports providing different interpretations.68–72 In more recent investigations, however, new evidences have suggested that such band would be produced by ExtraFramework ALuminum species (EFAL), as discussed more in detail in Section 6.08.2.2,73 or by BAS involved in H-bonds with framework oxygens.74

6.08.2.1.2

Protic molecules mediated hopping and solvation of the BAS

Protic molecules adsorbed on the BAS can mediate proton hopping between (adjacent) oxygen atoms, with a reduced barrier caused by the less tensioned transition state (Fig. 2B).59,75 The adsorption of a single water molecule on the BAS is related to the formation of a strong hydrogen bond. 76,77 Despite the BAS remaining bound to the framework, the hopping barrier around the Al is reduced from 117 to 17 kJ mol 1 in H-ZSM-5, according to Ryder et al.59 Due to the high acidity of the BAS, not only typical amphoteric molecules (such as water and alcohols) can facilitate the hopping, but also aromatics78,79, alkenes and even alkanes.80,81 While a single protic molecule adsorbed on the BAS can easily exchange protons, multiple sufficiently basic adsorbates (depending on the acidity of the zeotype material under study) can completely abstract it from the framework, forming positively charged clusters (Fig. 2C). Being the most studied protic molecule in zeolites, this section focuses on the computational and experimental evidence concerning the water solvation of protons in zeolites. Such behavior is, of course, more general and applies also to other interesting protic molecules, such as methanol. In Section 6.08.3.2.1, more details will be provided on how advance sampling techniques have played a key role in elucidating how the cooperative action of methanol molecules influences some key steps in the MTH process.82,83 The adsorption of water in zeolites provides a further degree of complexity to the catalyst. As shown by theoretical calculations in H-SSZ-1384 and Al K-edge X-ray Adsorption Near Edge Structure (XANES) of H-ZSM-5,30 about 50% of the protons is still solvated at water loadings as low as 1 water molecule per BAS. This clearly points towards a heterogenous distribution of the water molecules in the zeolite, that prefer to aggregate in cluster of 2–3 molecule while leaving some BAS dehydrated instead of homogenously adsorb over all active sites. This is caused by the large energy gain caused by proton solvation when a sufficiently large cluster of water molecules is interacting with the BAS. Addition of two water molecules on the BAS causes the proton to easily shuttle between the zeolite and the water cluster. Metadynamics was used to quantify the barrier of the process, that was found to be in the order of kBT at 330 K.85 Once three or more molecules are adsorbed on the BAS, the proton gets solvated as hydronium ion and the abstraction from the framework becomes fundamentally complete. This behavior is very common for all zeolite and zeotype materials with strong BAS and has received widespread support from inelastic neutron scattering86 and diffraction,28 1H NMR,87 IR spectroscopy28,88 and many theoretical works employing MD simulations, starting from the late 1990s.29,83,84,89–92 Upon further increasing the amount of adsorbed water, the size of the protonated clusters can grow up to 7–8 water molecules before condensation is observed (in H-ZSM-5, independently on the Si/Al ratio, see Fig. 4).93,94 Such clusters retain a structure that is mostly similar to gas-phase instead of solvated hydronium ions in pure liquid water.30 After this point, the enthalpy of adsorption of water converges to the heat of condensation94 and liquid water starts to condensate in the channels of the zeolite. This was also shown by an investigation of proton mobility with IR spectroscopy, monitoring the H/D exchange rate in the catalyst. The enthalpy variation associated with proton mobility decreases with increasing water content, which was interpreted by the formation of water chains connecting different Al sites, and thus the passage from isolated positive clusters migration towards a Grottuß-like mechanism.25,26

Fig. 4 (A) Number of water molecules per BAS as a function of the water partial pressure for H-ZSM-5 with different Si/Al ratio. As clearly visible, the amount of adsorbed water does not depend on the BAS density and stops at 7 water/BAS before condensation is observed. (B) Adsorption heat of water as a function of the number of water molecules adsorbed on the BAS. The value is initially high and converges to the condensation heat of water (45 kJ$mol 1) at 7 water molecules/BAS. Taken from https://doi.org/10.1002/anie.201812184.

170

Dynamic evolution of catalytic active sites within zeolite catalysis

It goes without saying that the formation of protonated clusters of solvent inside the zeolite catalyst has deep consequences in terms of reactivity, that has led to a significant amount of theoretical investigations. The presence of many (different) adsorbed molecules makes the use of static methodologies unsuitable as a lot of configurations should be taken into account95. Therefore, advanced sampling techniques have been employed to elucidate proton mobility, as presented in Section 6.08.3.2.1.

6.08.2.2

Framework-associated and extra framework aluminum (EFAL)

So far we discussed the transformations occurring at the BAS at reaction conditions without an effective modification of the zeolite skeleton. Now, the scope of active site modification will be extended in time by considering the breaking of framework bonds, which is normally associated with the (partial) extraction of aluminum defects. Zeolite dealumination is a wide and complex field of research. Many reactions can be involved, depending on zeolite loading, temperature, pH, neutralizing cation, etc. Moreover, dealumination is strongly related to desilication and therefore to the processes of mesopore formation and zeolite dissolution in medium and harsh water treatment, respectively. The purpose here is to provide a general overview of the current understanding in the EFAL chemistry of formation and effect on catalysis. For a more exhaustive discussion, the interested reader is referred to some recent literature reviews focused solely on the topic.6,96,97 By exposing zeolites to steam at high temperatures, to hot liquid water or, more in general, after hydrothermal synthesis, Al atoms can be extracted from the framework.96 If the Al still maintains covalent bonds with the framework, it is referred to as framework-associated Al, while if it forms a physisorbed cluster, it is called an Extra Framework ALuminum (EFAL).6 Despite EFALs being often associated with Lewis acidity, a survey of the available literature unravels that a lack of correlation exists between the amount of EFAL species and the measured Lewis acidity (Fig. 5).6 This is likely caused by the fact that many techniques exist to probe the Al location in the framework and the Lewis acid character, which may lead to various results depending on multiple factors (as a trivial example, the dimensions of a probe basic molecule when quantifying acidity with IR spectroscopy). Concerning the formation of EFAL and framework-associated Al, elucidating the mechanism of zeolite dealumination is a challenging task both experimentally and computationally.97,98 While the zeolite framework has mostly been considered inert unless subjected to relatively high temperatures and/or water pressures, it has recently become obvious that its bonds are much more dynamic than previously thought. Both SieO and AleO bonds have been found to be very labile already at room temperature when liquid water is present in the pores of the H-SSZ-13 zeolite. The barriers for SieO and AleO bond cleavage were calculated

Structural models Framework aluminium H TO

Measurement of Al acidity: Lewis acid sites Brønsted acid sites

O

Si

TO

OSi

Al

OT SiO

Measurement of Al coordina Three-coordinated Tetrahedral (Td) Penta-coordinated Octahedral (Oh)

OSi

Extra-framework aluminium

Techniques: FTIR of adsorbed probe molecules Structur e r Typical experimental c er dehydra vacuum or dry gas

OH2 HO OH Al H2O OH2 OH

HO H O

Al OH

OH

HO

OH

H O

Al OH

H

2+

OH

OH

Al

OH OH H

3+

O

Al

O O H O H Al HO

Techniques: ²⁷Al MAS NMR Al K-edge XAS XPS (Al KLL peak)

OH

Framework-associated aluminium Si Si

O O Si

OH Al O

Si OH2 OH2

Si

O O

OH Al OH

Si

Si OH2

O

OH2

HO

OH Al OH

OH2

Si

OH2

Si

Al

OH2

O

Si

O O

OH Al

OH2

Typical experimental c In hydrated state for ²⁷Al MAS NMR

OH

o structure

Number of Lewis acid sites (a.u.)

Linking acidity and coordina

O O

Si

OH

Amount of EFAL (a.u.)

Fig. 5 Common methods for the measurement of the EFAL amount (right) and Lewis acidity (left). The central green panel shows some possible structures for EFALs and framework-associated Al while the bottom graph the lack of correlation between number of LAS and EFAL with data from the available literature. Adapted from https://doi.org/10.1038/s41563-020-0751-3.

Dynamic evolution of catalytic active sites within zeolite catalysis

171

to be only  60 and  25 kJ mol 1, respectively.99 This lability has also been confirmed experimentally by monitoring the easy integration of 17O in multiple zeolite frameworks when exposed to H217O through NMR spectroscopy.100 In the same study, it was also highlighted how Si-O-Al bridging oxygens are exchanged faster than Si-O-Si ones, in line with the computational results. Similarly, T-O-T reversible bond breaking in H-SAPO-34 has been proved by introducing in the framework bulky organic molecules, that would not be able to diffuse through the material unless the cages-connecting windows are enlarged by fast and reversible hydrolysis of the framework bonds.101 Moving to harsher treatments, the full dealumination mechanism at steaming conditions (T > 673 K,  1 atm of steam) has been thoroughly investigated in the literature. Initial mechanistic investigations were performed using static DFT simulations.102,103 A first proposal by Malola et al.103 foresaw the passage through the formation of a vicinal disilanol104–106 defect (^Si(OH)2Al^) in H-SSZ-13, but was leading to prohibitively high reaction barriers (190 kJ mol 1) for dealumination. Afterwards, a much more affordable mechanism (barrier of  100 kJ mol 1) starting with a water molecule attacking the Al atom in anti with respect to the bridging hydroxyl group was proposed107–109 and later confirmed experimentally with 27Al NMR, with which it was found that Al sites located in the channel intersection of H-ZSM-5dwhose anti position is the most exposeddare more susceptible to abstraction.110 The role of the number of water molecules on the dealumination has also proven to be important. While static calculations can still provide interesting insights,111,112 a step forward towards the achievement of more realistic experimental conditions was performed by Nielsen et al., who used advanced sampling techniques to assess the influence of multiple water molecules on the dealumination barriers. The water molecules were shown to collectively cooperate in the extraction of the Al atom from the framework, by facilitating the proton transfers between different oxygens. The reaction proceeds through a stepwise cleavage of the AleO bonds, with initial expansion of the Al coordination sphere by anti water addition. The barriers for each step are in the order 90– 100 kJ mol 1, similar to the results from static calculations, but increase by 10–20 kJ mol 1 when the amount of water molecules per BAS is reduced from 3 to 1 (Fig. 6).113 Interestingly, the existence of partially hydrolyzed Al atoms, i.e., the intermediates in the proposed mechanism, has been identified multiple times in the literature based mainly on 27Al NMR experiments and they are expected to play an important catalytic role at reaction conditions (see Fig. 7C and D).114–116 Once the Al is extracted from the framework, it is still unclear in which form it is more likely to persist in the catalyst. Initial theoretical investigations assumed the existence of mononuclear hydroxy-aquo complexes of the form [Al(OH)x(H2O)n  x]3  x.117–119 27Al NMR experiments on H-ZSM-5 seem to suggest that a tetrahedral coordination is predominant, with octahedral Al appearing only at very long steaming times ( 20 h).120 Similarly, XANES measurements in H-b and H-MOR indicated that octahedral Al is present only at low temperatures and quantitatively converts in tetrahedral Al above 395 K.121 Octahedral Al can however be relevant in other conditions. Recent 27Al NMR experiments coupled with IR spectroscopy have shown that

Fig. 6 Free energy diagram of the stepwise dealumination in H-SSZ-13, as obtained from DFT-MD umbrella simulations. The values close to the arrows show the intrinsic free energy barriers of each step with one (gray dashed line) and three (blue line) water molecules per Al atom. Reprinted with permission from Nielsen, M.; Hafreager, A.; Brogaard, R.Y.; De Wispelaere, K.; Falsig, H.; Beato, P.; Van Speybroeck, V.; Svelle, S., Collective Action of Water Molecules in Zeolite Dealumination. Cat. Sci. Technol. 2019, 9(14), 3721–3725. Copyright 2019, The Royal Society of Chemistry.

172

Dynamic evolution of catalytic active sites within zeolite catalysis

Fig. 7 DFT-optimized structures of some recently proposed (partially) hydrolyzed aluminum species, as derived from NMR measurements. (A and B) EFAL in synergistic proximity of a BAS. (C and D) Framework-associated aluminum. (A and B) Adapted with permission from Chen, K.; Abdolrahmani, M.; Horstmeier, S.; Pham, T.N.; Nguyen, V.T.; Zeets, M.; Wang, B.; Crossley, S.; White, J.L., Brønsted–Brønsted Synergies Between Framework and Noncrystalline Protons in Zeolite H-ZSM-5. ACS Catal. 2019, 9(7), 6124–6136. Copyright 2019, American Chemical Society. (C and D) Adapted with permission from Chen, K.; Horstmeier, S.; Nguyen, V.T.; Wang, B.; Crossley, S.P.; Pham, T.; Gan, Z.; Hung, I.; White, J.L., Structure and Catalytic Characterization of a Second Framework Al (IV) Site in Zeolite Catalysts Revealed by NMR at 35.2 T. J. Am. Chem. Soc. 2020, 142(16), 7514–7523. Copyright 2020, American Chemical Society.

the amount of framework-associated octahedral Al correlates with the Lewis acidity in wet H-MOR, representing a significant amount of the overall number of Lewis acid sites.122 Tricoordinated extra-framework Al atoms with exceptional Lewis acidity have also been proposed for dealuminated HY zeolite, based on the results of NMR with trimethylphosphine as a probe molecule.123 The true structure of framework and extra-framework Al is thus still an open question in the literature, as the rich structural variety of Al likely makes many structures possibly coexisting in the catalyst6 (and references therein). Molecular modeling can in this case provide some useful insights. Pidko and co-workers proposed that the preferred state for EFAL species at reaction conditions is the polynuclear cluster [Al3O4H3]4þ in the faujasite zeolite, using a thermodynamic model coupled with the data from static DFT calculations.11,124 These findings are, however, still waiting for experimental validation. Interestingly, the same authors also calculated that the activity towards alkane cracking of a BAS next to the polynuclear EFAL cluster is increased.125 This increase in activity related to the synergistic proximity of BAS and EFAL species is of course deeply intertwined with the structural characterization of the EFAL itself and represents a remarkable subject of interest in the current investigation of zeolite chemistry. A synergistic enhancement of the Brønsted acidity in the H-Y zeolite caused by proximity of EFAL species was initially proposed based on NMR evidences coupled back by DFT simulations.126 The same authors also suggested that after hydration of the zeolite, the EFAL species tend to adopt an octahedral geometry and remain close to the framework aluminums because of strong H-bonds between the coordinating water molecules and the O atoms of the framework.127 Very recently, moreover, technological advances in NMR spectroscopy allowed to discover a minor, very deshielded and relatively broad peak at 12–15 ppm in the 1H spectrum of the H-ZSM-5 zeolite (Fig. 3).128 Such peak was proposed to be associated with a tetrahedral or triangular planar EFAL species in direct hydrogen bonding with a BAS (Fig. 7A and B),129 thus providing further evidence for a widespread proximity between framework and extra-framework aluminum. In conclusion, the past depiction of zeolite active sites as well-defined framework-bound protons has been deeply challenged in the last decades and a new picture of the catalyst, with a more complex chemistry has emerged. Not only the BAS can be mobilized at typical reaction conditions, but also the framework itself has been shown to be subject to significant chemical reactivity. Unraveling the nature and catalytic activity of framework-associated and extra-framework aluminum species is nowadays a very active research field. While the positive effect of Brønsted-Lewis and Brønsted-Brønsted synergies on catalytic reactions such as the prototypical alkane H/D exchange is accepted,7,130,131 future work will still be needed to achieve a realistic speciation of Al in zeolites at operating conditions, thereby leading to further improvement in our understanding of these fascinating materials.

6.08.2.3

Mobility of active sites in TM-exchanged zeolites

Transition metal (TM) exchanged zeolites have received significant attention over the past few decades as promising single-site catalytic materials.38,132–134 Akin to Brønsted acid sites, as outlined in the previous section, exchanged TM sites in zeolites can mobilize under certain environmental conditions leading to divergent catalytic behavior. Arguably the most notable example is the mobilization of copper sites in chabazite during the ammonia assisted selective catalytic reduction of nitrous oxides (NH3-SCR-NOx, Fig. 8), where summaries of key spectroscopic, kinetic and theoretical studies investigating this phenomena constitute the bulk of this section. It should be noted that while the role of mobilized Cu ions in SCR reactivity had been implied as early as the 90s,135 and mobility of Cu has been identified in Cu-Y and Cu-ZSM-5 systems under conditions relevant for NH3-SCR,136 this article focusses on the CHA topology. Other selected examples of TM site mobility are also outlined, such as Pd mobility in SSZ-13, Rh mobility in H-Y during ethene oligomerization & hydrogenation, and mobility of Ag clusters/nanoparticles in MFI during propane-assisted SCR reactivity. These are only a few examples which clearly show the dynamic rearrangements of TMexchanged zeolites, however, based on these cases, it is clear that dynamic evolution of active sites is a much more general concept within the field of zeolite catalysis.

Dynamic evolution of catalytic active sites within zeolite catalysis

173

Fig. 8 Illustration of mobilization of cationic Cu sites in the low-temperature ammonia-assisted selective catalytic reduction of NOx, catalyzed by Copper-exchanged chabazites. Cu ¼ orange, Si ¼ yellow, O ¼ red, Al ¼ purple, H ¼ white.

6.08.2.3.1

Mobility of copper sites in Cu-CHA during low-temperature NH3-SCR-NOx

NH3-SCR-NOx is a widely employed process in the global effort to combat high NOx emissions from industry and vehicle use, given the well-documented health impacts these gasses can cause if they remain largely abundant in the atmosphere.137 Metal-exchanged zeolites are well-studied materials for this purpose,138–140 offering efficient and robust catalytic systems that are less toxic than industrial vanadium-based catalysts such as V2O5-WO3/TiO2.141 NH3-SCR reactivity in zeolites can be categorized into three subtypes: (i) Standard SCR, in which NO is exclusively consumed upon reduction with NH3 and O2 4NO þ O2 þ 4NH3 /4N2 þ 6H2 O (ii) Fast SCR, in which both NO and NO2 are consumed in equal quantities: NO þ NO2 þ 2NH3 /2N2 þ 3H2 O (iii) NO2 SCR, in which NO2 is reduced exclusively upon reduction with NH3: 2NO2 þ 2NH3 /2N2 þ N2 O þ 3H2 O 6NO2 þ 8NH3 /7N2 þ 12H2 O

Copper-exchanged zeolites in particular have received significant attention for their propensity to facilitate “standard NH3-SCR” (in which NO is exclusively consumed upon reduction with NH3 and O2). Initial discoveries in the 1970s utilized faujasite (zeolite Y),142 and in the 1980s seminal work by Iwamoto and coworkers had identified Cu-SZM-5 as suitable catalysts for SCR reactivity.143–146 However, more recent reports emerged in the early 2010s outlining that Cu-exchanged chabazites, i.e., SSZ-13 and SAPO-34, are superior catalysts with greater catalytic efficacy147 and hydrothermal stability,148 and as a result Cu-SSZ-13 is now widely used in the commercial treatment of diesel engine emmisions.136,140 Despite these significant advances in the development of Cu-CHA catalytic systems, the speciation and nature of the active copper sites during NH3-SCR reactivity is not entirely resolved and is an active area of research in its own right.42 While this speciation is dependent on physiochemical factors such as gas-feed composition, temperature and also the framework composition itself such as the Si/Al ratio, cationic Cu-exchange has been shown to prevalently take place in which ions are anchored to 6-ring and 8-ring sites as Cu(I), Cu(II) and [CuOH]þ species (Fig. 9).149,150 Herein, a selection of key papers are outlined that elucidate the mobilization of these Cu catalytic sites during NH3-SCR-NOx reactivity, highlighting the importance of operando experimental and modeling techniques in unraveling this phenomenon. 6.08.2.3.1.1 Identification of catalytically relevant mobile [Cu(NH3)2]þ under low-temperature standard SCR Following initial reports identifying the catalytic capabilities of Cu-CHA, in situ studies revealed that solvation of Cu sites with ligating probe molecules can liberate them from the framework. Kwak et al. employed H2-temperature programmed reduction (TPR) and FTIR to probe the nature of the cationic copper sites in Cu-SSZ-13, identifying that addition of H2O (1%) into the H2 gas stream significantly altered the reduction profiles, interpreted as the Cu sites moving from their original positions anchored to the framework upon H2O-solvation. 151 Gao and coworkers probed Cu-SSZ-13 under conditions of NO oxidation, NH3 oxidation and NH3-SCR conditions using EPR spectroscopy and TPR studies,152 finding that prior to full dehydration of the sample hydrated Cu(II) species were present and mobile, as indicated by multiple EPR spectroscopic features identified at the high-field, and low-field hyperfine structures being partially shielded. The lack of these features at higher temperature (250  C) dehydrating conditions indicated that Cu ions are conversely immobile. The work of Gao and coworkers also provided some key evidence for the presence of transient dimers under low-temperature SCR conditions, which will be covered in Section 6.08.3.2.4. Szanyi and coworkers employed TP-XRD, XANES and vibrational (DRIFTS) spectroscopy to probe changes in the copper coordination environment in Cu-SSZ-13 during calcination, reduction with CO, and adsorption of CO and H2O.153 DRIFTS particularly proved to be a useful technique, where, for instance the adsorption of CO to form [Cu(CO)2]þ complexes at temperatures below 100  C could be seen to alter the Cu influence on the asymmetric T-O-T (nasym(TOT)) region of the IR spectra. Solvation of Cu(II) sites by NH3 was also identified by Beale and coworkers in combined theoretical and in situ FTIR studies.154 At 250  C, three distinct NH3

174

Dynamic evolution of catalytic active sites within zeolite catalysis

Fig. 9 Spectroscopically and computationally characterized locations of Cu(I) and Cu(II) ions at 6mr and 8mr rings upon cation-exchange. Cu ¼ orange, Si ¼ yellow, O ¼ red, Al ¼ purple, H ¼ white.

species were identified, where one of which, a [Cu(NH3)4]2þ complex, could be confirmed via FTIR with observation of frequencies in the 3100–3400 cm 1 range, at 1619 cm 1 and at 1278 cm 1, which correspond to different contributing NH3 vibrational modes. Theoretical absorbance bands obtained via Fourier-based analysis of AIMD simulations agreed with experimental bands and further supported formation of [Cu(NH3)4]2þ. Further PD, XAS and DFT studies by Beale and coworkers supported formation of untethered Cu species upon NH3-solvation at room temperature.155 For example, peaks at  2.0 Å and > 2.50 Å in the Cu K-edge k3-weighted EXAFS spectra fit with a N coordination number of 4 and Cu-N distance of 2.04 Å, and a N coordination number of 2 with a distance of 2.71 Å, consistent with a liberated [Cu(NH3)6]2þ species residing in the center of the cha cage. FTIR, XANES, XES and DFT studies by Bordiga and coworkers156 also showed that NH3 adsorption can result in the reduction of Cu(II) to Cu(I) with emergence of the 8982.5 eV peak in the XANES spectra (Fig. 10A), which, along with a high intensity of this peak, suggested a linear species in the form of OfweCu(NH3) (bound to one oxygen of the zeolite) or [Cu(NH3)2]þ. This was reflected in DFT calculations on cluster models (PBE/TZVP level) where one equivalent drives the Cu out of the 6mr plane during structure optimization, while two form mobilized [Cu(NH3)2]þ (Fig. 10C and D). While the studies outlined above demonstrate that adsorbates such as H2O, CO, NH3 can liberate the cationic Cu sites, the conditions were not altogether identical to those during NH3-SCR-NOx. Subsequent studies featuring operando spectroscopic techniques with complementary ab initio calculations could instead evaluate the speciation of Cu under conditions more explicitly comparable to NH3-SCR-NOx, and a selection of these are summarized below. Schneider and Gounder interrogated the speciation of Cu under various ex situ and in situ conditions in samples of Cu-SSZ-13, in which the 6-rings contain either 1 aluminum (1Al) or two (2Al), via XAS, static DFT and DFT-MD.157 Analysis under ex situ conditions indicated that Cu(II) preferentially populates the 2Al 6-ring sites to form “Z2Cu(II)” and at 1Al 6-ring sites to form “ZCuOH,” and under hydrating conditions Cu sites can mobilize via solvated copper aqua complexes, consistent with reports summarized previously. Exposure to a subset of catalytic conditions of 300 pm NO and NH3 at 473 K reduces all copper sites to Cu(I), where XANES was fitted to a Cu(I) fraction of 100%. AIMD simulations of Z[Cu(NH3)2]/[ZNH4] and Z[Cu(NH3)2] systems revealed the [Cu(NH3)2]þ species are highly mobile (Fig. 11), while XAS analysis found that a lone peak at 1.89 Å in the EXAFS region fits to a coordination number of 2 with respect to N or O. Conversely, analysis of samples with an oxidizing feed of 300 pm NH3 and 10% O2 indicated formation of [Cu(NH3)4]2þ complexes that are  50% less mobile than [Cu(NH3)2]þ. Operando EXAFS measurements on a Si:Al ¼ 25, Cu:Al ¼ 0.42 sample revealed a 60:40 Cu(I)/Cu(II) ratio, a lack of second-shell structure and a fit of 3.1 with respect to the CN, consistent with an admixture of [Cu(NH3)2]þ/[Cu(NH3)4]2þ, further evidencing that Cu sites are solvated and mobilized under catalytic conditions. Characterization of the active sites at Cu-SSZ-13 under NH3-SCR-NOx conditions between 150  C and 400  C was also undertaken by Bordiga and coworkers via operando XANES, XAS and vtc-XES.158 For reference, spectra of Cu(I) and Cu(II) aqua/amino complexes were obtained for linear combination fit (LCF) analyses. At low-temperature conditions (150  C) the XANES region

Dynamic evolution of catalytic active sites within zeolite catalysis

175

Fig. 10 (A) Evolution of the Cu K-edge XANES spectra of Cu-SSZ-13 with O2/He and a 1300 pm gas-feed of NH3 (in He) at 120  C. Inset: pre-edge region. (B) Cu Kß2.5 and Kß00 emission lines for initial and final states after exposure to NH3. (C and D) Local 6mr environment of the DFT-optimized zeolite clusters upon exposure to one (C) or two (D) equivalents of NH3 to Cu, computed at the PBE/TZVP level. Reprinted with permission from Giordanino, F.; Borfecchia, E.; Lomachenko, K.A.; Lazzarini, A.; Agostini, G.; Gallo, E.; Soldatov, A.V.; Beato, P.; Bordiga, S.; Lamberti, C., Interaction of NH3 With Cu-SSZ-13 Catalyst: A Complementary FTIR, XANES, and XES Study. J. Phys. Chem. Lett. 2014, 5(9), 1552–1559, with permission from the American Chemical Society.

was consistent with spectra obtained by Schneider and coworkers,157 and LCF analyses identified an admixture of 46% [Cu(NH3)2]þ, 25% [Cu(NH3)4]2þ and 26% remaining zeolite-bound Cu(II) species in the sample. The existence of mobilized Cu species could be validated via analysis of FT-EXAFS spectra, in which the second-shell region of  2.3 Å (a fingerprint for framework-bound Cu species) is perturbed at lower temperatures (Fig. 12), while at higher temperatures this peak is present and agrees well with reference Z-Cu(II) spectra, thus identifying a temperature-dependence of Cu mobility. Boronat and coworkers employed ab initio MD simulations, static DFT calculations and IR spectroscopy to probe the dynamic nature of copper mobility in Cu-SSZ-13 and Cu-SAPO-34 during NH3-SCR.159 AIMD simulations at 298 K and 523 K on unit cells featuring different gas feed adsorbates (NO, O2, NO2, NH3) were performed to probe changes in the copper coordination mode (Fig. 13A). NH3 could be seen to mobilize the Cu ions to form [Cu(NH3)2]þ, while NO, NO2 and O2 only mildly displaced Cu from the 6-ring, as shown by RMSD analysis of Cu positions in each simulation (Fig. 13B). IR spectra recorded for Cu(I) and Cu(II) states in Cu-SAPO-34 samples with different catalytically relevant feeds at low- and high-temperatures also unraveled a dynamic nature of copper mobility (Fig. 14). In the initial Cu(I) state n(TOT) vibrations in the 800–1000 cm 1 region disappear upon exposure to NH3 at lower-temperatures, and reemerge at high temperatures (623 K). With an initial Cu(II) state and a gas feed ratio of 5:2:1 O2:NO:Cu, the nasym(TOT) fingerprint remains throughout the 298– 623 K temperature range, and only when NH3 is added to the gas feed (2:1 NH3:Cu) at 423 K the n(TOT) fingerprint disappears again, in which complementary analysis of EXAFS identifies this is concomitant with Cu(II) / Cu(I) reduction. 6.08.2.3.1.2 The role of mobilized copper in the oxidation half-cycle during NH3-SCR-NOx With the identification of mobilized [Cu(NH3)2]þ species forming under low-temperature catalytic conditions, their role in the Cu(I) / Cu(II) oxidation half-cycle during NH3-SCR-NOx was not immediately clear. As previously discussed, Gao and coworkers employed kinetics studies and EPR measurements to probe Cu-SSZ-13 and identified highly mobile hydrated Cu species,152 where in samples with low to intermediate cationic copper loadings (Cu/Al  0.22), transient dimeric Cu species were identified to form under catalytic conditions via interaction of pairs of [Cu(NH3)2]þ, as indicated by a quadratic dependence on Cu site density on the SCR rate. Moreover, EPR spectroscopic features were consistent with dipolar interactions between Cu(II) ions. In a subsequent

176

Dynamic evolution of catalytic active sites within zeolite catalysis

Fig. 11 Copper cation positions (shown as gray balls) inside the SSZ-13 cage across 90 ps AIMD simulations. Reprinted from Paolucci, C.; Parekh, A.A.; Khurana, I.; Di Iorio, J. R.; Li, H.; Albarracin Caballero, J.D.; Shih, A.J.; Anggara, T.; Delgass, W.N.; Miller, J.T.; Ribeiro, F.H.; Gounder, R.; Schneider, W.F., Catalysis in a Cage: Condition-Dependent Speciation and Dynamics of Exchanged Cu Cations in SSZ-13 Zeolites. J. Am. Chem. Soc. 2016, 138(18), 6028–6048, with permission from the American Chemical Society.

Fig. 12 (A) FT-EXAFS spectra (top: moduli, bottom: imaginary parts of the Fourier transform) of the Cu-SSZ-13 sample collected during SCR reaction conditions at a range of temperatures. (B) FT-EXAFS spectra of references employed for analysis of the operando spectra, i.e., via LCF, by Borfecchia and coworkers. Reprinted from Lomachenko, K.A.; Borfecchia, E.; Negri, C.; Berlier, G.; Lamberti, C.; Beato, P.; Falsig, H.; Bordiga, S., The Cu-CHA deNOx Catalyst in Action: Temperature-Dependent NH3-Assisted Selective Catalytic Reduction Monitored by Operando XAS and XES. J. Am. Chem. Soc. 2016, 138(37), 12025–12028, with permission from the American Chemical Society.

Dynamic evolution of catalytic active sites within zeolite catalysis

177

Fig. 13 (A) Snapshots of Cu(I) and Cu(II) cationic species at the 6mr in Cu-SAPO-34 with probe adsorbates (labeled below each structure), which were generally most visited during 100 ps AIMD runs at 298 K. (B) Average and RMSD distances of the Cu cations traveled from the 6mr at 298 K (blue) and 523 K (green). Reprinted from Millan, R.; Cnudde, P.; Hoffman, A.E.J.; Lopes, C.W.; Concepción, P.; van Speybroeck, V.; Boronat, M., Theoretical and Spectroscopic Evidence of the Dynamic Nature of Copper Active Sites in Cu-CHA Catalysts under Selective Catalytic Reduction (NH3– SCR–NOx) Conditions. J. Phys. Chem. Lett. 2020, 11, 10060–10066, with permission from the American Chemical Society.

Fig. 14 Flow chart summarizing the oxidation states of cationic Cu sites (red) identified by EXAFS, the speciation of the Cu sites identified by IR (blue), and the presence of asymmetric v(TOT) peaks as a function of temperature and gas feed in Cu-SAPO-34 (also blue). Reprinted from Millan, R.; Cnudde, P.; Hoffman, A.E.J.; Lopes, C.W.; Concepción, P.; van Speybroeck, V.; Boronat, M., Theoretical and Spectroscopic Evidence of the Dynamic Nature of Copper Active Sites in Cu-CHA Catalysts Under Selective Catalytic Reduction (NH3–SCR–NOx) Conditions. J. Phys. Chem. Lett. 2020, 11, 10060–10066, with permission from the American Chemical Society.

study, Gao and coworkers also identified that the SCR rate (mol NO g 1 s 1) was quadratically dependent on Cu/Al ratio (Fig. 15B) under low temperature conditions (200  C), while a linear dependence is observed at 380  C (Fig. 15A). 160 DFT calculations (PBED3 level) revealed that O2 activation proceeds most accessibly with participation of two [Cu(NH3)2]þ complexes to form bridged [Cu(NH3)2]-O2-[Cu(NH3)2] species rather than activation over an individual [Cu(NH3)2]þ complex. Cu(I) oxidation at the bridged

178

Dynamic evolution of catalytic active sites within zeolite catalysis

Fig. 15 (A) SCR rate (mol NO g 1 s 1) against Cu/Al ratio at 200  C (top) and 380  C (lower). (B) SCR rate against (Cu/Al)2. Reprinted from Gao, F.; Mei, D.; Wang, Y.; Szanyi, J.; Peden, C.H.F., Selective Catalytic Reduction Over Cu/SSZ-13: Linking Homo- and Heterogeneous Catalysis. J. Am. Chem. Soc. 2017, 139(13), 4935–4942, with permission from the American Chemical Society.

species was then characterized to take place via NO addition and subsequent NO2 release to afford [(NH3)2-Cu(II)-O-Cu(II)NH3)2]2þ, followed by hydrolysis to yield [(NH3)2Cu(OH)]þ. At low Cu site densities diffusion of [Cu(NH3)2]þ to a neighboring cage was proposed to be rate-limiting, while formation of the [(NH3)2-Cu(II)-O-Cu(II)-NH3)2]2þ dimer is proposed as rate-limiting at higher site densities. A computed diffusion coefficient (D) for [Cu(NH3)2]þ of 4.4  10 12 m2 s 1 was obtained via AIMD (NVT, 200  C, D ¼ RMSD/6t) which supports a diffusion-controlled process at low temperatures and Cu-loadings. Separate Cu-density dependent kinetic schemes were established by Paolucci and coworkers, where for densities over 1.9  10 4 A 3 the rate increases linearly, while at lower densities of < 1.13  10 4 A 3 the quadratic dependence was observed. 20 LCF analyses of Cu K-edge XANES region of different Cu-CHA samples outlined an inverse relation of Cu-Cu distance with Cu(I) fraction, providing evidence for an alternative non-single site mechanism at play whereby the rate of Cu(I) oxidation increases with greater site densities. This behavior was rationalized by an O2-consuming step in the oxidation half-cycle being sensitive to the Cu density. Monitoring the rate of decay of Cu(I) under O2 in samples with differing site densities via Cu K-edge XANES could also identify decay is second-order in Cu(I), implicating the following pseudobimolecular mechanism:  þ 2þ  2 CuðNH3 Þ2 þ O2 / ðNH3 Þ2Cu–O2 –CuðNH3 Þ2 Static DFT calculations (PBE-D2 level) on a 12-T site supercell with two [Cu(NH3)2]þ complexes charge-compensated by 2 Al atoms, revealed that initial [Cu(NH3)2]þ diffusion to a neighboring cage is facile with an activation barrier of 35 kJ mol 1, and subsequent coordination of O2 to the two close by Cu(I) ions is favored ( 59 kJ mol 1) over a single-ion binding mode ( 26 kJ mol 1) to form a triplet dimeric species. Spin-forbidden rearrangement to form a di-oxo structure then takes place, where

Dynamic evolution of catalytic active sites within zeolite catalysis

179

each Cu(II) ion adopts a square-planar geometry. However, to gain more accurate insight into the energetics of [Cu(NH3)2]þ diffusion, metadynamics simulations were performed (see Section 6.08.3.2.4). Recently XAS and diffuse reflectance UV-Vis-NIR spectroscopic studies have been used to further probe the structure of the [Cu2(NH3)4O2]2þ species.161 A starting state of mobilized [Cu(NH3)2]þ could be formed via exposure of pre-treated Cu-CHA to 1000 pm NO and 1000 pm NH3 at 200  C and confirmed by the 8982.5 eV fingerprint in the Cu K-edge EXAFS spectra. Exposure of this sample to O2 (10% in He) revealed an almost complete disappearance of the fingerprint, indicating the oxidation of most [Cu(NH3)2]þ species. This was further supported by UV-Vis-NIR with a red-shift of the ligand-to-metal charge transfer (LMCT) peak from 35 k to 25 k, along with emergence of a d-d absorption peak at 13850 cm 1. Signals in the NIR region (6515 cm 1 and 4970 cm 1) arising from combination modes of NH3 and NH4þ, also indicated NH3 is still ligated to the oxidized Cu(II) sites. The precise structure of this state was probed by fitting of the FT-EXAFS spectra against gas-phase DFT structures (M06-HF-D3/Def2-TZVP level), and the best fit obtained supported a side-on m-h2,h2-peroxo diamino dicopper(II) species [Cu2(NH3)4O2]2þ (Fig. 16A(iii)) over an end-on trans m-1,2-peroxo binding mode (Fig. 16A(ii)). Subsequent exposure of the sample featuring [Cu2(NH3)2O2]2þ to NH3 shows a partial re-emergence of the 8982.5 eV peak, rationalized by partial reformation of monomeric [Cu(NH3)2]þ, while in the UV-Vis-NIR spectrum a shift from 13,800–14,400 cm 1 and a lowering in intensity also suggests some Cu(II) remains. LCF analysis of the XANES region also identified partial reduction of Cu(II) to Cu(I), with an admixture of 65% [Cu(NH3)2]þ and 35% [Cu(NH3)3(X)]þ, which was considered to be either [Cu(NH3)3(OH)]þ, as previously predicted by Paolucci and coworkers,157 and computationally identified by Shimizu and coworkers,162 or [Cu(NH3)3(OO)]þ, which was proposed based on LCF analysis of the XANES region spectra. Conversely, exposure of the oxidized [Cu2(NH3)2O2]2þ species to NO disaggregates the Cu centers along with reduction to Cu(I), as seen by the reemergence of the 8982.5 eV peak, with concomitant formation of N2. UV-Vis-NIR spectra shows the minimization of the intensity of the 13,850 cm 1 peak corresponding to [Cu2(NH3)2O2]2þ and emergence of a series of peaks at around 20,000, 16,350, 13,300 and 10,600 cm 1. The Cu K-edge XANES region of the XAS spectra is also consistent with a Z[Cu(NH3)] species (Fig. 16B(ii)). The aforementioned operando studies identifying the role of mobile [Cu(NH3)2]þ and resulting dimeric [Cu(NH3)2O2]2þ in the low-temperature NH3-SCR pathway can inform new proposals of full catalytic cycles for the low-temperature NH3-SCR-NOx pathway. Moreover, as transient intermediates can be unidentifiable, even via operando spectroscopic techniques, DFT calculations can fill in the gaps, serving to provide insight in proposing a complete mechanism for such an intriguing reaction. A recent example of the application of DFT to provide a full working catalytic cycle of NH3-SCR-NOx is by Grönbeck and coworkers.163 Calculations at the PBE þ U-D3 (a LOT which the authors identified as a strong performer based on prior benchmarking studies)164, revealed

Fig. 16 (A) Forms of [Cu(NH3)2O2]2þ proposed and considered in the study by Berlier and coworkers: (i) trans-m-1,2-peroxo diamino dicopper(II), (ii) bis-m-oxo diamino dicopper(III), and (iii) m-h2,h2-peroxo diamino dicopper(II), supported by fitting against operando XAS measurements. (B) Proposed products of (i) NH3 addition and (ii) NO addition to m-h2,h2-peroxo diamino dicopper(II).

180

Dynamic evolution of catalytic active sites within zeolite catalysis

Fig. 17 Two catalyst cycles proposed and characterized with DFT at the PBEþU-D3 level, by Grönbeck and coworkers, differing in the binding mode of NO at the oxidized [Cu2(NH3)4(O2)]2þ species.

initial activation of O2 at a mobilized [Cu(NH3)2]þ pair takes place to form the m-h2,h2-peroxo dimer, supporting the work of Berlier and coworkers.161 Two divergent cycles were characterized based on the initial binding mode of NO to the resulting dimer (Fig. 17). Cycle I considers NO coordination at Cu2þ and subsequent NH3 addition to form a NONH3 ligated complex. The overall formation of this [Cu2(NH3)4(O2)(NONH3)]2þ species from [Cu(NH3)2]þ monomers is turnover-limiting. Proton transfer from the -NH3 moiety can then take place to release H2NNO and form a bridging OH group, where H2NNO decomposes to N2 and H2O upon diffusion from the Cu sites to an available Bronsted acid site. The process of NO þ NH3 uptake to release H2NNO is repeated to form a bridged hydroxide dimer, upon which further addition of NO and final release of HONO can complete the cycle. Cycle II conversely begins with NO binding to O2 to form a [Cu(NO2)2(NH3)2]2þ species, upon which Cu separation takes place to form [Cu(NH3)2(NO2)]þ pairs. At each Cu site addition of NO and NH3, followed by release of H2NNO then takes place, where formation of HONO and H2NNO is turnover-limiting.

Dynamic evolution of catalytic active sites within zeolite catalysis 6.08.2.3.2

181

Solvation and mobility of Pd in SSZ-13

Much like in Cu-CHA, palladium sites in SSZ-13 have also been recently identified to mobilize under hydrating conditions by Paolucci et al. via combined in situ spectroscopic, kinetics and theoretical studies.165 Ab initio thermodynamic analysis of a series of candidate “Z[PdHXOY]” species was carried out at the HSE06-D3(BJ) level at two separate temperature regimes (298 K and 773 K). This identified that at lower temperatures Pd(II) sites preferentially form [Pd(H2O)4]2þ under hydrating conditions in the chabazite cages, while framework-coordinated Z2Pd(II) at 6-ring sites featuring two Al-substitutions is preferred at higher temperatures (Fig. 18). This temperature-dependent mobilization of Pd, analogous to Cu mobilization outlined in the previous section, was supported by in situ XAS experiments, where the XANES region of Pd-SSZ-13 exposed to 10% O2 and 3.1% H2O at 473 K agreed with reference homogeneous aqueous Pd(II) spectra. Moreover CO oxidation was more amenable at the low-temperature hydrating conditions, implicating the role of mobilized H2O-solvated Pd species in other Pd-zeolite catalytic transformations in hydrating conditions, such as the Wacker oxidation process. 166,167

6.08.2.3.3

Mobility of Rh in zeolite Y and consequences for ethene hydrogenation & oligomerization

Rh-exchanged zeolites have been studied for both ethylene oligomerization and hydrogenation, with efforts tracing back to the 1970s in which Rh-Y was employed for this purpose168–170 with more recent efforts by Gates have also focusing on Rh-Y for ethene dimerizations.171,172 As part of this line of research, in 2011 Gates identified that the speciation of Rhodium in cation-exchanged Rh-Y can interconvert between monomeric Rh(C2H4)2 sites, and small-clustered Rhn sites based on the ratio of H2:C2H4 in the gas feed, thus revealing a mobility of Rh sites during an interconvertible speciation process.22 Reaction of Rh(C2H4)2(acac) with dealuminated zeolite HY (Si/Al ¼ 30) afforded spatially uniform and well-defined Rh(C2H4) sites anchored via two RheO bonds, supported by IR and EXAFS measurements, where IR spectra revealed that  25% of the Bronsted acid sites are Rh-exchanged. Upon exposure of the Rh-Y sample to a flow of H2 at room temperature, small Rh clusters formed, as evidenced by EXAFS measurements, which fitted to a Rh-Rh coordination number of 1.9. Moreover the Rh-O coordination number remained unaffected, indicating that the clusters are anchored to the zeolite framework. This identified that disengagement and mobilization of Rh sites en route to cluster formation can take place by tuning the gas feed composition. The Rh-Y sample was then probed for its catalytic efficacy in ethene hydrogenation and oligomerization as a function of C2H4:H2 ratio in the gas feed, with rates of ethane (hydrogenation) and butane (oligomerization) formation measured (Fig. 19, top) along with operando FT-EXAFS spectra, where cycles of (1) C2H4-rich, (2) H2-rich, (3) pure H2 and (4) H2-rich again were used in the gas feed tuning procedure. As a result, changes in the product selectivity and magnitudes of the FT-EXAFS spectra were seen, indicating changes from monomeric to multinuclear clusters were coinciding with changes in selectivity towards oligomerization or hydrogenation.

Fig. 18 (A) Computed free energies (HSE06-D3(BJ) level) of various “[ZPd(HxOy)z]” states of Pd-speciation at 298 K (blue) and 773 K (red). (B) Phase diagram of Pd-speciation as a function of temperature & pressure constructed via ab initio thermodynamic analysis, showing the preference for solvated [Pd(NH3)4]2þ species at low-temperatures (cobalt blue) to a preference for framework-anchored Z2Pd(II) species at higher-temperatures (yellow). Reprinted from Mandal, K.; Gu, Y.; Westendorff, K.S.; Li, S.; Pihl, J.A.; Grabow, L.C.; Epling, W.S.; Paolucci, C., Condition-Dependent Pd Speciation and NO Adsorption in Pd/Zeolites. ACS Catal. 2020, 10(21), 12801–12818.

182

Dynamic evolution of catalytic active sites within zeolite catalysis

Cycle Step 1

2

3

4

1

2

3

4

1

2

2.1 1.1

1.6

R (Å)

2.6

10-12

a.u

10-11

10-10

He

Time (h)

1

2

3

4

5

Fig. 19 (Top) Changes in selectivity towards butanes (red line) and ethane (green) formation catalyzed by Rh(C2H4)2 supported in zeolite HY, and (bottom) time-resolved Fourier-transform (FT)-k3-weighted EXAFS data, outlining the evolution of the magnitude of the FT over time (bottom X-axis) and composition feed-cycles (top X-axis). Reprinted from Serna, P.; Gates, B.C., Zeolite-Supported Rhodium Complexes and Clusters: Switching Catalytic Selectivity by Controlling Structures of Essentially Molecular Species. J. Am. Chem. Soc. 2011, 133(13), 4714–4717, with permission from the American Chemical Society.

6.08.2.3.4

Mobility of Ag sites in MFI during C3H8-SCR reactivity

Another example of mobility and clustering of active TM sites under reductive conditions, leading to subsequent consequences for propane-assisted SCR (C3H8-SCR) of NO was identified in Ag-MFI by Shibata and coworkers.173,174 Under 673 K the rate of C3H8SCR catalyzed by Ag-MFI-58 was significantly enhanced by the addition of H2, where a maximum NO conversion of 48% was seen at 573 K in the presence of 0.5% H2, and a lower  10% conversion was measured in the absence of 0.5 H2 (Fig. 20).174 This enhanced reactivity assisted by exposure to H2 was shown to be reversible, where upon H2 addition and complementary increased NO conversion rates, removal of H2 suppresses C3H8-SCR to levels observed prior to H2 addition (Fig. 20). The speciation of the cationic Ag(I) sites in both the presence and absence of H2 was then probed via UV-Vis spectroscopy. Upon exposure of the sample to H2, bands at  260 and  284 nm emerged, which, based on comparison with other UV-Vis characterizations of Ag clusters,175–180 was assigned as Agnþ clusters (2  n  4). Subsequent treatment of the H2 addition sample under C3H8-SCR conditions in the absence of H2 led to these Agnþ signals to disappear. Assuming cationic monomeric Ag(I) sites are framework bound in Ag-MFI-58 in the absence of H2, this revealed that H2 exposure mobilizes the Ag cations to form cationic Agnþ clusters. Further assessment of the rate of C3H8-SCR in a range of samples with varying Ag/Al ratios also identified that in the presence of H2 an increase in Ag/Al ratio increases the NO conversion to N2 with a reaction order above 1, thus supporting the notion that the cationic Agnþ clusters, reversibly formed upon H2-induced mobilization of Ag sites, are the catalytically active species in the C3H8-SCR process. Based on kinetic analyses of Ag(I) reduction to Ag2þ by H2-TPR studies, reported previously in AgCHA by Beyer and coworkers,181 the authors proposed that the following mechanism is taking place to form the Agnþ clusters: H2 #2H ðXÞ H þ AgðIÞ#½AgHþ ðYÞ ½AgHþ þ AgðIÞ#Ag2 þ þ Hþ ðZÞ Subsequently Shimizu and coworkers extensively probed the structure and nature of these identified Ag clusters by a combination of H2-TPR, XRD, UV-Vis, and Ag K-edge EXAFS studies.173 The H2-TPR results at Ag-MFI-58 revealed two peaks of H2 consumption, which combined with XRD studies indicated that all Agþ cations first reduce to small clusters, followed by further aggregation

Dynamic evolution of catalytic active sites within zeolite catalysis

183

Fig. 20 Time evolution of C3H8-SCR of NO in Ag-MFI-58 in both the presence and absence of H2. Reprinted from Shibata, J.; Takada, Y.; Shichi, A.; Satokawa, S.; Satsuma, A.; Hattori, T., Ag Cluster as Active Species for SCR of NO by Propane in the Presence of Hydrogen Over Ag-MFI. J. Catal. 2004, 222(2), 368–376, with permission from Elsevier. Table 2

Curve-fitting analysis of FT-EXAFS for a range of Ag samples.

Samples

Scatter atom

N

r (A˚)

Ag-MFI(22)-48 Ag-MFI(22)-58 Ag-MFI(13)-30 Ag-MFI(7.7)-33 Ag-MFI(22)-58a Ag foilb Ag2Ob

Ag Ag Ag O Ag Ag O

1.3 1.9 1.0 0.9 3.3 12 2

2.71 2.73 2.72 2.11 2.73 2.89 2.04

a

Measured after H2 treatment at 573 K. Parameters from crystal structures. 173

b

to larger particles. Curve-fitting analysis of FT-EXAFS spectra of a range of different Ag-MFI samples with varying Ag/Al ratios (Table 2), after C3H8-SCR in the presence of 0.5% H2 revealed that, the coordination number of Ag increases as a function of Ag/Al ratio, where in the Ag-MFI-58 sample after H2 treatment at 573 K a coordination number of 3.3 is fit with Ag-Ag distances of 2.73 Å, indicating that under these conditions the clusters are around 3–4 atoms in size. Based on the combined results from these different experimental studies, the authors proposed a most probable structure of Ag4þ clusters forming and serving as the active catalytic sites during lower-temperature and H2-rich conditions of C3H8-SCR of NO. Overall, this section outlines some selected and important examples of mobility of TM centers within zeolites, with a particular focus on operando experimental and computational studies elucidating Cu-mobility during standard SCR reactivity in chabazites. These highlighted examples demonstrate that the arrangement of such active sites during catalysis should not necessarily be assumed to be static, where one could anticipate that with further experimental & computational interrogations of TM-zeolite catalysts through state-of-the-art operando techniques, this phenomena could also be unraveled in other examples in the future.

6.08.3

Computational assessment of active site mobility in zeolites

6.08.3.1

Overview of enhanced sampling methods over static methods

A fundamental understanding of the functioning of active sites in zeolite catalysts at the molecular level is essential to improve catalytic processes or design new, highly selective catalysts. In this regard, the inherent characteristic dynamic behavior of zeolite materials at realistic working conditions introduces a major complexity. Modeling the dynamic nature of the active sites and their catalytic function at realistic operating conditions requires advanced operando spectroscopic and computational techniques as many zeolite catalyzed processes are typically occurring in a complex molecular environment with multiple guest species present

184

Dynamic evolution of catalytic active sites within zeolite catalysis

at higher temperatures and pressures. In recent years, the field of computational chemistry has progressively shifted from a standard static modeling approach towards these operando models. In order to properly characterize the complex dynamic environment of the zeolite catalyst, a range of methods encompassing molecular dynamics simulations, microkinetic models, machine learning algorithms, etc. are being explored.2,182,183 The standard quantum chemical modeling approach describes the 0 K potential energy surface (PES) of the reactive system by a discrete set of points, typically corresponding to the reactant, product and transition states. This static description of the working catalytic material can be a huge oversimplification as the structure of the active sites are presumed to remain unaltered by the reaction conditions. In reality, however, these assumptions become invalid as competitive pathways determining the product selectivity may be operational and various guest molecules may be present in the zeolite micropores, transforming the nature of the active sites. A second disadvantage of the static approach is that temperature effects are often ignored or underestimated. While at low temperature, intermediates and reactants are often well-defined stationary states on the potential energy surface, the intermediates and active sites behave dynamically at high temperature and their mobility is hence improperly accounted for.183,184 In zeolite catalysis, the presence of various guest species and the high reaction temperatures urge to make the transition from the PES to the free energy surface (FES) at realistic conditions. Due to the specific microporous nature of the catalyst, the FES for zeolite systems can be rather complex, exhibiting multiple local stationary states. Dynamic techniques rely on a sampling protocol to scan a larger part of the configurational space, while accounting for framework flexibility, finite temperature effects and anharmonic motions. Two general methodologies are distinguished, namely Monte Carlo (MC) simulations and Molecular Dynamics (MD) simulations, which will be the focus of this article. In MD simulations, the real-time evolution of the system on the FES is explored and an ensemble of different configurations with actual paths connecting these physical states is sampled. The temperature and pressure can be controlled by coupling a thermostat or barostat to the simulation engine in order to mimic the true experimental conditions. The energy surface may be constructed both with first-principle methods such as DFT as well as classical force-field methods, although for the description of bond formation and cleavage only a first-principle description is suitable. Since the interaction of guest molecules with the active sites often involves a rearrangement of chemical bonds and interactions, the application of DFT methods is inevitable. Currently, the attainable time scales for first-principle MD simulations of zeolite systemsdwhich are typically in the order of hundreds of psdare sufficient for the description of stable configurations and local minima, while activated processes such as chemical transformations can hardly be observed in the time span of the simulation. The situation can be compared to a topographical map of a mountain landscape, as shown in Fig. 21. During a regular MD simulation of a few hundred picoseconds, the sampling would be limited to the mountain valleys only. However, to describe the activated processes, one would also have to visit the mountain passes with sufficiently high probabilities. Various free energy methods have been developed to enhance the sampling in low probability regions of configurational space, thus allowing to reconstruct the free energy surface for activated transitions. Some techniques improve the sampling of all degrees of freedom such as Transition Path Sampling (TPS) which, due to the high computational cost, has only limitedly found its way into the field of heterogeneous catalysis.184–186 Alternatively, the sampling can also be enhanced along selected degrees of freedom only, the so-called collective variables (CV). The selection of these CVs is essential for a successful description of the activated transition, though identifying all degrees of freedom involved in the reaction is not always trivial.187,188 The premier criterion for a proper CV is its ability to uniquely distinguish between the initial state, the final state and all intermediate states while for computational efficiency, the dimensionality

high free energy

Y

low free energy X Fig. 21

Exploration of a hypothetical free energy surface by driving the sampling in various directions.

Dynamic evolution of catalytic active sites within zeolite catalysis

185

should be kept as low as possible. The CVs are described as a function of the microscopic coordinates of the system, employing often geometric parameters such as distances, angles, coordination numbers, etc. The same process can be described by multiple appropriate CVs, resulting in differently shaped free energy surfaces. A recent study on methylation reaction kinetics in zeolite ZSM-5 has shown that activation barriers estimated by employing a different set of CVs are in good agreement, provided the width of the reactant and transition state region of the free energy profiles is taken into account.189 Three enhanced sampling techniques which have been successfully demonstrated for zeolite-catalyzed reaction systems are Metadynamics (MTD), Umbrella Sampling (US) and Thermodynamic Integration (TI),37,189–192 schematically represented in Fig. 22. In principle, all techniques should yield equal free energy profiles for the same reaction process if the configuration space is sampled sufficiently well.189 Metadynamics is a non-equilibrium technique, developed by Laio and Parrinello,193–195 in which the sampling of low probability regions is facilitated by the introduction of a bias potential acting on the selected CVs. This potential is constructed on the fly by gradually spawning Gaussian shaped hills along the system trajectory in the CV space until the potential energy of the local minima artificially increases so that reaction barriers can be overcome. The free energy profile as a function of the predefined reaction coordinate can be reconstructed by taking the opposite of the bias potential function. While the MTD method is an excellent tool to scan the configurational space of complex systems and explore reaction mechanisms, the statistical error of the method is dependent on the choice of the Gaussian hill parameters which is less desirable.196 To speed up the convergence of the free energy profile in MTD simulations, the multiple walker MTD scheme has been proposed. Herein, multiple simulations are run in parallel which all simultaneously sample the same free energy surface though the communication between the simulations allows for a more efficient screening of different regions of the free energy surface.197 In the Umbrella Sampling technique, introduced by Torrie and Valleau,198 the CV range is divided into a set of windows and for each window a biased MD simulation is carried out. An external potential restricts the sampling of the configurational space to the individual window only, thus ensuring the sampling is equally well in all regions of the CV space. The free energy profile can be reconstructed by employing a post-processing algorithm such as the weighted histogram analysis method (WHAM).199 Also within Thermodynamic Integration, proposed by Kirkwood,200 a number of points along the CV range are selected and for each point a constrained MD simulation at fixed CV values is performed. The free energy profile is then obtained by integrating over the averaged free energy derivative in terms of the CVs. Both techniques are computationally very efficient thanks to their high parallelizability, although a proper knowledge on the reaction mechanism and its intermediates has to be known in advance.201 For more detailed information on free energy methods, the reader is referred to dedicated reviews.202–204 Through the application of MD techniques and free energy methods, several studies were able to characterize the important role of dynamic active sites in zeolite catalysis. On the one hand, Bronsted acid sites originally located on the lattice can be captured by protic molecules to form reactive clusters as demonstrated in the context of the MTO process33,92 or even by unsaturated hydrocarbons present in the pores of the material to form carbocation intermediates in the context of alkene cracking.36,190,205 In particular, the existence of alkyl carbocations has long been debated due to their unstable nature at low temperature. However, finite temperature effects have an important contribution to the stabilization of these intermediates and their occurrence in the zeolite environment at elevated reaction temperatures could be demonstrated by performing MD simulations in contrast to static DFT calculations.36 MD studies in the field of selective catalytic reduction of nitrogen oxides also evidenced that the structure of Lewis acid sites can dynamically evolve in the course of the reaction.20 Below, three examples demonstrating the beneficial use of enhanced sampling techniques for characterizing the mobility of Brønsted and Lewis acid sites are discussed in detail.

Fig. 22 Schematic representation of the regular MD and enhanced sampling techniques, metadynamics (MTD), umbrella sampling (US) and thermodynamic integration (TI) for a fictitious free energy profile F(q), bias potential U(q) and sampling probability p(q) as a function of collective variable q.

186

Dynamic evolution of catalytic active sites within zeolite catalysis

6.08.3.2 6.08.3.2.1

Case studies Proton mobility in zeolites

As explained in Section 6.08.2.1, the BAS in zeolites becomes completely solvated once few protic molecules are adsorbed in the framework. This represents a very interesting case study for enhanced sampling techniques. Indeed, with the addition of solvent molecules the number of available configurations becomes so large that static calculations are effective only if a very large number of structures is optimized.206 Here, we discuss two cases in which advanced sampling techniques have been effectively used to gain insight into zeolite-catalyzed reactions with mobilized protons, namely ethanol dehydration in H-b at high water loadings and the formation of surface methoxide species (SMS) at high methanol loadings, an initial reaction of the MTH process.82 Bates et al. have recently performed a thorough analysis of the ethanol dimerization to diethyl ether in the presence of water, catalyzed by the H-b zeolite.207 The impact of water in the zeolite pores on the reaction kinetics and mechanism is well established. For example the velocity of H/D exchange for isobutane208 and benzene209 in H-ZSM-5 with relatively low Si/Al ratio increases with a loading of  1 H2O/BAS, to then decrease again when the loading exceeds 1–3 H2O/BAS. While a few water molecules on the BAS likely make the proton transfer entropically easier, more water molecules in the proton solvation remarkably stabilize it, reducing its acidity. Also, Liu et al. studied the dehydration of cyclohexanol in water, comparing the homogenous H3PO4 and the H-bzeolite as catalysts, finding that the reaction proceeds remarkably faster in the latter.210 The authors invoked a favorable association between the cyclohexanol and the hydronium ion in the zeolite with respect to pure liquid water. When ethanol dehydration to diethyl ether is considered, a  1 order in the reaction kinetics with respect to water is found. Using a combination of in situ IR spectroscopy and ab initio molecular dynamics simulations, it was shown that in the zeolite pores a reactive cluster in the form of (C2H5OH)(H3Oþ)(H2O)4–5 is present, with more extended water networks around it whose size is dependent on the applied water pressure. The introduction of ethanol in the water cluster becomes less favorable as the cluster grows, because of the alkyl group disrupting the hydrogen bond network. Metadynamics was used to study the transition state of the reaction (Fig. 23). It was found that the ethanol molecules do not significantly disrupt the protonated water cluster and prefer to reside at its periphery. Similarly, the transition state as well resides at the periphery of the protonated water cluster. With the growth of the water cluster, the formation of the hydrophobic diethyl ether disrupts the network of hydrogen bonds significantly compared to the reacting ethanol molecules, therefore explaining the negative dependence of the reaction kinetics with respect to water pressure. A second case in which the presence of protic molecules absorbed in the catalyst pores is known to be of fundamental importance is the MTH process which gained a lot of interest in the contemporary chemical industry,3 as it is a possible link between sustainable raw materials and commodity chemicals. A complete overview of the intriguing MTH mechanism is outside the scope of this contribution and the interested reader is therefore referred to specialized reviews on the topic.13,82,211,212 Here, the focus will lie on the transformations occurring at the BAS in the initial stages of the process, when methanol starts to adsorb in the catalyst pores. Once methanol (or any small alcohol in general) is introduced into the framework of a zeotype material, the acidity of the BAS immediately catalyzes dehydration reactions. These reactions can produce dimethyl ether, if methanol reacts with itself, or surface methoxide species (SMS), if the methyl group is transferred to the framework.213,214 Interestingly, the higher basicity of methanol with respect to water implies that even less molecules are needed to solvate the BAS. This has been confirmed with AIMD since the late 1990s.215 For instance, in the aluminophosphate zeotype material H-SAPO-34 the BAS is solvated for more than 50% of the time with only three methanol molecules, while seven water molecules are needed to reach the same result.83

Fig. 23 (A) Reaction profile for the dehydration of ethanol to diethyl ether in H-b with 5 co-adsorbed water molecules, as obtained from metadynamics. Snapshots of the simulation are also shown. (B) The original 2-dimensional free energy surface from which the profile in (A) was derived. The minimum free energy path is shown in red. Adapted with permission from Bates, J.S.; Bukowski, B.C.; Greeley, J.; Gounder, R., Structure and Solvation of Confined Water and Water–Ethanol Clusters Within Microporous Brønsted Acids and Their Effects on Ethanol Dehydration Catalysis. Chem. Sci. 2020, 11(27), 7102–7122. Copyright 2020, The Royal Society of Chemistry.

Dynamic evolution of catalytic active sites within zeolite catalysis

187

When the formation of SMS in H-ZSM-5 was for the first time studied with advance sampling techniques,216 it was found that static and dynamic simulations depict a very different image. Indeed, when a second methanol molecule is introduced in the reaction environment, static calculations suggest that the free energy barrier to form SMS increases (provided that thermal corrections are taken into account217,218), since the optimization of the transition state leads to a ring-like structure whose large entropic penalty compensates the modest enthalpic stabilization. By using metadynamics, on the other hand, the barrier was shown to decrease as a consequence of an easier proton abstraction from the framework by the two protic molecules, subsequently facilitating the SMS formation from the protonated methanol molecule. Later on, De Wispelaere et al.32 investigated the role of water in the MTH process. The differences between extra water and extra methanol molecules present during the SMS formation were studied in H-SAPO-34 with a combination of enhanced sampling techniques and in situ microspectroscopy. Also in this case it was found that the formation of protonated clusters, either from water or methanol, helps to reduce the reaction barrier for SMS formation (Fig. 24). However, this effect is smaller for water, as it actually directly competes with methanol for the access to the BAS. This reflects experimentally in a longer induction period for MTH when water is cofeed with methanol. Based on experimental evidence suggesting that the formation of SMS can already occur at room temperature,219–222 Nastase et al.33 recently investigated such phenomena as a function of the acid site density and methanol loading in H-ZSM-5. By performing metadynamics, the authors demonstrated that not only larger methanol loadings can lower the methylation barriers, but also a closer proximity between Al defects, in agreement with the experimental observations. Nonetheless, the decrease in the barrier height was not as pronounced as in the experimental case, therefore the authors proposed that more Al locations or alternatively different type of active sites, such as framework-associated and extra-framework aluminum, should also be considered.

6.08.3.2.2

Ni-SSZ-24 for ethene oligomerization

Interested in the mechanism of ethene oligomerization in Ni-exchanged zeolites, Brogaard and coworkers employed advanced Umbrella Sampling simulations to probe the mechanism at operating conditions in Ni-SSZ-23 (25 bar and 120  C) and to calculate intrinsic reaction barriers and construct free energy profiles.23 The initial reactant state modeled was a [Ni(C2H4)(C2H5)]þ species anchored to the zeolite via two Ni – O bonds, based on a key state in the CosseeArlman mechanism identified in a previous study.223 DFT-MD simulations were carried out at the revPBE-D3 level, and the constructed free energy profile is outlined in Fig. 25, while Fig. 26 and Table 3 highlight the choice of collective variable employed for each process. Initial ethene coordination was characterized to take place with accompanying formation of a ß-agostic interaction with the ethyl ligand and cleavage of both NieO bonds from the zeolite framework, thus mobilizing the Ni(II) cation to afford [Ni(C2H5)(C2H4)2]þ. This was supported by histogram analysis of Ni-Al distances where an increase from  3 to 6 Å is identified in the reactant and product sampling regions, respectively (Fig. 27A, inset). The replacement of a relatively weak-field Ofw ligand with a stronger h2 ethene ligand also induces a tetrahedral/trigonal-planar to square-planar geometry change in [Ni(C2H5)(C2H4)2]þ, bringing the two coupling partners closer together. C-C coupling can then take place with an intrinsic activation barrier of 37 kJ mol 1, where, in reverse to the mobilization observed in the prior step, re-coordination of Ni(II) to the AlO2 site of the framework concomitantly takes place (again supported by shortening of the average Ni-Al interatomic distance). Following formation of the anchored [Ni(C4H9)(C2H4)]þ species, coordination of additional ethene can take place with an intrinsic reaction barrier of 37 kJ mol 1, followed by ß-hydrogen transfer (where DGz ¼ 20 kJ mol 1) to afford a mobilized [Ni(C4H8)(C2H5)(C2H4)]þ species. Subsequent ethene coordination and 1-butene desorption can take place to regenerate the mobilized [Ni(C2H4)2(C2H5)]þ species. Alternatively, from the [Ni(C4H9)(C2H4)]þ adduct, a process of ethene coordination and C-C coupling to afford the n-hexyl intermediate was also modeled, which was found to be kinetically disfavored. A microkinetic model constructed from the computed energetics identified that rate of 1-butene formation is approximately two orders of magnitude higher than 1-hexene formation, which was supported by continuous flow experiments where 98% selectivity towards butenes formation was observed over hexenes. This study neatly demonstrates the power of operando modeling techniques such as US to probe and reveal the tendency of active TM sites in zeolites to mobilize under catalytic conditions. In this case, Ni(II) reversibly is shown to mobilize in the form of four-coordinate cationic organonickel complexes interacting with the zeolite framework via longer-range coulombic interactions, resembling an organometallic nickel complex interacting with solvent molecules under homogenous catalytic conditions. Ethene dimerization can take place from this state with accompanying re-anchoring of the Ni site, and comparison of the computationallyderived microkinetic model was found to be consistent with experiment.

6.08.3.2.3

Mobility of active sites in H-SSZ13 during fast NH3-SCR-NOx

In 2017, Schneider and coworkers employed a combination of static DFT calculations, ab initio molecular dynamics, metadynamics and complementary experimental kinetics studies to unravel the role of the Brønsted acid site in fast NH3-SCR-NOx in H-SSZ-13.34 Three distinct mechanistic routes of N2O3 consumption, an intermediate postulated as a transient intermediate in fast SCR, were considered and modeled with this combined approach: 1. A “NH4þ” route, in which a NH4þ cation interacting with the Brønsted acid site reacts with N2O3 to afford H2NNO and HONO: ZNH4 þ N2 O3 /ZH þ H2 NNO þ HONO

188

Dynamic evolution of catalytic active sites within zeolite catalysis

Fig. 24 2D FES at 330  C for methoxide formation in pure methanol (1:0)mw,sim and (5:0)mw,sim (A, B) and the (1:4)mw,sim methanol-water mixture (C) occluded in H-SAPO-34, with indication of the least free energy path. The insets show snapshots of the TS region. (x:y)mw,sim stands for x MeOH and y H2O molecules per BAS. Adapted with permission from De Wispelaere, K.; Wondergem, C.S.; Ensing, B.; Hemelsoet, K.; Meijer, E.J.; Weckhuysen, B.M.; Van Speybroeck, V.; Ruiz-Martinez, J., Insight Into the Effect of Water on the Methanol-to-Olefins Conversion in H-SAPO-34 From Molecular Simulations and In Situ Microspectroscopy. ACS Catal. 2016, 6(3), 1991–2002. Copyright 2016, American Chemical Society.

Dynamic evolution of catalytic active sites within zeolite catalysis

189

Fig. 25 Free energy profile (in kJ/mol) of ethene oligomerization by Ni-SSZ-24 constructed from results of Umbrella Sampling calculations of the elementary steps at 25 bar and 120  C. Reprinted from Brogaard, R.Y.; Kømurcu, M.; Dyballa, M.M.; Botan, A.; Van Speybroeck, V.; Olsbye, U.; De Wispelaere, K., Ethene Dimerization on Zeolite-Hosted Ni Ions: Reversible Mobilization of the Active Site. ACS Catal. 2019, 9(6), 5645–5650, with permission from the American Chemical Society.

Fig. 26 Illustrative structures to outline the atoms used in the CNs used to sample each process, along with parameters. Si ¼ yellow, O ¼ red, Al ¼ purple, N ¼ blue, H ¼ white, Ni ¼ turquoise.

2. A “ZH” route, in which the protonated Brønsted Acid site reacts with N2O3 to form a short-lived “ZNO” intermediate, which is subsequently consumed by NH3 to regenerate “ZH”: ZH þ N2 O3 /ZNO þ HONO ZNO þ NH3 /ZH þ H2 NNO

190

Dynamic evolution of catalytic active sites within zeolite catalysis Table 3

CN parameters for the reaction steps illustrated in Fig. 26.

Reaction step

Coordination number

r0 [A˚]

n

m

Coordination of ethene Migratory insertion Hydrogen transfer A Hydrogen transfer B 1-butene desorption

CN (Ni, C in surrounding ethenes) CN (C in ethene, Ca in alkyl) CN (ß-H in butyl, Ni) CN (ß-H in butyl, C in ethene) CN(Ni, C in C]C of 1-butene)

2.1 1.9 1.5 1.8 2.7

6 6 6 6 6

8 12 10 12 10

Fig. 27 Free energy profile of ethene coordination from [Ni(C2H4)(C2H5)]þ complex (left) with Ni-Al distance histograms of the reactant and product states (left, inset), and from [Ni(C2H4)(C4H9)]þ complex (bottom right). MD snapshots (top right) of the anchored and mobilized Ni(II) states, with the time-elapsed position of the Ni atom (turquoise) and Ni-bound ethyl C atom (blue) in the MD trajectories. Reprinted from Brogaard, R.Y.; Kømurcu, M.; Dyballa, M.M.; Botan, A.; Van Speybroeck, V.; Olsbye, U.; De Wispelaere, K., Ethene Dimerization on Zeolite-Hosted Ni Ions: Reversible Mobilization of the Active Site. ACS Catal. 2019, 9(6), 5645–5650, with permission from the American Chemical Society.

3. A “physisorbed NH3” route, in which the tethered NH4þ active site is solvated with an equivalent of ammonia to form a “ZNH4.NH3” adduct, which in turn can react with N2O3: ZNH4 $NH3 þ N2 O3 /ZNH4 þ H2 NNO þ HONO

Metadynamics calculations were performed on a 12-T-site SSZ structure (Si/Al ¼ 11) via the CPMD method at the PBE level of theory. Initial experimental rate measurements revealed that the rate of NOx consumption varies non-linearly with respect to temperature (Fig. 28A), where two distinct activation energies were identified depending on the temperature range. An activation energy of  30 kJ mol 1 could be observed in the temperature range of 613–653 K while EA ¼ 21 kJ mol 1 was identified for 498–533 K. AIMD simulations combined with a Potential Mean Force (PMF) approach were employed to assess the free energy of adsorption of NH3. Free energies could be obtained by integrating the obtained constraint force along a predefined one-dimensional grid of separation distance between the adsorbate (NH3) and active site (either ZH or ZNH4þ). The coverages of NH3 across the temperature range employed experimentally revealed that almost all Brønsted Acid sites take the form of the NH4þ site, and at higher temperatures a physisorbed NH3 site of ZNH4.NH3 is unlikely to form. The reactivity of the proposed active sites with N2O3 were then investigated computationally via metadynamics simulations, with two CVs sampled in each case (outlined in Fig. 29 and Table 4).

Dynamic evolution of catalytic active sites within zeolite catalysis

191

Fig. 28 (A) Rate of NOx consumption under fast SCR conditions per gram of H-SSZ-13, versus temperature. (B) Corresponding Arrhenius plots, detailing two separate kinetic schemes at different temperature ranges. Reprinted from Li, S.; Zheng, Y.; Gao, F.; Szanyi, J.; Schneider, F.; Experimental and Computational Interrogation of Fast SCR Mechanism and Active Sites on H-Form SSZ-13. ACS Catal. 2017, 7(8), 5087–5096, with permission from the American Chemical Society.

Fig. 29 Schematic structures of starting points and collective variables used in the metadynamics simulations of the (A) NH4þ route, (B) ZH route and (C) physisorbed NH3 route. Si ¼ yellow, O ¼ red, Al ¼ purple, N ¼ blue, H ¼ white. Table 4 Route

Collective Variables for the reaction steps illustrated in Fig. 29. CV1

CV2

Ea (kJ mol 1)

(a) NH4þ CN[N1-N2] SUM(CN[N2-H1,2,3,4]) 80 (b) ZH CN[N1-H] TOTAL(CN[N1,2-O1,2,3,4]) 100 (c) Physisorbed NH3 SUM(CN[N1-N2,3]) TOTAL(CN[N2,3-Hall]) 60

The metadynamics simulations revealed that the physisorbed NH3 route (Fig. 30C) proceeds via an activation free energy of 80 kJ mol 1, where NO addition to NH3 takes place concomitantly with hydrogen transfer from NH3 to NO2 to form HONO and H2NNO. In the NH4þ route (Fig. 30A), NH3 partially desorbs from the Al site before reacting in a similar manner with N2O3 as the physisorbed route, where EA ¼ 100 kJ mol 1. Finally, the ZH route (Fig. 30B) proceeds with proton transfer from the Brønsted acid site to NO2 to form a ZNO site takes place, with an activation energy of 60 kJ mol 1. The authors subsequently constructed a kinetic model based on the computed barriers of the three routes via metadynamics simulations, in which individual rates of the three routes were calculated as a function of temperature along with a total rate (Fig. 31). This revealed that at temperatures below 570 K, the physisorbed NH3 reaction route dominates and contributes the most to the total rate of fast SCR in the catalytic system, and up to 570 K temperature the rate is invariant with respect to temperature. Beyond 570 K the NH4þ route dominates, and the initial identification of two separate temperature-dependent kinetic regimes was uncovered. These studies therefore demonstrate the power and application of enhanced sampling techniques in the accurate operando modeling of mobilized catalytically active sites when used in concert with experiment. While the identification of two separate

192

Dynamic evolution of catalytic active sites within zeolite catalysis

(A) 3.5

–10

–30

3.0 CV2

–40 –50 2.5 –60

Free energy (kJ mol–1)

–20

–70 2.0

0

0.2

0.4

0.6 CV1

0.8

1.0

3.5

–20

3.0

–40

–30

CV2

–50 –60

2.5

–70 –80

2.0

Free energy (kJ mol–1)

–10

(B)

–90 –100

1.5 0

0.2

0.4 0.6 CV1

0.8

1.0

(C)

–20 0.6

CV2

–30 0.4 –40 0.2 –50

Free energy (kJ mol–1)

–10

3.8

0 –60 0

0.2

0.4

0.6 CV1

0.8

1.0

Fig. 30 2D free energy profiles for the (A) Physisorbed NH3 route, (B) NH4þ route and (C) ZH route. Reprinted from Li, S.; Zheng, Y.; Gao, F.; Szanyi, J.; Schneider, F.; Experimental and Computational Interrogation of Fast SCR Mechanism and Active Sites on H-Form SSZ-13. ACS Catal. 2017, 7(8), 5087–5096, with permission from the American Chemical Society.

temperature dependent reaction-schemes in fast NH3-SCR-NOx reactivity in H-SSZ-13 was identified by experiment, complementary metadynamics simulations could pinpoint the mechanistic origins of this temperature-dependent behavior.

6.08.3.2.4

Mobility of multinuclear Cu sites in chabazites for the selective catalytic reduction (SCR) of nitrogen oxides

As previously outlined (Section 6.08.2.3.1.2), Gounder and coworkers employed a combination of XAS, static and dynamic DFT calculations to probe the speciation and role of mobile cationic copper sites in low-temperature NH3-SCR.20 The mobility of the [Cu(NH3)2]þ complexes were assessed using metadynamics simulations via CPMD. The Cu-Al CN was chosen as the CV to describe

Dynamic evolution of catalytic active sites within zeolite catalysis

193

Fig. 31 (A) Computed total rate (black), and rates of the three modeled mechanistic routes as a function of simulation temperature. (B) Arrhenius plot of total rate. Reprinted from Li, S.; Zheng, Y.; Gao, F.; Szanyi, J.; Schneider, F.; Experimental and Computational Interrogation of Fast SCR Mechanism and Active Sites on H-Form SSZ-13. ACS Catal. 2017, 7(8), 5087–5096, with permission from the American Chemical Society.

Fig. 32 Illustration of the diffusion of [Cu(NH3)2]þ species into a neighboring chabazite cage, with Cu-Al interatomic distances (r, in Å) and free energies (F, in kJ/mol) of the reactant state (I), saddle point (II) and product (III), calculated via metadynamics simulations by Schneider and Gounder.20

the diffusion of [Cu(NH3)2]þ to the adjacent CHA cage. The longest Cu-Al distance in the pre-equilibration MD simulation was chosen for the CN parameter d0. The metadynamics simulations, and the corresponding free energy profile revealed the [Cu(NH3)2]þ equilibrates in the aluminated cage with a Cu-Al distance of 4.7 Å (state I, Fig. 32), upon which the free energy increases as the Cu species approaches the window to the adjacent cage, hitting a peak at a Cu-Al distance of 8 Å and an activation barrier of 55 kJ mol 1 (state II, Fig. 32). Finally, as [Cu(NH3)2]þ diffuses into the neighboring cage, a local minimum (state III, Fig. 32) is observed at a Cu-Al distance of 8.5 Å. This activation barrier of 55 kJ mol 1 was used to estimate a rate of molecular diffusion of 6  106 s 1, and the fraction of mobilized [Cu(NH3)2]þ to “anchored” Cu at the charge-compensated cage (i.e., the population of state III relative to state I at equilibrium) was estimated to be 1.4  10 3 at operating conditions (473 K). Moreover, the electrostatic contribution to the chargeseparation of [Cu(NH3)2]þ and the Al site was estimated using Coulomb’s law, where Cu and Al were assumed to be positive and negative point-charges, respectively. This indicated that the electrostatic interactions between Cu and Al is a dominating contributor to the energetics of [Cu(NH3)2]þ diffusion, until around 7 Å, where steric hindrance likely proceeds to play a larger contributing role in diffusing through the 8-ring window.

194

6.08.4

Dynamic evolution of catalytic active sites within zeolite catalysis

Conclusions and perspectives

Within this article, the current knowledge on the nature and evolution of active sites within zeolite catalysis has been reviewed. Recently a consensus has been reached that active sites may become very dynamic under operating conditions, where the latter may refer to altered temperatures, partial pressures, etc. In this context, the state of understanding of proton mobility, mobility of Al sites via framework decomposition, and mobility of TM-exchanged sites within zeolites has been surveyed, highlighting in particular the use of operando spectroscopic and computational methods to identify these phenomena. As many catalytic processes in nanoporous materials occur at high temperatures and are influenced by the guest adsorption in the pores, operando modeling of the catalytic function is of critical importance to accurately describe the reaction systems. The more “standard” static modeling approach starts from a few points on the potential energy surface, such as reactants, transition states and products. This is a huge oversimplification of the working catalytic material, and in reality the scene is much more complex. Competitive pathways may be operational which are essential for determining product selectivities, various guest molecules may be present in the pores of the material which may alter the nature of the active site or facilitate certain reaction paths, and the operating temperature window may change the nature of the reactive intermediates.37,40,83 The field of computational modeling and spectroscopic characterization has evolved substantially, enabling more precise information on the nature of the active site and its mobility during operation. Using enhanced sampling MD techniques based on a first-principle description of the PES, one can map the free energy surface at realistic conditions. As such, a dynamic reorganization of catalytically active sites could be better characterized for various processes, as outlined in selected case studies. For Brønsted acidic sites, protic molecules in the zeolite pores were able to capture the proton originally located on the lattice to form protonated reactive clusters.32,204 For Ni-zeolites used in ethene oligomerization it was shown that ethene molecules reversibly mobilize the active site and exchange with the zeolite as ligands during reaction.23 Such reactant-mobilized active sites have also been observed in the selective catalytic reduction of nitrogen oxides with ammonia in H-SSZ-13,34 and Cu-SSZ-13 zeolites in standard SCR and some other cases.20 In the quest for conversion of new feedstocks such as biomass, new heterogeneous catalysts will have to be designed, containing more complex active sites, combining for example Brønsted and Lewis acid functions in close proximity.224 Ideally one could atomically design so-called single active sites at the surface of a solid catalyst, which gives the desired function and is robust in the desired operation window.225–227 To achieve this goal, the role of enhanced sampling techniques, in concert with accurate operando spectroscopic techniques, is expected to be crucial and increasingly prevalent in heterogeneous catalysis. Insight into the characterization of dynamic and mobile catalytic sites will both aid in the rationalization of the observed reactivities and selectivities, and will in turn inform further predictions of refined catalyst systems.

References 1. Vogt, E. T. C.; Weckhuysen, B. M. Fluid Catalytic Cracking: Recent Developments on the Grand Old Lady of Zeolite Catalysis. Chem. Soc. Rev. 2015, 44 (20), 7342–7370. 2. Van Speybroeck, V.; Hemelsoet, K.; Joos, L.; Waroquier, M.; Bell, R. G.; Catlow, C. R. A. Advances in Theory and their Application Within the Field of Zeolite Chemistry. Chem. Soc. Rev. 2015, 44, 7044–7111. 3. Tian, P.; Wei, Y.; Ye, M.; Liu, Z. Methanol to Olefins (MTO): From Fundamentals to Commercialization. ACS Catal. 2015, 5 (3), 1922–1938. 4. Jacobs, P. A.; Dusselier, M.; Sels, B. F. Will Zeolite-Based Catalysis Be as Relevant in Future Biorefineries as in Crude Oil Refineries? Angew. Chem. Int. Ed. 2014, 53 (33), 8621–8626. 5. Dusselier, M.; Davis, M. E. Small-Pore Zeolites: Synthesis and Catalysis. Chem. Rev. 2018, 118 (11), 5265–5329. 6. Ravi, M.; Sushkevich, V. L.; van Bokhoven, J. A. Towards a Better Understanding of Lewis Acidic Aluminium in Zeolites. Nat. Mater. 2020, 19 (10), 1047–1056. 7. Zhang, Y.; Zhao, R.; Sanchez-Sanchez, M.; Haller, G. L.; Hu, J.; Bermejo-Deval, R.; Liu, Y.; Lercher, J. A. Promotion of Protolytic Pentane Conversion on H-MFI Zeolite by Proximity of Extra-Framework Aluminum Oxide and Brønsted Acid Sites. J. Catal. 2019, 370, 424–433. 8. Gounder, R.; Jones, A. J.; Carr, R. T.; Iglesia, E. Solvation and Acid Strength Effects on Catalysis by Faujasite Zeolites. J. Catal. 2012, 286, 214–223. 9. Almutairi, S. M. T.; Mezari, B.; Filonenko, G. A.; Magusin, P. C. M. M.; Rigutto, M. S.; Pidko, E. A.; Hensen, E. J. M. Influence of Extraframework Aluminum on the Brønsted Acidity and Catalytic Reactivity of Faujasite Zeolite. ChemCatChem 2013, 5 (2), 452–466. 10. Shamzhy, M.; Opanasenko, M.; Concepción, P.; Martínez, A. New Trends in Tailoring Active Sites in Zeolite-Based Catalysts. Chem. Soc. Rev. 2019, 48 (4), 1095–1149. 11. Li, G.; Pidko, E. A. The Nature and Catalytic Function of Cation Sites in Zeolites: A Computational Perspective. ChemCatChem 2019, 11 (1), 134–156. 12. Hereijgers, B. P. C.; Bleken, F.; Nilsen, M. H.; Svelle, S.; Lillerud, K.-P.; Bjørgen, M.; Weckhuysen, B. M.; Olsbye, U. Product Shape Selectivity Dominates the Methanol-toOlefins (MTO) Reaction Over H-SAPO-34 Catalysts. J. Catal. 2009, 264 (1), 77–87. 13. Olsbye, U.; Svelle, S.; Bjørgen, M.; Beato, P.; Janssens, T. V. W.; Joensen, F.; Bordiga, S.; Lillerud, K. P. Conversion of Methanol to Hydrocarbons: How Zeolite Cavity and Pore Size Controls Product Selectivity. Angew. Chem. Int. Ed. 2012, 51 (24), 5810–5831. 14. Cnudde, P.; Demuynck, R.; Vandenbrande, S.; Waroquier, M.; Sastre, G.; Speybroeck, V. V. Light Olefin Diffusion during the MTO Process on H-SAPO-34: A Complex Interplay of Molecular Factors. J. Am. Chem. Soc. 2020, 142 (13), 6007–6017. 15. Cnudde, P.; Redekop, E. A.; Dai, W.; Porcaro, N. G.; Waroquier, M.; Bordiga, S.; Hunger, M.; Li, L.; Olsbye, U.; Van Speybroeck, V. Experimental and Theoretical Evidence for Promotional Effect of Acid Sites on the Diffusion of Alkenes Through Small-Pore Zeolites. Angew. Chem. Int. Ed. 2021, 60 (18), 10016–10022. 16. Wang, S.; He, Y.; Jiao, W.; Wang, J.; Fan, W. Recent Experimental and Theoretical Studies on Al Siting/Acid Site Distribution in Zeolite Framework. Curr. Opin. Chem. Eng. 2019, 23, 146–154. 17. Yokoi, T.; Mochizuki, H.; Namba, S.; Kondo, J. N.; Tatsumi, T. Control of the Al Distribution in the Framework of ZSM-5 Zeolite and Its Evaluation by Solid-State NMR Technique and Catalytic Properties. J. Phys. Chem. C 2015, 119 (27), 15303–15315. 18. Nishitoba, T.; Yoshida, N.; Kondo, J. N.; Yokoi, T. Control of Al Distribution in the CHA-Type Aluminosilicate Zeolites and its Impact on the Hydrothermal Stability and Catalytic Properties. Ind. Eng. Chem. Res. 2018, 57 (11), 3914–3922. 19. Di Iorio, J. R.; Gounder, R. Controlling the Isolation and Pairing of Aluminum in Chabazite Zeolites Using Mixtures of Organic and Inorganic Structure-Directing Agents. Chem. Mater. 2016, 28 (7), 2236–2247.

Dynamic evolution of catalytic active sites within zeolite catalysis

195

20. Paolucci, C.; Khurana, I.; Parekh, A. A.; Li, S.; Shih, A. J.; Li, H.; Iorio, J. R. D.; Albarracin-Caballero, J. D.; Yezerets, A.; Miller, J. T.; Delgass, W. N.; Ribeiro, F. H.; Schneider, W. F.; Gounder, R. Dynamic Multinuclear Sites Formed by Mobilized Copper Ions in NOx Selective Catalytic Reduction. Science 2017, 357 (6354), 898–903. 21. Moors, S. L. C.; De Wispelaere, K.; Van der Mynsbrugge, J.; Waroquier, M.; Van Speybroeck, V. Molecular Dynamics Kinetic Study on the Zeolite-Catalyzed Benzene Methylation in ZSM-5. ACS Catal. 2013, 3 (11), 2556–2567. 22. Serna, P.; Gates, B. C. Zeolite-Supported Rhodium Complexes and Clusters: Switching Catalytic Selectivity by Controlling Structures of Essentially Molecular Species. J. Am. Chem. Soc. 2011, 133 (13), 4714–4717. 23. Brogaard, R. Y.; Kømurcu, M.; Dyballa, M. M.; Botan, A.; Van Speybroeck, V.; Olsbye, U.; De Wispelaere, K. Ethene Dimerization on Zeolite-Hosted Ni Ions: Reversible Mobilization of the Active Site. ACS Catal. 2019, 9 (6), 5645–5650. 24. Schoonheydt, R. A.; Geerlings, P.; Pidko, E. A.; van Santen, R. A. The framework basicity of zeolites. J. Mater. Chem. 2012, 22 (36), 18705–18717. 25. Losch, P.; Joshi, H.; Stegmann, N.; Vozniuk, O.; Schmidt, W. Studying Proton Mobility in Zeolites by Varying Temperature Infrared Spectroscopy. Molecules 2019, 24 (17), 3199. 26. Losch, P.; Joshi, H. R.; Vozniuk, O.; Grünert, A.; Ochoa-Hernández, C.; Jabraoui, H.; Badawi, M.; Schmidt, W. Proton Mobility, Intrinsic Acid Strength, and Acid Site Location in Zeolites Revealed by Varying Temperature Infrared Spectroscopy and Density Functional Theory Studies. J. Am. Chem. Soc. 2018, 140 (50), 17790–17799. 27. Osuga, R.; Yokoi, T.; Doitomi, K.; Hirao, H.; Kondo, J. N. Infrared Investigation of Dynamic Behavior of Brønsted Acid Sites on Zeolites at High Temperatures. J. Phys. Chem. C 2017, 121 (45), 25411–25420. 28. Smith, L.; Cheetham, A. K.; Morris, R. E.; Marchese, L.; Thomas, J. M.; Wright, P. A.; Chen, J. On the Nature of Water Bound to a Solid Acid Catalyst. Science 1996, 271 (5250), 799–802. 29. Termath, V.; Haase, F.; Sauer, J.; Hutter, J.; Parrinello, M. Understanding the Nature of Water Bound to Solid Acid Surfaces. Ab Initio Simulation on HSAPO-34. J. Am. Chem. Soc. 1998, 120 (33), 8512–8516. 30. Vjunov, A.; Wang, M.; Govind, N.; Huthwelker, T.; Shi, H.; Mei, D.; Fulton, J. L.; Lercher, J. A. Tracking the Chemical Transformations at the Brønsted Acid Site Upon WaterInduced Deprotonation in a Zeolite Pore. Chem. Mater. 2017, 29 (21), 9030–9042. 31. Zheng, A.; Han, B.; Li, B.; Liu, S.-B.; Deng, F. Enhancement of Brønsted Acidity in Zeolitic Catalysts Due to an Intermolecular Solvent Effect in Confined Micropores. Chem. Commun. 2012, 48 (55), 6936–6938. 32. De Wispelaere, K.; Wondergem, C. S.; Ensing, B.; Hemelsoet, K.; Meijer, E. J.; Weckhuysen, B. M.; Van Speybroeck, V.; Ruiz-Martıínez, J. Insight into the Effect of Water on the Methanol-to-Olefins Conversion in H-SAPO-34 from Molecular Simulations and in Situ Microspectroscopy. ACS Catal. 2016, 6 (3), 1991–2002. 33. Nastase, S. A. F.; Cnudde, P.; Vanduyfhuys, L.; De Wispelaere, K.; Van Speybroeck, V.; Catlow, C. R. A.; Logsdail, A. J. Mechanistic Insight Into the Framework Methylation of H-ZSM-5 for Varying Methanol Loadings and Si/Al Ratios Using First-Principles Molecular Dynamics Simulations. ACS Catal. 2020, 10 (15), 8904–8915. 34. Li, S.; Zheng, Y.; Gao, F.; Szanyi, J.; Schneider, W. F. Experimental and Computational Interrogation of Fast SCR Mechanism and Active Sites on H-Form SSZ-13. ACS Catal. 2017, 7 (8), 5087–5096. 35. Ristanovic, Z.; Chowdhury, A. D.; Brogaard, R. Y.; Houben, K.; Baldus, M.; Hofkens, J.; Roeffaers, M. B. J.; Weckhuysen, B. M. Reversible and Site-Dependent ProtonTransfer in Zeolites Uncovered at the Single-Molecule Level. J. Am. Chem. Soc. 2018, 140 (43), 14195–14205. 36. Cnudde, P.; De Wispelaere, K.; Van der Mynsbrugge, J.; Waroquier, M.; Van Speybroeck, V. Effect of Temperature and Branching on the Nature and Stability of Alkene Cracking Intermediates in H-ZSM-5. J. Catal. 2017, 345, 53–69. 37. Cnudde, P.; De Wispelaere, K.; Vanduyfhuys, L.; Demuynck, R.; Van der Mynsbrugge, J.; Waroquier, M.; Van Speybroeck, V. How Chain Length and Branching Influence the Alkene Cracking Reactivity on H-ZSM-5. ACS Catal. 2018, 8 (10), 9579–9595. 38. Kosinov, N.; Liu, C.; Hensen, E. J. M.; Pidko, E. A. Engineering of Transition Metal Catalysts Confined in Zeolites. Chem. Mater. 2018, 30 (10), 3177–3198. 39. Bailleul, S.; Yarulina, I.; Hoffman, A. E. J.; Dokania, A.; Abou-Hamad, E.; Chowdhury, A. D.; Pieters, G.; Hajek, J.; De Wispelaere, K.; Waroquier, M.; Gascon, J.; Van Speybroeck, V. A Supramolecular View on the Cooperative Role of Brønsted and Lewis Acid Sites in Zeolites for Methanol Conversion. J. Am. Chem. Soc. 2019, 141 (37), 14823–14842. 40. Yarulina, I.; Wispelaere, K. D.; Bailleul, S.; Goetze, J.; Radersma, M.; Abou-Hamad, E.; Vollmer, I.; Goesten, M.; Mezari, B.; Hensen, E. J. M.; Martínez-Espín, J. S.; Morten, M.; Mitchell, S.; Perez-Ramirez, J.; Olsbye, U.; Weckhuysen, B. M.; Speybroeck, V. V.; Kapteijn, F.; Gascon, J. Structure–Performance Descriptors and the Role of Lewis Acidity in the Methanol-to-Propylene Process. Nat. Chem. 2018, 10 (8), 804. 41. Dapsens, P. Y.; Mondelli, C.; Pérez-Ramírez, J. Design of Lewis-Acid Centres in Zeolitic Matrices for the Conversion of Renewables. Chem. Soc. Rev. 2015, 44 (20), 7025–7043. 42. Paolucci, C.; Di Iorio, J. R.; Schneider, W. F.; Gounder, R. Solvation and Mobilization of Copper Active Sites in Zeolites by Ammonia: Consequences for the Catalytic Reduction of Nitrogen Oxides. Acc. Chem. Res. 2020, 53 (9), 1881–1892. 43. Vercammen, J.; Bocus, M.; Neale, S.; Bugaev, A.; Tomkins, P.; Hajek, J.; Van Minnebruggen, S.; Soldatov, A.; Krajnc, A.; Mali, G.; Van Speybroeck, V.; De Vos, D. E. ShapeSelective C–H Activation of Aromatics to Biarylic Compounds Using Molecular Palladium in Zeolites. Nat. Catal. 2020, 3 (12), 1002–1009. 44. Buurmans, I. L. C.; Weckhuysen, B. M. Heterogeneities of Individual Catalyst Particles in Space and Time as Monitored by Spectroscopy. Nat. Chem. 2012, 4 (11), 873–886. 45. Weckhuysen, B. M. Chemical Imaging of Spatial Heterogeneities in Catalytic Solids at Different Length and Time Scales. Angew. Chem. Int. Ed. 2009, 48 (27), 4910–4943. 46. Chowdhury, A. D.; Houben, K.; Whiting, G. T.; Chung, S.-H.; Baldus, M.; Weckhuysen, B. M. Electrophilic Aromatic Substitution over Zeolites Generates Wheland-Type Reaction Intermediates. Nat. Catal. 2018, 1 (1), 23–31. 47. Hunger, M. Brønsted Acid Sites in Zeolites Characterized by Multinuclear Solid-State NMR Spectroscopy. Catal. Rev. 1997, 39 (4), 345–393. 48. Freude, D.; Oehme, W.; Schmiede, H.; Staudte, B. NMR Investigation of Proton Mobility in Zeolites. J. Catal. 1974, 32 (1), 137–143. 49. Baba, T.; Inoue, Y.; Shoji, H.; Uematsu, T.; Ono, Y. Temperature-Dependent Lineshape of H-1 Magic-Angle-Spinning Nuclear-Magnetic-Resonance Spectra of Acidic HydroxylGroups in Zeolites. Microporous Mater. 1995, 3 (6), 647–655. 50. Sarv, P.; Tuherm, T.; Lippmaa, E.; Keskinen, K.; Root, A. Mobility of the Acidic Proton in Bronsted Sites of H-Y, H-Mordenite, and H-Zsm-5 Zeolites, Studied by HighTemperature H-1 Mas NMR. J. Phys. Chem. 1995, 99 (38), 13763–13768. 51. Baba, T.; Komatsu, N.; Ono, Y.; Sugisawa, H. Mobility of the Acidic Protons in H-ZSM-5 as Studied by Variable Temperature H-1 MAS NMR. J. Phys. Chem. B 1998, 102 (5), 804–808. 52. Baba, T.; Komatsu, N.; Ono, Y.; Sugisawa, H.; Takahashi, T. Nature of the Acidic Protons in H-Mordenite and H-MCM-22 as Studied by Variable Temperature H-1 MAS NMR1. Microporous Mesoporous Mater. 1998, 22 (1–3), 203–210. 53. Baba, T.; Ono, Y. Dynamic Properties of Protons in Solid Acids as Studied by Variable Temperature 1H MAS NMR. Appl. Catal. Gen. 1999, 181 (2), 227–238. 54. Kanellopoulos, J.; Gottert, C.; Schneider, D.; Knorr, B.; Prager, D.; Ernst, H.; Freude, D. NMR Investigation of Proton Mobility in Zeolites. J. Catal. 2008, 255 (1), 68–78. 55. Asakawa, N.; Motokura, K.; Yashima, T.; Koyama, T.; O-nuki, T.; Miyaji, A.; Baba, T. Proton Exchange Reaction Between Hydroxyl Groups in the Supercage and Those in the Sodalitecage of Y Zeolite As Studied by Variable Temperature H-1 MAS NMR. J. Phys. Chem. C 2012, 116 (33), 17734–17738. 56. Osuga, R.; Yokoi, T.; Doitomi, K.; Hirao, H.; Kondo, J. N. Infrared Investigation of Dynamic Behavior of Bronsted Acid Sites on Zeolites at High Temperatures. J. Phys. Chem. C 2017, 121 (45), 25411–25420. 57. Franke, M.; Simon, U. Characteristics of Proton Hopping in Zeolite H-ZSM5. Phys. Status Solidi B 2000, 218 (1), 287–290. 58. Franke, M.; Simon, U. Proton Mobility in H-ZSM5 Studied by Impedance Spectroscopy. Solid State Ion. 1999, 118 (3–4), 311–316. 59. Ryder, J. A.; Chakraborty, A. K.; Bell, A. T. Density Functional Theory Study of Proton Mobility in Zeolites: Proton Migration and Hydrogen Exchange in ZSM-5. J. Phys. Chem. B 2000, 104 (30), 6998–7011.

196

Dynamic evolution of catalytic active sites within zeolite catalysis

60. Sierka, M.; Sauer, J. Proton Mobility in Chabazite, Faujasite, and ZSM-5 Zeolite Catalysts, Comparison Based on Ab Initio Calculations. J. Phys. Chem. B 2001, 105 (8), 1603–1613. 61. Sierka, M.; Sauer, J. Finding Transition Structures in Extended Systems: A Strategy Based on a Combined Quantum Mechanics-Empirical Valence Bond Approach. J. Chem. Phys. 2000, 112 (16), 6983–6996. 62. Huo, H.; Peng, L.; Grey, C. P. Low Temperature 1H MAS NMR Spectroscopy Studies of Proton Motion in Zeolite HZSM-5. J. Phys. Chem. C 2009, 113 (19), 8211–8219. 63. Lee, B. J.; Kondo, J. N.; Wakabayashi, F.; Domen, K. Infrared Spectroscopic Study of High Temperature Behavior of the Bronsted Acidic Hydroxyl Groups on Zeolites. Bull. Chem. Soc. Jpn. 1998, 71 (9), 2149–2152. 64. Franke, M. E.; Sierka, M.; Simon, U.; Sauer, J. Translational Proton Motion in Zeolite H-ZSM-5. Energy Barriers and Jump Rates from DFT Calculations. Phys. Chem. Chem. Phys. 2002, 4 (20), 5207–5216. 65. Wang, Y.; Zhou, D. H.; Yang, G.; Liu, X. C.; Ma, D.; Liang, D. B.; Bao, X. H. Density Functional Theory Study of Proton Hopping in MCM-22 Zeolite. Chem. Phys. Lett. 2004, 388 (4–6), 363–366. 66. Fermann, J. T.; Blanco, C.; Auerbach, S. Modeling Proton Mobility in Acidic Zeolite Clusters. I. Convergence of Transition State Parameters From Quantum Chemistry. J. Chem. Phys. 2000, 112 (15), 6779–6786. 67. Tuma, C.; Sauer, J. A Hybrid MP2/Planewave-DFT Scheme for Large Chemical Systems: Proton Jumps in Zeolites. Chem. Phys. Lett. 2004, 387 (4–6), 388–394. 68. Beck, L. W.; White, J. L.; Haw, J. F. 1H {27Al} Double-Resonance Experiments in Solids: An Unexpected Observation in the 1H MAS Spectrum of Zeolite HZSM-5. J. Am. Chem. Soc. 1994, 116 (21), 9657–9661. 69. Brunner, E.; Beck, K.; Koch, M.; Heeribout, L.; Karge, H. Verification and Quantitative Determination of a New Type of Brønsted Acid Sites in H-ZSM-5 by 1H Magic-Angle Spinning Nuclear Magnetic Resonance Spectroscopy. Microporous Mater. 1995, 3 (4–5), 395–399. 70. Freude, D. Enhanced Resolution in the 1H NMR Spectra of Zeolite H-ZSM-5 by Heteronuclear Dipolar-Dephasing Spin-Echo MAS. Chem. Phys. Lett. 1995, 235 (1–2), 69–75. 71. Hunger, M. Multinuclear Solid-State NMR Studies of Acidic and Non-Acidic Hydroxyl Protons in Zeolites. Solid State Nucl. Magn. Reson. 1996, 6 (1), 1–29. 72. Peng, L.; Huo, H.; Liu, Y.; Grey, C. P. 17O Magic Angle Spinning NMR Studies of Brønsted Acid Sites in Zeolites HY and HZSM-5. J. Am. Chem. Soc. 2007, 129 (2), 335–346. 73. Abdolrahmani, M.; Chen, K.; White, J. L. Assessment, Control, and Impact of Brønsted Acid Site Heterogeneity in Zeolite HZSM-5. J. Phys. Chem. C 2018, 122 (27), 15520– 15528. 74. Schroeder, C.; Siozios, V.; Mück-Lichtenfeld, C.; Hunger, M.; Hansen, M. R.; Koller, H. Hydrogen Bond Formation of Brønsted Acid Sites in Zeolites. Chem. Mater. 2020, 32 (4), 1564–1574. 75. Benco, L.; Demuth, T.; Hafner, J.; Hutschka, F. Spontaneous Proton Transfer between O-Sites in Zeolites. Chem. Phys. Lett. 2000, 324 (5–6), 373–380. 76. Krossner, M.; Sauer, J. Interaction of Water With Brønsted Acidic Sites of Zeolite Catalysts. Ab Initio Study of 1: 1 and 2: 1 Surface Complexes. J. Phys. Chem. 1996, 100 (15), 6199–6211. 77. Jobic, H.; Tuel, A.; Krossner, M.; Sauer, J. Water in Interaction With Acid Sites in H-ZSM-5 Zeolite Does Not Form Hydroxonium Ions. A Comparison between Neutron Scattering Results and Ab Initio Calculations. J. Phys. Chem. 1996, 100 (50), 19545–19550. 78. Beck, L. W.; Xu, T.; Nicholas, J. B.; Haw, J. F. Kinetic NMR and Density Functional Study of Benzene H/D Exchange in Zeolites, the Most Simple Aromatic Substitution. J. Am. Chem. Soc. 1995, 117 (46), 11594–11595. 79. Wang, C.; Li, S.; Mao, X.; Caratzoulas, S.; Gorte, R. J. HD Exchange of Simple Aromatics as a Measure of Brønsted-Acid Site Strengths in Solids. Catal. Lett. 2018, 148 (11), 3548–3556. 80. Rybicki, M.; Sauer, J. Ab Initio Prediction of Proton Exchange Barriers for Alkanes at Brønsted Sites of Zeolite H-MFI. J. Am. Chem. Soc. 2018, 140 (51), 18151–18161. 81. Truitt, M. J.; Toporek, S. S.; Rovira-Truitt, R.; White, J. L. Alkane C  H Bond Activation in Zeolites: Evidence for Direct Protium Exchange. J. Am. Chem. Soc. 2006, 128 (6), 1847–1852. 82. Van Speybroeck, V.; De Wispelaere, K.; Van der Mynsbrugge, J.; Vandichel, M.; Hemelsoet, K.; Waroquier, M. First Principle Chemical Kinetics in Zeolites: The Methanol-toOlefin Process as a Case Study. Chem. Soc. Rev. 2014, 43 (21), 7326–7357. 83. De Wispelaere, K.; Ensing, B.; Ghysels, A.; Meijer, E. J.; Van Speybroeck, V. Complex Reaction Environments and Competing Reaction Mechanisms in Zeolite Catalysis: Insights From Advanced Molecular Dynamics. Chem. Eur. J. 2015, 21 (26), 9385–9396. 84. Vener, M. V.; Rozanska, X.; Sauer, J. Protonation of Water Clusters in the Cavities of Acidic Zeolites:(H 2 O) n$ H-Chabazite, n ¼ 1–4. Phys. Chem. Chem. Phys. 2009, 11 (11), 1702–1712. 85. Liu, P.; Mei, D. Identifying Free Energy Landscapes of Proton Transfer Processes Between Brønsted Acid Site and Water Clusters Inside the Zeolite Pores. J. Phys. Chem. C 2020, 124 (41), 22568–22576. 86. Jiménez-Ruiz, M.; Gahle, D.; Lemishko, T.; Valencia, S.; Sastre, G.; Rey, F. Evidence of Hydronium Formation in Water–Chabazite Zeolite Using Inelastic Neutron Scattering Experiments and Ab Initio Molecular Dynamics Simulations. J. Phys. Chem. C 2020, 124 (9), 5436–5443. 87. Wang, M.; Jaegers, N. R.; Lee, M.-S.; Wan, C.; Hu, J. Z.; Shi, H.; Mei, D.; Burton, S. D.; Camaioni, D. M.; Gutiérrez, O. Y. Genesis and Stability of Hydronium Ions in Zeolite Channels. J. Am. Chem. Soc. 2019, 141 (8), 3444–3455. 88. Zecchina, A.; Geobaldo, F.; Spoto, G.; Bordiga, S.; Ricchiardi, G.; Buzzoni, R.; Petrini, G. FTIR Investigation of the Formation of Neutral and Ionic Hydrogen-Bonded Complexes by Interaction of H-ZSM-5 and H-Mordenite With CH3CN and H2O: Comparison With the H-NAFION Superacidic System. J. Phys. Chem. 1996, 100 (41), 16584–16599. 89. Nusterer, E.; Blöchl, P.; Schwarz, K. Interaction of Water and Methanol With a Zeolite at High Coverages. Chem. Phys. Lett. 1996, 253 (5–6), 448–455. 90. Schwarz, K.; Nusterer, E.; Blöchl, P. E. First-Principles Molecular Dynamics Study of Small Molecules in Zeolites. Catal. Today 1999, 50 (3–4), 501–509. 91. Heard, C. J.; Grajciar, L.; Nachtigall, P. The Effect of Water on the Validity of Löwenstein’s Rule. Chem. Sci. 2019, 10 (22), 5705–5711. 92. Bailleul, S.; Rogge, S. M. J.; Vanduyfhuys, L.; Van Speybroeck, V. Insight Into the Role of Water on the Methylation of Hexamethylbenzene in H-SAPO-34 From First Principle Molecular Dynamics Simulations. ChemCatChem 2019, 11 (16), 3993–4010. 93. Olson, D.; Haag, W.; Borghard, W. Use of Water as a Probe of Zeolitic Properties: Interaction of Water With HZSM-5. Microporous Mesoporous Mater. 2000, 35, 435–446. 94. Eckstein, S.; Hintermeier, P. H.; Zhao, R.; Baráth, E.; Shi, H.; Liu, Y.; Lercher, J. A. Influence of Hydronium Ions in Zeolites on Sorption. Angew. Chem. Int. Ed. 2019, 58 (11), 3450–3455. 95. Collinge, G.; Yuk, S. F.; Nguyen, M.-T.; Lee, M.-S.; Glezakou, V.-A.; Rousseau, R. Effect of Collective Dynamics and Anharmonicity on Entropy in Heterogenous Catalysis: Building the Case for Advanced Molecular Simulations. ACS Catal. 2020, 10 (16), 9236–9260.  96. Heard, C. J.; Grajciar, L.; Uhlík, F.; Shamzhy, M.; Opanasenko, M.; Cejka, J.; Nachtigall, P. Zeolite (In) Stability Under Aqueous or Steaming Conditions. Adv. Mater. 2020;, 2003264. 97. Chizallet, C. Toward the Atomic Scale Simulation of Intricate Acidic Aluminosilicate Catalysts. ACS Catal. 2020, 10 (10), 5579–5601. 98. Silaghi, M.-C.; Chizallet, C.; Raybaud, P. Challenges on Molecular Aspects of Dealumination and Desilication of Zeolites. Microporous Mesoporous Mater. 2014, 191, 82–96. 99. Heard, C. J.; Grajciar, L.; Rice, C. M.; Pugh, S. M.; Nachtigall, P.; Ashbrook, S. E.; Morris, R. E. Fast Room Temperature Lability of Aluminosilicate Zeolites. Nat. Commun. 2019, 10 (1), 1–7. 100. Pugh, S. M.; Wright, P. A.; Law, D. J.; Thompson, N.; Ashbrook, S. E. Facile, Room-Temperature 17O Enrichment of Zeolite Frameworks Revealed by Solid-State NMR Spectroscopy. J. Am. Chem. Soc. 2019, 142 (2), 900–906. 101. Sun, T.; Xu, S.; Xiao, D.; Liu, Z.; Li, G.; Zheng, A.; Liu, W.; Xu, Z.; Cao, Y.; Guo, Q. Water-Induced Structural Dynamic Process in Molecular Sieves Under Mild Hydrothermal Conditions: Ship-in-a-Bottle Strategy for Acidity Identification and Catalyst Modification. Angew. Chem. 2020, 132 (46), 20853–20862.

Dynamic evolution of catalytic active sites within zeolite catalysis

197

102. Lisboa, O.; Sánchez, M.; Ruette, F. Modeling Extra Framework Aluminum (EFAL) Formation in the Zeolite ZSM-5 Using Parametric Quantum and DFT Methods. J. Mol. Catal. A Chem. 2008, 294 (1–2), 93–101. 103. Malola, S.; Svelle, S.; Bleken, F. L.; Swang, O. Detailed Reaction Paths for Zeolite Dealumination and Desilication From Density Functional Calculations. Angew. Chem. Int. Ed. 2012, 51 (3), 652–655. 104. Sokol, A.; Catlow, C.; Garces, J.; Kuperman, A. Defect Centers in Microporous Aluminum Silicate Materials. J. Phys. Chem. B. 1998, 102 (52), 10647–10649. 105. Sokol, A. A.; Catlow, C. R. A.; Garces, J.; Kuperman, A. Computational Investigation Into the Origins of Lewis Acidity in Zeolites. Adv. Mater. 2000, 12 (23), 1801–1805. 106. Sokol, A. A.; Catlow, C. R. A.; Garcés, J. M.; Kuperman, A. Local States in Microporous Silica and Aluminum Silicate Materials. 1. Modeling Structure, Formation, and Transformation of Common Hydrogen Containing Defects. J. Phys. Chem. B. 2002, 106 (24), 6163–6177. 107. Silaghi, M.-C.; Chizallet, C.; Petracovschi, E.; Kerber, T.; Sauer, J.; Raybaud, P. Regioselectivity of Al–O Bond Hydrolysis During Zeolites Dealumination Unified by Brønsted– Evans–Polanyi Relationship. ACS Catal. 2015, 5 (1), 11–15. 108. Silaghi, M.-C.; Chizallet, C.; Sauer, J.; Raybaud, P. Dealumination Mechanisms of Zeolites and Extra-Framework Aluminum Confinement. J. Catal. 2016, 339, 242–255. 109. Nielsen, M.; Brogaard, R. Y.; Falsig, H.; Beato, P.; Swang, O.; Svelle, S. Kinetics of Zeolite Dealumination: Insights From H-SSZ-13. ACS Catal. 2015, 5 (12), 7131–7139. 110. Holzinger, J.; Beato, P.; Lundegaard, L. F.; Skibsted, J. Distribution of Aluminum Over the Tetrahedral Sites in ZSM-5 Zeolites and Their Evolution After Steam Treatment. J. Phys. Chem. C 2018, 122 (27), 15595–15613. 111. Valdiviés-Cruz, K.; Lam, A.; Zicovich-Wilson, C. M. Chemical Interaction of Water Molecules With Framework Al in Acid Zeolites: A Periodic Ab Initio Study on H-Clinoptilolite. Phys. Chem. Chem. Phys. 2015, 17 (36), 23657–23666. 112. Stanciakova, K.; Ensing, B.; Göltl, F.; Bulo, R. E.; Weckhuysen, B. M. Cooperative Role of Water Molecules During the Initial Stage of Water-Induced Zeolite Dealumination. ACS Catal. 2019, 9 (6), 5119–5135. 113. Nielsen, M.; Hafreager, A.; Brogaard, R. Y.; De Wispelaere, K.; Falsig, H.; Beato, P.; Van Speybroeck, V.; Svelle, S. Collective Action of Water Molecules in Zeolite Dealumination. Cat. Sci. Technol. 2019, 9 (14), 3721–3725. 114. Sazama, P.; Wichterlova, B.; Dedecek, J.; Tvaruzkova, Z.; Musilova, Z.; Palumbo, L.; Sklenak, S.; Gonsiorova, O. FTIR and 27Al MAS NMR Analysis of the Effect of Framework Al-and Si-Defects in Micro-and Micro-Mesoporous H-ZSM-5 on Conversion of Methanol to Hydrocarbons. Microporous Mesoporous Mater. 2011, 143 (1), 87–96. 115. Brus, J.; Kobera, L.; Schoefberger, W.; Urbanová, M.; Klein, P.; Sazama, P.; Tabor, E.; Sklenak, S.; Fishchuk, A. V.; Dedecek, J. Structure of Framework Aluminum Lewis Sites and Perturbed Aluminum Atoms in Zeolites as Determined by 27Al {1H} REDOR (3Q) MAS NMR Spectroscopy and DFT/Molecular Mechanics. Angew. Chem. Int. Ed. 2015, 54 (2), 541–545. 116. Chen, K.; Horstmeier, S.; Nguyen, V. T.; Wang, B.; Crossley, S. P.; Pham, T.; Gan, Z.; Hung, I.; White, J. L. Structure and Catalytic Characterization of a Second Framework Al (IV) Site in Zeolite Catalysts Revealed by NMR at 35.2 T. J. Am. Chem. Soc. 2020, 142 (16), 7514–7523. 117. Ruiz, J. M.; McAdon, M. H.; Garces, J. M. Aluminum Complexes as Models for Broensted Acid Sites in Zeolites: Structure and Energetics of [Al (OH)4]-,[Al (H2O)6]3þ, and Intermediate Monomeric Species [Al(OH) x (H2O) n-x$mH2O]3-x Obtained by Hydrolysis. J. Phys. Chem. B. 1997, 101 (10), 1733–1744. 118. Benco, L.; Demuth, T.; Hafner, J.; Hutschka, F.; Toulhoat, H. Extraframework Aluminum Species in Zeolites: Ab Initio Molecular Dynamics Simulation of Gmelinite. J. Catal. 2002, 209 (2), 480–488. 119. Bhering, D. L.; Ramírez-Solís, A.; Mota, C. J. A Density Functional Theory Based Approach to Extraframework Aluminum Species in Zeolites. J. Phys. Chem. B. 2003, 107 (18), 4342–4347. 120. Ong, L. H.; Dömök, M.; Olindo, R.; van Veen, A. C.; Lercher, J. A. Dealumination of HZSM-5 Via Steam-Treatment. Microporous Mesoporous Mater. 2012, 164, 9–20. 121. van Bokhoven, J. A.; Van der Eerden, A. M.; Koningsberger, D. C. Three-Coordinate Aluminum in Zeolites Observed With In Situ X-ray Absorption near-Edge Spectroscopy at the Al K-Edge: Flexibility of Aluminum Coordinations in Zeolites. J. Am. Chem. Soc. 2003, 125 (24), 7435–7442. 122. Ravi, M.; Sushkevich, V. L.; van Bokhoven, J. A. Lewis Acidity Inherent to the Framework of Zeolite Mordenite. J. Phys. Chem. C 2019, 123 (24), 15139–15144. 123. Yi, X.; Liu, K.; Chen, W.; Li, J.; Xu, S.; Li, C.; Xiao, Y.; Liu, H.; Guo, X.; Liu, S.-B. Origin and Structural Characteristics of Tri-Coordinated Extra-Framework Aluminum Species in Dealuminated Zeolites. J. Am. Chem. Soc. 2018, 140 (34), 10764–10774. 124. Liu, C.; Li, G.; Hensen, E. J.; Pidko, E. A. Nature and Catalytic Role of Extraframework Aluminum in Faujasite Zeolite: A Theoretical Perspective. ACS Catal. 2015, 5 (11), 7024–7033. 125. Liu, C.; Li, G.; Hensen, E. J.; Pidko, E. A. Relationship between Acidity and Catalytic Reactivity of Faujasite Zeolite: A Periodic DFT Study. J. Catal. 2016, 344, 570–577. 126. Li, S.; Zheng, A.; Su, Y.; Zhang, H.; Chen, L.; Yang, J.; Ye, C.; Deng, F. Brønsted/Lewis Acid Synergy in Dealuminated HY Zeolite: A Combined Solid-State NMR and Theoretical Calculation Study. J. Am. Chem. Soc. 2007, 129 (36), 11161–11171. 127. Li, S.; Zheng, A.; Su, Y.; Fang, H.; Shen, W.; Yu, Z.; Chen, L.; Deng, F. Extra-Framework Aluminium Species in Hydrated Faujasite Zeolite as Investigated by Two-Dimensional Solid-State NMR Spectroscopy and Theoretical Calculations. Phys. Chem. Chem. Phys. 2010, 12 (15), 3895–3903. 128. Chen, K.; Abdolrhamani, M.; Sheets, E.; Freeman, J.; Ward, G.; White, J. L. Direct Detection of Multiple Acidic Proton Sites in Zeolite HZSM-5. J. Am. Chem. Soc. 2017, 139 (51), 18698–18704. 129. Chen, K.; Abdolrahmani, M.; Horstmeier, S.; Pham, T. N.; Nguyen, V. T.; Zeets, M.; Wang, B.; Crossley, S.; White, J. L. Brønsted–Brønsted Synergies Between Framework and Noncrystalline Protons in Zeolite H-ZSM-5. ACS Catal. 2019, 9 (7), 6124–6136. 130. Schallmoser, S.; Ikuno, T.; Wagenhofer, M.; Kolvenbach, R.; Haller, G.; Sanchez-Sanchez, M.; Lercher, J. Impact of the Local Environment of Brønsted Acid Sites in ZSM-5 on the Catalytic Activity in n-Pentane Cracking. J. Catal. 2014, 316, 93–102. 131. Xue, N.; Vjunov, A.; Schallmoser, S.; Fulton, J. L.; Sanchez-Sanchez, M.; Hu, J. Z.; Mei, D.; Lercher, J. A. Hydrolysis of Zeolite Framework Aluminum and its Impact on Acid Catalyzed Alkane Reactions. J. Catal. 2018, 365, 359–366. 132. Naccache, C.; Taarit, Y. B. Transition Metal Exchanged Zeolites: Physical and Catalytic Properties. In Zeolites: Science and Technology; Ribeiro, F. R., Rodrigues, A. E., Rollmann, L. D., Naccache, C., Eds., Springer Netherlands: Dordrecht, 1984; pp 373–396. 133. Sachtler, W. M. H.; Zhang, Z. Zeolite-Supported Transition Metal Catalysts*. In Advances in Catalysis; Eley, D. D., Pines, H., Weisz, P. B., Eds.; vol. 39; Academic Press, 1993; pp 129–220. 134. Armor, J. N. Metal-Exchanged Zeolites as catalysts. Microporous Mesoporous Mater. 1998, 22 (1), 451–456. 135. Pârvulescu, V. I.; Grange, P.; Delmon, B. Catalytic Removal of NO. Catal. Today 1998, 46 (4), 233–316. 136. Deka, U.; Lezcano-Gonzalez, I.; Weckhuysen, B. M.; Beale, A. M. Local Environment and Nature of Cu Active Sites in Zeolite-Based Catalysts for the Selective Catalytic Reduction of NOx. ACS Catal. 2013, 3 (3), 413–427. 137. EPA, U. S. Integrated Science Assessment (ISA) for Oxides of NitrogendHealth Criteria (Final Report, Jan 2016), U.S. Environmental Protection Agency: Washington, DC, 2016. EPA/600/R-15/068. 138. Li, J.; Chang, H.; Ma, L.; Hao, J.; Yang, R. T. Low-Temperature Selective Catalytic Reduction of NOx With NH3 over Metal Oxide and Zeolite CatalystsdA Review. Catal. Today 2011, 175 (1), 147–156. 139. Brandenberger, S.; Kröcher, O.; Tissler, A.; Althoff, R. The State of the Art in Selective Catalytic Reduction of NOx by Ammonia Using Metal-Exchanged Zeolite Catalysts. Catal. Rev. 2008, 50 (4), 492–531. 140. Beale, A. M.; Gao, F.; Lezcano-Gonzalez, I.; Peden, C. H. F.; Szanyi, J. Recent Advances in Automotive Catalysis for NOx Emission Control by Small-Pore Microporous Materials. Chem. Soc. Rev. 2015, 44 (20), 7371–7405. 141. Busca, G.; Lietti, L.; Ramis, G.; Berti, F. Chemical and Mechanistic Aspects of the Selective Catalytic Reduction of NOx by Ammonia over Oxide Catalysts: A Review. Appl. Catal. Environ. 1998, 18 (1), 1–36.

198

Dynamic evolution of catalytic active sites within zeolite catalysis

142. Seiyama, T.; Arakawa, T.; Matsuda, T.; Takita, Y.; Yamazoe, N. Catalytic Activity of Transition Metal Ion Exchanged Y Zeolites in the Reduction of Nitric Oxide With Ammonia. J. Catal. 1977, 48 (1), 1–7. 143. Iwamoto, M.; Furukawa, H.; Mine, Y.; Uemura, F.; Mikuriya, S.-I.; Kagawa, S. Copper(II) Ion-Exchanged ZSM-5 Zeolites as Highly Active Catalysts for Direct and Continuous Decomposition of Nitrogen Monoxide. J. Chem. Soc. Chem. Commun. 1986, (16), 1272–1273. 144. Iwamoto, M.; Hamada, H. Removal of Nitrogen Monoxide From Exhaust Gases Through Novel Catalytic Processes. Catal. Today 1991, 10 (1), 57–71. 145. Iwamoto, M.; Yahiro, H.; Tanda, K.; Mizuno, N.; Mine, Y.; Kagawa, S. Removal of Nitrogen Monoxide Through a Novel Catalytic Process. 1. Decomposition on Excessively Copper-Ion-Exchanged ZSM-5 Zeolites. J. Phys. Chem. 1991, 95 (9), 3727–3730. 146. Iwamoto, M.; Yahiro, H.; Mizuno, N.; Zhang, W. X.; Mine, Y.; Furukawa, H.; Kagawa, S. Removal of Nitrogen Monoxide Through a Novel Catalytic Process. 2. Infrared Study on Surface Reaction of Nitrogen Monoxide Adsorbed on Copper Ion-Exchanged ZSM-5 Zeolites. J. Phys. Chem. 1992, 96 (23), 9360–9366. 147. Kwak, J. H.; Tonkyn, R. G.; Kim, D. H.; Szanyi, J.; Peden, C. H. F. Excellent Activity and Selectivity of Cu-SSZ-13 in the Selective Catalytic Reduction of NOx With NH3. J. Catal. 2010, 275 (2), 187–190. 148. Kwak, J. H.; Tran, D.; Burton, S. D.; Szanyi, J.; Lee, J. H.; Peden, C. H. F. Effects of Hydrothermal Aging on NH3-SCR Reaction Over Cu/Zeolites. J. Catal. 2012, 287, 203–209. 149. Borfecchia, E.; Beato, P.; Svelle, S.; Olsbye, U.; Lamberti, C.; Bordiga, S. Cu-CHAdA Model System for Applied Selective Redox Catalysis. Chem. Soc. Rev. 2018, 47 (22), 8097–8133. 150. Paolucci, C.; Di Iorio, J. R.; Ribeiro, F. H.; Gounder, R.; Schneider, W. F. Chapter OnedCatalysis Science of NOx Selective Catalytic Reduction With Ammonia Over Cu-SSZ-13 and Cu-SAPO-34. In Advances in Catalysis; Song, C., Ed.; 59; Academic Press, 2016; pp 1–107. 151. Hun Kwak, J.; Zhu, H.; Lee, J. H.; Peden, C. H. F.; Szanyi, J. Two Different Cationic Positions in Cu-SSZ-13? Chem. Commun. 2012, 48 (39), 4758–4760. 152. Gao, F.; Walter, E. D.; Kollar, M.; Wang, Y.; Szanyi, J.; Peden, C. H. F. Understanding Ammonia Selective Catalytic Reduction Kinetics over Cu/SSZ-13 From Motion of the Cu Ions. J. Catal. 2014, 319, 1–14. 153. Kwak, J. H.; Varga, T.; Peden, C. H. F.; Gao, F.; Hanson, J. C.; Szanyi, J. Following the Movement of Cu Ions in a SSZ-13 Zeolite during Dehydration, Reduction and Adsorption: A Combined in Situ TP-XRD, XANES/DRIFTS Study. J. Catal. 2014, 314, 83–93. 154. Lezcano-Gonzalez, I.; Deka, U.; Arstad, B.; Van Yperen-De Deyne, A.; Hemelsoet, K.; Waroquier, M.; Van Speybroeck, V.; Weckhuysen, B. M.; Beale, A. M. Determining the Storage, Availability and Reactivity of NH3 within Cu-Chabazite-Based Ammonia Selective Catalytic Reduction Systems. Phys. Chem. Chem. Phys. 2014, 16 (4), 1639–1650. 155. Lezcano-Gonzalez, I.; Wragg, D. S.; Slawinski, W. A.; Hemelsoet, K.; Van Yperen-De Deyne, A.; Waroquier, M.; Van Speybroeck, V.; Beale, A. M. Determination of the Nature of the Cu Coordination Complexes Formed in the Presence of NO and NH3 Within SSZ-13. J. Phys. Chem. C 2015, 119 (43), 24393–24403. 156. Giordanino, F.; Borfecchia, E.; Lomachenko, K. A.; Lazzarini, A.; Agostini, G.; Gallo, E.; Soldatov, A. V.; Beato, P.; Bordiga, S.; Lamberti, C. Interaction of NH3 With Cu-SSZ-13 Catalyst: A Complementary FTIR, XANES, and XES Study. J. Phys. Chem. Lett. 2014, 5 (9), 1552–1559. 157. Paolucci, C.; Parekh, A. A.; Khurana, I.; Di Iorio, J. R.; Li, H.; Albarracin Caballero, J. D.; Shih, A. J.; Anggara, T.; Delgass, W. N.; Miller, J. T.; Ribeiro, F. H.; Gounder, R.; Schneider, W. F. Catalysis in a Cage: Condition-Dependent Speciation and Dynamics of Exchanged Cu Cations in SSZ-13 Zeolites. J. Am. Chem. Soc. 2016, 138 (18), 6028–6048. 158. Lomachenko, K. A.; Borfecchia, E.; Negri, C.; Berlier, G.; Lamberti, C.; Beato, P.; Falsig, H.; Bordiga, S. The Cu-CHA deNOx Catalyst in Action: Temperature-Dependent NH3Assisted Selective Catalytic Reduction Monitored by Operando XAS and XES. J. Am. Chem. Soc. 2016, 138 (37), 12025–12028. 159. Millan, R.; Cnudde, P.; Hoffman, A. E. J.; Lopes, C. W.; Concepción, P.; van Speybroeck, V.; Boronat, M. Theoretical and Spectroscopic Evidence of the Dynamic Nature of Copper Active Sites in Cu-CHA Catalysts under Selective Catalytic Reduction (NH3–SCR–NOx) Conditions. J. Phys. Chem. Lett. 2020, 10060–10066. 160. Gao, F.; Mei, D.; Wang, Y.; Szanyi, J.; Peden, C. H. F. Selective Catalytic Reduction Over Cu/SSZ-13: Linking Homo- and Heterogeneous Catalysis. J. Am. Chem. Soc. 2017, 139 (13), 4935–4942. 161. Negri, C.; Selleri, T.; Borfecchia, E.; Martini, A.; Lomachenko, K. A.; Janssens, T. V. W.; Cutini, M.; Bordiga, S.; Berlier, G. Structure and Reactivity of Oxygen-Bridged Diamino Dicopper(II) Complexes in Cu-Ion-Exchanged Chabazite Catalyst for NH3-Mediated Selective Catalytic Reduction. J. Am. Chem. Soc. 2020, 142 (37), 15884–15896. 162. Liu, C.; Kubota, H.; Toyao, T.; Maeno, Z.; Shimizu, K.-I. Mechanistic Insights Into the Oxidation of Copper(I) Species During NH3-SCR Over Cu-CHA Zeolites: A DFT Study. Cat. Sci. Technol. 2020, 10 (11), 3586–3593. 163. Chen, L.; Janssens, T. V. W.; Vennestrøm, P. N. R.; Jansson, J.; Skoglundh, M.; Grönbeck, H. A Complete Multisite Reaction Mechanism for Low-Temperature NH3-SCR over Cu-CHA. ACS Catal. 2020, 10 (10), 5646–5656. 164. Chen, L.; Janssens, T. V. W.; Grönbeck, H. A Comparative Test of Different Density Functionals for Calculations of NH3-SCR Over Cu-Chabazite. Phys. Chem. Chem. Phys. 2019, 21 (21), 10923–10930. 165. Mandal, K.; Gu, Y.; Westendorff, K. S.; Li, S.; Pihl, J. A.; Grabow, L. C.; Epling, W. S.; Paolucci, C. Condition-Dependent Pd Speciation and NO Adsorption in Pd/Zeolites. ACS Catal. 2020, 10 (21), 12801–12818. 166. Kubota, T.; Kumada, F.; Tominaga, H.; Kunugi, T. Oxidation of Propylene Over a Pd(II)-Cu(II)-Y Zeolite Catalyst. Int. Chem. Eng. 1973, 13, 539–545. 167. Arai, H.; Yamashiro, T.; Kubo, T.; Tominaga, H. The Catalysis of Palladium and Cupric Ion-Exchanged Zeolite for Oxidation of Ethylene. J. Jpn. Pet. Inst. 1976, 18. 168. Lapidus, A. L.; Maganya, M. I.; Maltsev, V. V. Ethylene Oligomerization Over Palladium and Rhodium Aluminosilicate Catalysts. Neftekhimiya 1978, 18, 376–379. 169. Takahashi, N.; Fujiwara, Y.; Mijin, A. Reaction of Ethylene/Propylene Mixture over Rh$Y Zeolite. Zeolites 1985, 5 (6), 363–364. 170. Yashima, T.; Ushida, Y.; Ebisawa, M.; Hara, N. Polymerization of Ethylene Over Transition-Metal Exchanged Y Zeolites. J. Catal. 1975, 36 (3), 320–326. 171. Ogino, I.; Gates, B. C. Role of the Support in Catalysis: Activation of a Mononuclear Ruthenium Complex for Ethene Dimerization by Chemisorption on Dealuminated Zeolite Y. Chem. A Eur. J. 2009, 15 (28), 6827–6837. 172. Serna, P.; Gates, B. C. A Bifunctional Mechanism for Ethene Dimerization: Catalysis by Rhodium Complexes on Zeolite HY in the Absence of Halides. Angew. Chem. Int. Ed. 2011, 50 (24), 5528–5531. 173. Shibata, J.; Shimizu, K.-I.; Takada, Y.; Shichi, A.; Yoshida, H.; Satokawa, S.; Satsuma, A.; Hattori, T. Structure of Active Ag Clusters in Ag Zeolites for SCR of NO by Propane in the Presence of Hydrogen. J. Catal. 2004, 227 (2), 367–374. 174. Shibata, J.; Takada, Y.; Shichi, A.; Satokawa, S.; Satsuma, A.; Hattori, T. Ag Cluster as Active Species for SCR of NO by Propane in the Presence of Hydrogen over Ag-MFI. J. Catal. 2004, 222 (2), 368–376. 175. Linnert, T.; Mulvaney, P.; Henglein, A.; Weller, H. Long-Lived Nonmetallic Silver Clusters in Aqueous Solution: Preparation and Photolysis. J. Am. Chem. Soc. 1990, 112 (12), 4657–4664. 176. Mulvaney, P.; Henglein, A. Long-Lived Nonmetallic Silver Clusters in Aqueous Solution: A Pulse Radiolysis Study of Their Formation. J. Phys. Chem. 1990, 94 (10), 4182–4188. 177. Ershov, B. G.; Janata, E.; Henglein, A. Growth of Silver Particles in Aqueous Solution: Long-Lived “Magic” Clusters and Ionic Strength Effects. J. Phys. Chem. 1993, 97 (2), 339–343. 178. Gachard, E.; Belloni, J.; Subramanian, M. A. Optical and EPR Spectroscopic Studies of Silver Clusters in Ag,Na-Y Zeolite by g-Irradiation. J. Mater. Chem. 1996, 6 (5), 867–870. 179. Michalik, J.; Sadlo, J.; Kodaira, T.; Shimomura, S.; Yamada, H. ESR and Optical Studies of Cationic Silver Clusters in Zeolite Rho. J. Radioanal. Nucl. Chem. 1998, 232 (1), 135–138. 180. Sato, K.; Yoshinari, T.; Kintaichi, Y.; Haneda, M.; Hamada, H. Remarkable Promoting Effect of Rhodium on the Catalytic Performance of Ag/Al2O3 for the Selective Reduction of NO With Decane. Appl. Catal. Environ. 2003, 44 (1), 67–78.

Dynamic evolution of catalytic active sites within zeolite catalysis

199

181. Chemical Evidence for Charged Clusters in Silver Zeolites. In Studies in Surface Science and Catalysis; Beyer, H. K., Jacobs, P. A., Jacobs, P. A., Jaeger, N. I., JírŮ, P., Schulz-Ekloff, G., Eds.; vol. 12; Elsevier, 1982; pp 95–102. 182. Bruix, A.; Margraf, J. T.; Andersen, M.; Reuter, K. First-Principles-Based Multiscale Modelling of Heterogeneous Catalysis. Nat. Catal. 2019, 2 (8), 659–670. 183. Grajciar, L.; Heard, C. J.; Bondarenko, A. A.; Polynski, M. V.; Meeprasert, J.; Pidko, E. A.; Nachtigall, P. Towards Operando Computational Modeling in Heterogeneous Catalysis. Chem. Soc. Rev. 2018, 47 (22), 8307–8348. 184. Bucko, T.; Benco, L.; Hafner, J.; Ángyán, J. G. Monomolecular Cracking of Propane Over Acidic Chabazite: An Ab Initio Molecular Dynamics and Transition Path Sampling Study. J. Catal. 2011, 279 (1), 220–228. 185. Dellago, C.; Bolhuis, P. G.; Geissler, P. L. Transition Path Sampling. In Advances in Chemical Physics; Prigogine, I., Rice, S. A., Eds., John Wiley & Sons, Inc., 2002; pp 1–78. 186. van Erp, T. S.; Bolhuis, P. G. Elaborating Transition Interface Sampling Methods. J. Comput. Phys. 2005, 205 (1), 157–181. 187. Rohrdanz, M. A.; Zheng, W.; Clementi, C. Discovering Mountain Passes Via Torchlight: Methods for the Definition of Reaction Coordinates and Pathways in Complex Macromolecular Reactions. Annu. Rev. Phys. Chem. 2013, 64 (1), 295–316. 188. Krivov, S. V.; Karplus, M. Hidden Complexity of Free Energy Surfaces for Peptide (Protein) Folding. PNAS 2004, 101 (41), 14766–14770. 189. Bailleul, S.; Dedecker, K.; Cnudde, P.; Vanduyfhuys, L.; Waroquier, M.; Van Speybroeck, V. Ab Initio Enhanced Sampling Kinetic Study on MTO Ethene Methylation Reaction. J. Catal. 2020, 388, 38–51. 190. Rey, J.; Raybaud, P.; Chizallet, C.; Bucko, T. Competition of Secondary Versus Tertiary Carbenium Routes for the Type B Isomerization of Alkenes Over Acid Zeolites Quantified by Ab Initio Molecular Dynamics Simulations. ACS Catal. 2019, 9813–9828. 191. De Wispelaere, K.; Bailleul, S.; Van Speybroeck, V. Towards Molecular Control of Elementary Reactions in Zeolite Catalysis by Advanced Molecular Simulations Mimicking Operating Conditions. Cat. Sci. Technol. 2016, 6 (8), 2686–2705. 192. Rey, J.; Bignaud, C.; Raybaud, P.; Bucko, T.; Chizallet, C. Dynamic Features of Transition States for b-Scission Reactions of Alkenes Over Acid Zeolites Revealed by AIMD Simulations. Angew. Chem. 2020, 132 (43), 19100–19104. 193. Laio, A.; Rodriguez-Fortea, A.; Gervasio, F. L.; Ceccarelli, M.; Parrinello, M. Assessing the Accuracy of Metadynamicsy. J. Phys. Chem. B 2005, 109 (14), 6714–6721. 194. Laio, A.; Parrinello, M. Escaping Free-Energy Minima. PNAS 2002, 99 (20), 12562–12566. 195. Laio, A.; Gervasio, F. L. Metadynamics: A Method to Simulate Rare Events and Reconstruct the Free Energy in Biophysics, Chemistry and Material Science. Rep. Prog. Phys. 2008, 71 (12), 126601. 196. Barducci, A.; Bonomi, M.; Parrinello, M. Metadynamics. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2011, 1 (5), 826–843. 197. Raiteri, P.; Laio, A.; Gervasio, F. L.; Micheletti, C.; Parrinello, M. Efficient Reconstruction of Complex Free Energy Landscapes by Multiple Walkers Metadynamics. J. Phys. Chem. B 2006, 110 (8), 3533–3539. 198. Torrie, G. M.; Valleau, J. P. Nonphysical Sampling Distributions in Monte Carlo Free-Energy Estimation: Umbrella Sampling. J. Comput. Phys. 1977, 23 (2), 187–199. 199. Kumar, S.; Rosenberg, J. M.; Bouzida, D.; Swendsen, R. H.; Kollman, P. A. THE Weighted Histogram Analysis Method for Free-Energy Calculations on Biomolecules. I. THE Method. J. Comput. Chem. 1992, 13 (8), 1011–1021. 200. Kirkwood, J. G. Statistical Mechanics of Fluid Mixtures. J. Chem. Phys. 1935, 3 (5), 300–313. 201. Kästner, J. Umbrella sampling. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2011, 1 (6), 932–942. 202. Hansen, N.; van Gunsteren, W. F. Practical Aspects of Free-Energy Calculations: A Review. J. Chem. Theory Comput. 2014, 10 (7), 2632–2647. 203. Abrams, C.; Bussi, G. Enhanced Sampling in Molecular Dynamics Using Metadynamics, Replica-Exchange, and Temperature-Acceleration. Entropy 2013, 16 (1), 163–199. 204. Pietrucci, F. Strategies for the Exploration of Free Energy Landscapes: Unity in Diversity and Challenges Ahead. Rev. Phys. 2017, 2, 32–45. 205. Hajek, J.; Van der Mynsbrugge, J.; De Wispelaere, K.; Cnudde, P.; Vanduyfhuys, L.; Waroquier, M.; Van Speybroeck, V. On the Stability and Nature of Adsorbed Pentene in Brønsted Acid Zeolite H-ZSM-5 at 323 K. J. Catal. 2016, 340, 227–235. 206. Hoffman, A. J.; Bates, J. S.; Di Iorio, J. R.; Nystrom, S. V.; Nimlos, C. T.; Gounder, R.; Hibbitts, D. Rigid Arrangements of Ionic Charge in Zeolite Frameworks Conferred by Specific Aluminum Distributions Preferentially Stabilize Alkanol Dehydration Transition States. Angew. Chem. Int. Ed. 2020, 59 (42), 18686–18694. 207. Bates, J. S.; Bukowski, B. C.; Greeley, J.; Gounder, R. Structure and Solvation of Confined Water and Water–Ethanol Clusters Within Microporous Brønsted Acids and Their Effects on Ethanol Dehydration Catalysis. Chem. Sci. 2020, 11 (27), 7102–7122. 208. Chen, K.; Damron, J.; Pearson, C.; Resasco, D.; Zhang, L.; White, J. L. Zeolite Catalysis: Water Can Dramatically Increase or Suppress Alkane C–H Bond Activation. ACS Catal. 2014, 4 (9), 3039–3044. 209. Chen, K.; Gumidyala, A.; Abdolrhamani, M.; Villines, C.; Crossley, S.; White, J. L. Trace Water Amounts Can Increase Benzene H/D Exchange Rates in an Acidic Zeolite. J. Catal. 2017, 351, 130–135. 210. Liu, Y.; Vjunov, A.; Shi, H.; Eckstein, S.; Camaioni, D. M.; Mei, D.; Baráth, E.; Lercher, J. A. Enhancing the Catalytic Activity of Hydronium Ions Through Constrained Environments. Nat. Commun. 2017, 8 (1), 1–8. 211. Hemelsoet, K.; Van der Mynsbrugge, J.; De Wispelaere, K.; Waroquier, M.; Van Speybroeck, V. Unraveling the Reaction Mechanisms Governing Methanol-to-Olefins Catalysis by Theory and Experiment. ChemPhysChem 2013, 14 (8), 1526–1545. 212. Yarulina, I.; Chowdhury, A. D.; Meirer, F.; Weckhuysen, B. M.; Gascon, J. Recent Trends and Fundamental Insights in the Methanol-to-Hydrocarbons Process. Nat. Catal. 2018, 1 (6), 398–411. 213. Wang, W.; Seiler, M.; Hunger, M. Role of Surface Methoxy Species in the Conversion of Methanol to Dimethyl Ether on Acidic Zeolites Investigated by in Situ Stopped-Flow MAS NMR Spectroscopy. J. Phys. Chem. B. 2001, 105 (50), 12553–12558. 214. Wang, W.; Buchholz, A.; Seiler, M.; Hunger, M. Evidence for an Initiation of the Methanol-to-Olefin Process by Reactive Surface Methoxy Groups on Acidic Zeolite Catalysts. J. Am. Chem. Soc. 2003, 125 (49), 15260–15267. 215. Gale, J.; Shah, R.; Payne, M.; Stich, I.; Terakura, K. Methanol in Microporous Materials From First Principles. Catal. Today 1999, 50 (3–4), 525–532. 216. Van der Mynsbrugge, J.; Moors, S. L.; De Wispelaere, K.; Van Speybroeck, V. Insight into the Formation and Reactivity of Framework-Bound Methoxide Species in H-ZSM-5 From Static and Dynamic Molecular Simulations. ChemCatChem 2014, 6 (7), 1906–1918. 217. Lesthaeghe, D.; Van Speybroeck, V.; Marin, G. B.; Waroquier, M. Understanding the Failure of Direct C-C Coupling in the Zeolite-Catalyzed Methanol-to-Olefin Process. Angew. Chem. 2006, 118 (11), 1746–1751. 218. Maihom, T.; Boekfa, B.; Sirijaraensre, J.; Nanok, T.; Probst, M.; Limtrakul, J. Reaction Mechanisms of the Methylation of Ethene With Methanol and Dimethyl Ether Over HZSM-5: An ONIOM Study. J. Phys. Chem. C 2009, 113 (16), 6654–6662. 219. Salvador, P.; Kladnig, W. Surface Reactivity of Zeolites Type HY and Na-Y With Methanol. J. Chem. Soc., Faraday Trans. 1 1977, 73, 1153–1168. 220. O’malley, A. J.; Parker, S. F.; Chutia, A.; Farrow, M. R.; Silverwood, I. P.; García-Sakai, V.; Catlow, C. R. A. Room Temperature Methoxylation in Zeolites: Insight Into a Key Step of the Methanol-to-Hydrocarbons Process. Chem. Commun. 2016, 52 (14), 2897–2900. 221. Matam, S. K.; Howe, R. F.; Thetford, A.; Catlow, C. R. A. Room Temperature Methoxylation in Zeolite H-ZSM-5: An Operando DRIFTS/Mass Spectrometric Study. Chem. Commun. 2018, 54 (91), 12875–12878. 222. Matam, S. K.; Nastase, S. A.; Logsdail, A. J.; Catlow, C. R. A. Methanol Loading Dependent Methoxylation in Zeolite H-ZSM-5. Chem. Sci. 2020, 11 (26), 6805–6814. 223. Brogaard, R. Y.; Olsbye, U. Ethene Oligomerization in Ni-Containing Zeolites: Theoretical Discrimination of Reaction Mechanisms. ACS Catal. 2016, 6 (2), 1205–1214. 224. Sudarsanam, P.; Peeters, E.; Makshina, E. V.; Parvulescu, V. I.; Sels, B. F. Advances in Porous and Nanoscale Catalysts for Viable Biomass Conversion. Chem. Soc. Rev. 2019, 48 (8), 2366–2421.

200

Dynamic evolution of catalytic active sites within zeolite catalysis

225. Rogge, S. M. J.; Bavykina, A.; Hajek, J.; Garcia, H.; Olivos-Suarez, A. I.; Sepulveda-Escribano, A.; Vimont, A.; Clet, G.; Bazin, P.; Kapteijn, F.; Daturi, M.; RamosFernandez, E. V.; Xamena, F.; Van Speybroeck, V.; Gascon, J. Metal-Organic and Covalent Organic Frameworks as Single-Site Catalysts. Chem. Soc. Rev. 2017, 46 (11), 3134–3184. 226. Copéret, C.; Comas-Vives, A.; Conley, M. P.; Estes, D. P.; Fedorov, A.; Mougel, V.; Nagae, H.; Núñez-Zarur, F.; Zhizhko, P. A. Surface Organometallic and Coordination Chemistry Toward Single-Site Heterogeneous Catalysts: Strategies, Methods, Structures, and Activities. Chem. Rev. 2016, 116 (2), 323–421. 227. Thomas, J. M.; Raja, R.; Lewis, D. W. Single-Site Heterogeneous Catalysts. Angew. Chem. Int. Ed. 2005, 44 (40), 6456–6482.

6.09

Nanocluster heterogeneous catalysts: Insights from theory

Geng Sun and Philippe Sauteta,b,c, a Department of Chemical and Biomolecular Engineering, University of California, Los Angeles, Los Angeles, CA, United States; b Department of Chemistry and Biochemistry, University of California, Los Angeles, Los Angeles, CA, United States; and c California Nano Systems Institute, Los Angeles, CA, United States a

© 2023 Elsevier Ltd. All rights reserved.

6.09.1 6.09.1.1 6.09.1.2 6.09.2 6.09.2.1 6.09.2.2 6.09.2.3 6.09.2.4 6.09.2.4.1 6.09.2.4.2 6.09.2.4.3 6.09.3 References

Unique properties in nanocluster catalysis Reactant induced cluster reconstruction Cluster fluxionality Methods Descriptors for adsorption BEP relationship Global optimization methods to explore structures of nanocluster catalysts Machine learning methods in nanocluster catalysis Machine learning aided global optimizations Machine learning for surface chemistry of nanoclusters Machine learning methods for extracting structural information for x-ray adsorption spectroscopy Conclusion and perspective

202 202 205 208 208 211 211 214 214 215 215 219 219

Abstract Small nanoclusters are important materials holding the promise of being versatile catalysts for a wide range of applications. The intricate structures of small nanoclusters endow them with novel properties toward catalytic applications, however, the emerging structural complexity, in turn, renders the understanding of the reaction mechanisms difficult and hence hampers the rational improvement of nanocluster catalysts. One key characteristic of nanocluster catalysts is their intrinsic nonuniform active sites from either surface heterogeneity or structural diversity, which requires a comprehensive and statistical understanding of all surface sites. Theoretical studies are powerful tools for studying small nanoclusters. In this chapter, we discuss the theoretical progress that is dedicated to understanding the catalytic properties of nanoclusters bearing structure and site heterogeneity. This chapter starts by discussing nanocluster reconstructions under catalytic reaction conditions and highlights adsorbate-induced reconstruction or Ostwald ripening. We also discuss the role of minority species in catalysis. Follow next the theoretical methods that are particularly important in this field, such as global optimization methods, descriptors as well as machine learning approaches. These methods provide key concepts and protocols that can be used for a wide range of studies.

Nanoclusters consisting of a few to a few hundred atoms have been widely used for a range of innovative applications such as electrochemistry,1 catalysis,2 photoluminescence,3 etc. Nanoclusters manifest unique properties that originate from their unusual geometric or electronic structures, such as abundant undercoordinated sites, the opening of a gap between occupied and vacant states, and the strong size dependent properties.4–7 Although the small nanoclusters hold promises as novel catalysts, most advances in the domain are still based on a trial-and-error strategy. There exist practical challenges in explaining the properties of nanoclusters and in rationally designing new ones. One of the key challenges is the heterogeneity of nanocluster catalysts which renders the study of the catalytic reaction complicated.8–10 Indeed nanoclusters are intrinsically heterogeneous systems. In applications, they are supported on a different solid or suspended in a solution. The interaction between the nanocluster and the medium plays a key role in determining the properties of nanoclusters and, in consequence, the interaction between nanoclusters and the reactant/product molecules appears to be essential in understanding their catalytic properties. In addition, nanocluster catalysts are generally composed of a mixture of cluster sizes, compositions as well as geometries. Although the advanced synthesis of the well-defined nanoclusters has made important progress in the past decades, these techniques can be only used for a few metal clusters.11,12 In some cases, the as-prepared nanoclusters can reconstruct in the presence of the adsorbates, hence heterogenizing the uniform prepared structure. Since the heterogeneity of nanoclusters is intrinsic and inevitable, the statistical description of the nanocluster catalysts is critical for the interpretation of experimental results and for predictive simulations.8 Theoretical studies, that are based on ab initio calculations, are very important tools for the rationale of the experimental data as well as for the discovery of the fundamental principles to design a new catalyst. Theoretical approaches are particularly helpful in the studies of nanocluster catalysts because they can provide critical atomic insights, which are not available from the experiments, thanks to the progress in density functional theory simulations and to the emergence of many efficient software. Today theoretical

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00007-8

201

202

Nanocluster heterogeneous catalysts: Insights from theory

chemists can simulate nanoclusters containing a few hundred atoms, which are very close to the realistic catalytic systems (size < 2 nm) applied in industrial processes. Many well-established theories and tools are accessible. For example, local and transition-state optimization methods (NEB and DIMER), kinetic simulations (mean-field kinetic simulations and kinetic Monte Carlo method), as well as molecular dynamic simulations, have constructed the paradigm of multi-scale simulations, which can help to solve the mechanism of various catalytic reactions. The current chapter focusses on nanocluster catalysis. Many previous theoretical studies attempt to obtain information on nanoclusters from extended periodic surface models, and they will not be considered here, though we may re-examine the validity of a few of these conclusions when considering explicit nanocluster models. We will focus on the phenomena, methods, and theories that are unique or particularly important to understand catalysis with nanoclusters. A special interest will be placed in the approaches that try to address the heterogeneity of nanocluster catalysts. Finally, the nanocluster mentioned in this chapter have a size from a few atoms to a few hundred atoms, corresponding to the so-called non-scalable regime. Larger nanoparticles, where extended surfaces dominate, are not considered. This chapter will be organized as follows. First, we will discuss several important phenomena that take place when nanoclusters are put under catalytic reaction conditions. We will emphasize the role of adsorbates on the structures and properties of nanoclusters. We will also discuss the intrinsic fluxionality of small nanoclusters which may introduce metastable isomers at finite reaction temperature. The emergence of metastability requires a statistical description of nanocluster catalysts. In the second part, we will focus on the theories and tools that are critical to theoretical nanocluster catalysis. We will discuss the development of reactivity descriptors that handle the surface heterogeneity of nanoclusters and examine the global optimization tools to study their structure. Finally, we will discuss the quickly-developing machine learning methods that are assisting the studies of nanocluster catalysts in global optimization, handling surface heterogeneity of nanoclusters, as well as interpreting the experimental X-ray absorption spectrum.

6.09.1

Unique properties in nanocluster catalysis

6.09.1.1

Reactant induced cluster reconstruction

One important aspect for the theoretical simulations of nanocluster catalysts is their reconstruction induced by the reactants from the environment. Nanoclusters consist of many undercoordinated sites that are very active. In reaction conditions, these active sites strongly adsorb reactants and intermediates. In many cases, the strong interactions between the reactant and the undercoordinated atoms will induce ample cluster reconstruction. For example, Pt clusters are widely used for light alkane dehydrogenation producing alkene.13,14 Small Pt clusters are very active catalysts with high initial activity. However, the reaction also suffers from unwanted coke formation that quickly blocks the active sites and deactivates the catalyst. In practice, H2 is usually co-fed in the reactant stream to prevent catalyst deactivation. To understand the role of hydrogen in preventing catalyst deactivation, it is essential to know the structures of Pt clusters with H adsorbed and the role of co-adsorbed H in the dehydrogenation reaction. Mager-Maury et al. investigated the impact of H adsorption on the Pt13 cluster supported on g-Al2O3.15 They studied a wide range of H coverage (from 0 to 38H atoms) on Pt13 cluster. The optimal structures for selected hydrogen coverages are shown in Fig. 1. The morphology of the Pt13 cluster exhibits a transition from a biplanar shape at low hydrogen coverage (Fig. 1A–C) to a cuboctahedral shape at high hydrogen coverage (Fig. 1E and F) in which the morphology transition takes place at a H coverage of n(H) ¼ 18. The morphology transition is also accompanied by changes in the charge transfer between Pt13Hn clusters and the support. At low H coverage the Pt13 cluster is biplanar and shows a strong interaction with Al2O3. About 0.8 |e | electron transfers from the g-Al2O3 support to the cluster. However, at high H coverage, the cuboctahedral Pt13 cluster only weakly interacts with the support and there is only a small (0.2 |e |) charge transfer. The optimal hydrogen coverage and structures of Pt13 can be determined by ab initio thermodynamics16 at arbitrary temperature and pressure. Therefore, one should select for reactivity study the corresponding stable structure under reaction conditions rather than the bare Pt13 cluster at zero coverage due to their distinct geometry and electronic structure. The theoretical prediction of the work of Mager-Maury is further supported by experimental studies including X-ray near-edge adsorption17 and temperatureprogrammed desorption.18 In realistic reaction conditions, the driving force for the cluster reconstruction not only comes from the co-adsorbate as we discussed in the work of Mager-Maury15 but also comes from the reactant and the intermediates. During the catalytic reaction, reactant undergoes a series of intermediate structures, and they will compete with the co-fed adsorbate in determining the cluster structure. In the case of ethane dehydrogenation forming ethynyl, one possible pathway involves five different adsorbed intermediates (i.e. ethyl, ethylene, vinyl, acetylene and ethynyl). The interactions between intermediates and cluster, H and cluster as well as the cluster-support ones produce a very complicated reaction pathway in which the cluster catalyst demonstrates distinct structures and formula at different states of the catalytic reactions (Fig. 2).20 The reconstruction of the nanoclusters under reaction condition is even more complicated if we take the sintering of small nanoclusters from the Ostwald ripening process into account. The Ostwald ripening processes have been widely discussed in the literature for nanocluster sintering on surfaces or in solvents. Although the physical picture for Ostwald ripening is clear and qualitatively explains a wide range of phenomena, the quantitative description of the parameters for Ostwald ripening is challenging for both experimental and theoretical approaches. Small supported nanoclusters are generally not very stable compared with large nanoparticles because of their high surface energy. Several factors play a role in determining the stable cluster size

Nanocluster heterogeneous catalysts: Insights from theory

203

Fig. 1 The optimized structures of H covered Pt13 cluster. The formula is Pt13Hn. (A–F) show structures with different numbers of hydrogen atoms. (A) n ¼ 0, (B) n ¼ 6, (C) n ¼ 18, (D) n ¼ 38 (metastable structure) (E) n ¼ 20, (F) n ¼ 34.15 From Ref. Mager-Maury, C.; Bonnard, G.; Chizallet, C. et al. H2-Induced Reconstruction of Supported Pt Clusters: Metal-Support Interaction Versus Surface Hydride. ChemCatChem 2011, 3, 200–207.

and the rate of cluster disintegration or sintering, such as cluster-metal oxide and cluster-adsorbate interactions. These factors render the system complex and very challenging to control for obtaining the stability of the nanocluster catalysts. Ouyang et al. exploit the model proposed by Campbell et al.21,22 and evaluate key parameters from density functional theory calculations and complementary experiments, and finally show a simple yet effective framework to qualitatively describe the Ostwald Ripening of nano Rh clusters on a TiO2 surface.23 This method exploits ab initio thermodynamics, the Wulff construction as well as the Gibbs  Thomson equation, converting the key parameters in Ostwald ripening to data that are available from DFT simulations and experimental observations. The scheme is shown in Fig. 3.23 A metal particle of radius R is deposited on a support surface. The chemical potential of the f metal in the particle is DmNP(R). A metal monomer may dissociate from the cluster with the formation energy DEma , and its diffusion d d barrier on the support is Ema (or Ecarb when the metal monomer adsorbs CO forming carbonyls). CO adsorption can lower the chemical potential of a monomer by DGCO. Hence two key parameters are involved, one is the stability of the monomer in the presence/absence of CO environment, and the other one is the chemical potential of the metal in nanoclusters.

204

Nanocluster heterogeneous catalysts: Insights from theory

Fig. 2 Proposed reaction pathway for ethane dehydrogenation. Top: shows the schematic pathway. Middle and bottom show the top and side view of the optimal intermediate (the chemical potential of H and ethane are set by fixing partial pressure of them as 10.0 bar and temperature as 800 K).19 From Ref. Raybaud, P.; Chizallet, C.; Mager-Maury, C. et al. From G-Alumina to Supported Platinum Nanoclusters in Reforming Conditions: 10 Years of Dft Modeling and Beyond. J. Catal. 2013, 308, 328–340.

The first one is calculated as:

 h
i Em ¼ Ema=ox  Eox  EB þ Ecarb=ox  Ema=ox  n  ðECO þ DmCO ðT; pÞÞ

(1)

In Eq. (1), the term (Ema/ox  Eox  EB) computes the chemical potential of a bare single atom on the surface compared with bulk metal. The last term [Ecarb/ox  Ema/ox  n  (ECO þ DmCO(T, p))] calculates the CO adsorption free energy. Evaluating the chemical potential of supported nanoclusters is very challenging. The most straightforward method is to compute the energies of optimized nanoclusters on the supports and then to evaluate the chemical potential of the metal from the DFT

Fig. 3 A schematic illustration of the CO reactant induced Ostwald Ripening of supported metal clusters.23DmNP(R) and Dm
NP ðR Þ are the chemical potential of metals in two nanoparticles whose radius is R. DEmaf(R) is the electronic formation energy of the metal monomer on the metal oxide surface. DGCO is the formation energy of the metal monomer with CO adsorbed. Emad is the diffusion barrier. From Ref. Ouyang, R.; Liu, J.-X.; Li, W.X. Atomistic Theory of Ostwald Ripening and Disintegration of Supported Metal Particles under Reaction Conditions. J. Am. Chem. Soc. 2013, 135, 1760–1771.

Nanocluster heterogeneous catalysts: Insights from theory

205

calculated electronic energy. There are however some practical difficulties, since the DFT calculations are limited to clusters of small size and the optimum shape of the nanoclusters are not known. The work of Ouyang et al. exploits a different approach that uses the Gibbs  Thomson (G–T) relation to convert the chemical potential of the metal in the nanoclusters to a geometrical parameter R which is the radius of the semi-sphere encapsulating the supported nanoclusters.21,22

DmNP ¼

2Ugme R

(2)

In which DmNP is the chemical potential of the nanoclusters, R is the radius of the nanocluster, U is the molecular volume of bulk metal. gme is the surface energy of the metal clusters, which can be evaluated by averaging the surface energies and area ratio of the different exposed surfaces estimated by Wulff construction. In the end, the unknown parameter R can be obtained directly from the experimental measurements of the height-width ratio of Rh nanoclusters. With all data from DFT calculations and the parameter R observed from the experiments, one can compare the relative stability of small nano Rh clusters (of size 2.0 nm) and that of monomers Rh(CO) and Rh(CO)2. Fig. 4 predicts the transition of the monomer complex from Rh(CO) to Rh(CO)2, and also shows that at low temperature, the cluster should thermodynamically disintegrate and form Rh(CO)2 surface monomers. The most cumbersome parameter in the above model is the adhesion energy between metal clusters and the metal oxide support. Recent studies show that the adhesion energy is linearly correlated with the oxophilicity of the metal and the surface oxygen density,24–26 hence one can obtain the adhesion energy by interpolating a few experimental data. The model works well for clusters as small as 1 nm,24 while the atomic details of the clusters and influence of the adsorbates on the adhesion energies are still unclear.

6.09.1.2

Cluster fluxionality

The nanocluster catalyst may present a large number of low energy metastable isomers besides the global minimum structure. We have seen that small nanoclusters are influenced by environmental reactants. The interaction between nanoclusters and adsorbates changes the preferred morphology of nanoclusters. The remaining question is whether the structure of a nanocluster catalyst is well represented by only one optimal geometry at given adsorbate coverage. It is commonly acknowledged that the structure of a lowindex extended surface is generally rigid, i.e. the melting point is high. Unless the reaction temperature is very high, surface atoms only vibrate around the equilibrium geometry. Nanoclusters, however, generally present a much lower melting point than their bulk counterparts,27,28 and the catalytic reaction temperature, or even room temperature, is sufficient to populate low-energy metastable isomers. Experimental shreds of evidence also show the existence of cluster isomerization and assorted geometries that coexist for a Pt4 cluster.29 Several theoretical studies are focused on this question. H. Zhai et al. theoretically studied the isomerization pathways of Pt7 clusters on a-Al2O3(0001) surface. They find that the potential energy of Pt7 cluster is rather flat, and the isomerization barriers are very small (< 0.3 eV).10 The small isomerization barriers of cluster isomerization imply that different isomers may co-exist in the reaction system. The interaction of metal clusters and adsorbates not only changes the preferable geometry but also alters the isomerization rate. For example, X-P. Xing et al. find that CO and H2O adsorption lowers the isomerization barriers of Au6 cation clusters.30 The isomerization transforms the Au6 cluster from a triangular shape to an incomplete hexagonal shape. The transition barrier is 0.38 eV in the absence of adsorbate, while the barrier is much lower in the presence of 3 CO (0.14 eV) or with H2O

Fig. 4 The relative stability of a Rh nanocluster of diameter 20 Å, versus Rh(CO) and Rh(CO)2 complex under 0.1 mbar of CO. The temperature range is from 100 K to 1000 K. y axis is DGNPdis is the disintegration free energy which measures the thermodynamic tendency of the cluster to disintegrate into Rh complex Rh(CO) or Rh(CO)2. From Ref. Ouyang, R.; Liu, J.-X.; Li, W.-X. Atomistic Theory of Ostwald Ripening and Disintegration of Supported Metal Particles under Reaction Conditions. J. Am. Chem. Soc. 2013, 135, 1760–1771.

206

Nanocluster heterogeneous catalysts: Insights from theory

Fig. 5 The isomerization of two different Au6 cations (A) bare Au6 cation (B) Au6 cation with 3CO adsorbed, (C) Au6 cation with 3CO and 2H2O coadsorbed. The structures of transition state, the reaction energies and the activation barriers from DFT simulations are also shown. From Ref. Xing, X.; Li, X.; Yoon, B. et al. Dynamic Fluxionality and Enhanced CO Adsorption in the Presence of Coadsorbed H2O on Free Gold Cluster Cations. Int. J. Mass Spectrom. 2015, 377, 393–402.

( 0 eV) adsorbed (Fig. 5). Since the reactant is an integrated part of catalysis, the metal clusters in realistic conditions will be unavoidably covered with adsorbates. The isomerization of different clusters could be either accelerated or suppressed, therefore, it is necessary to study the cluster isomerization and the reactivity of the obtained metastable isomers in reaction conditions. The clusters with different energies show distinct stabilities and occurrence probabilities. If the isomerization of a cluster is much faster than the catalytic reaction, the relative population can be estimated by the Boltzmann distribution pi ¼ exp ( DEi/kBT). Under a high reaction temperature, the population of metastable isomers will be more significant. With the Boltzmann distribution, the occurrence probability of one metastable isomer decreases exponentially with the increasing energy. At 400  C, the population of a metastable isomer, whose energy is 0.2 eV higher than the global minimum, is only 3% of the population of the global minimum. Their small population renders the metastable structures hidden in nearly all kinds of experimental measurements. However, the metastable isomer could be very active for the catalytic reaction and finally contribute even more than the global minimum structure. One example is shown by Sun et al. using the small Pt13 cluster and methane activation.8 A modified genetic algorithm was exploited to sample the structures of hydrogen covered Pt13 clusters. The aim of the sampling is to obtain not only the global minimum structure but also the low energy isomers in a selected energy window of 0.5 eV. In the paper of Sun et al., the sampling generated a complete set of structures from the global minimum to 0.5 eV above that. Ab initio thermodynamic analysis selects Pt13H26 as the most stable chemical formula under considered reaction condition. There are 19 isomers identified in a window of 0.5 eV above the global minimum (see Fig. 6). Those Pt13H26 clusters share a similar cuboctahedral Pt13 core but have different hydrogen adsorption patterns. In particular, the surface Pt site features different H coordination numbers, such as PtH2, PtH3, PtH4, and PtH5. The PtH5 is the saturated site and can not accommodate more adsorbate, while others (PtH2, PtH3, and PtH4) are investigated for methane activation (see Fig. 6). The reaction rate constant for each cluster isomer is evaluated with the harmonic transition state theory using the DFT calculated activation barrier. However, we’ve acknowledged that the nanocluster isomers are not uniformly distributed. The more stable is the nanocluster isomer, the more occurrence probability the cluster gains. Therefore, we can estimate the population density of each isomer using Eq. (3).

Nanocluster heterogeneous catalysts: Insights from theory

207

Fig. 6 The low-energy metastable ensemble for Pt13H26 clusters. (A–C) show the structures of the first three metastable isomers. The spectrum below shows the relative energies starting from the global minimum. From Ref. Sun, G.; Sautet, P. Metastable Structures in Cluster Catalysis from First-Principles: Structural Ensemble in Reaction Conditions and Metastability Triggered Reactivity. J. Am. Chem. Soc. 2018, 140, 2812–2820.

  exp  kDBETi   pi ¼ PNLEME DE j exp  j kB T

(3)

In Eq. (3), DEi is the energy of isomer i. The denominator is used to normalize the population for all low energy metasble structures. In order to compare the contributions from two isomers to the reaction, we can calculate the effective reaction rates from Eq. (4). r ieffective ¼ riTST  pi ri

(4)

TST

is the reaction rate constant calculated from transition state theory. The CH4 activation barriers are calculated for the surface Pt sites on the first three stable isomers shown in Fig. 6.8 Eq. (4) is then used to calculate the contributions from different isomers. The results are shown in Fig. 7. The first metastable isomer of Pt13H26, which is 0.173 eV less stable than the global minimum, presents a PtH2 surface site, i.e. the Pt atom has only two hydrogen atoms. Although this low-coordinated Pt site is generally unfavored in the sense of stability, it is very active for CH4 dissociation. The activation barrier is only 0.6 eV on the PtH2 site, which is much lower than these from the global minimum (> 1.0 eV). Finally, the weighted reaction rate of the first metastable isomer is almost 2 orders of magnitude higher than that of the global minimum. The work of Sun et al. shows the general competing factors that are important for catalysis studies. The flexible geometry of nanocluster promotes the diverse active sites that may arise in the ensemble of nanocluster catalysts. Some local arrangements are not

Fig. 7 The weighted rate constants for methane activation on surface Pt sites of three low energy isomers of the Pt13H26 cluster. S indicates the surface site. GMn (n ¼ 0–2) indicates the three low energy isomers. PtHm (m ¼ 2–4) indicates the numbers of hydrogen neighbors of the considered Pt site.8 From Ref. Sun, G.; Sautet, P. Metastable Structures in Cluster Catalysis from First-Principles: Structural Ensemble in Reaction Conditions and Metastability Triggered Reactivity. J. Am. Chem. Soc. 2018, 140, 2812–2820.

208

Nanocluster heterogeneous catalysts: Insights from theory

very favorable in providing the most stable structure but very active in catalyzing the reactions. The decrease in the activation barrier speeds up the catalytic reaction in an exponential manner and can compensate for the loss in population. Therefore, the role of metastable isomers in catalytic reactions can not be safely ignored.

6.09.2

Methods

6.09.2.1

Descriptors for adsorption

There exist extensive efforts in the theoretical community to develop accurate and accessible descriptors for predicting the chemical properties of materials.31,32 Descriptors can be simple geometrical features or electronic structure parameters that can be easily calculated. The predictions using the descriptors are quantities or properties that are difficult or time-consuming to compute. Obviously, this is a very appealing approach because not only one can quickly screen a large number of materials in searching for an efficient catalyst but also the successful validation of a descriptor also underlines critical physical insights. In the field of nanocluster catalysis, the adsorption energies on diverse surface sites are key parameters to screen for the nanomaterial in catalytic reactions. The nanoclusters naturally exhibit diverse surface sites with different coordination numbers and compositions. A comprehensive study of a nanocluster for a catalytic reaction requires detailed examination of all surface sites, for the global minimum and low energy metastable structures of the cluster, even those with low-population in reaction conditions. A brute-force calculation for all reaction sites of a nanocluster with DFT is therefore expensive and not practical. Therefore, nanocluster catalysis studies demand efficient adsorption energy descriptors. The first example is the linearity between the adsorption energy and the d-band center of transition metal atoms, developed by Nørskov et al., initially for extended systems.33 In a schematic explanation, the d-band center theory considers the interaction between transition metal surface and adsorbate with a two state interaction model. The more electrons occupy the d orbital of the metal atoms or the lower the energy of the d orbital of the metal atoms, the more electrons fill the antibonding state between adsorbate and metal d orbital. Then a linear relationship between adsorbate adsorption energy and average d orbital energy (d band center) is observed across the transition metals. This method has been widely used for a range of catalyst, especially on extended surfaces.33–36 Although a transition metal nanocluster is a molecule with a finite number of d states, the number of d orbitals is generally large enough to apply the d band center theory for the specific surface sites. Tang et al. studied the dissociative O2 adsorption on the terrace sites of Pd-M alloy clusters which contains 19 heteroatomic core atoms and 70 Pd shell atoms.37 The adsorption energies of two O atoms correlate very well with the average d orbital energies of Pd atoms (Fig. 8), and in addition the same linear relation also applies to extended surface (Au, Pt, Pd) with similar adsorption geometry. For the nanocluster catalyst, the electronic structure of the component element is not the only factor that determines the reaction activity. The surface sites of nanoclusters, even mono-component ones, exhibit diverse coordination environments. The distinct site geometries also affect the adsorption energy independently to the identity of the site species. The accurate description of properties of diverse undercoordinated sites is a fundamental question for nanoclusters due to their large surface-volume ratio. The d band center theory discussed above normally assumes a consistent adsorption geometry on all considered materials, which is not applicable for diverse surface sites. Another approach is to use the generalized coordination number (GCN) developed by Calle-Vallejo et al. as descriptor of the binding site, which provides an accurate linear relationship with the adsorption energy when different undercoordinated sites are compared.38

Fig. 8 The dissociative adsorption energies of O2 on a series of single-element surfaces or core-shell alloy clusters (Pd is the shell) versus the average d orbital energies of Pd. All clusters have 79 atoms consisting of 19 core atoms and 70 shell Pt atoms.37 From Ref. Tang, W.; Henkelman, G. Charge Redistribution in Core-Shell Nanoparticles to Promote Oxygen Reduction. J. Chem. Phys. 2009, 130, 194504.

Nanocluster heterogeneous catalysts: Insights from theory

209

The GCN follows the principle of bond order conservation, which means that the unconsumed bond order of a site i (site i can be atop, bridge or hollow site) should be inversely proportional to the number of first neighbors of site i. Let us say these first neighbors are atom j (j ¼ 1 / Nj). However, not all the atom j equivalently interact with site i. The conservation of bond order for atom j are also exploited, and the bond order between atom j and site i is also inversely proportional to the number of first neighbors for atom j (which is also the second neighbors of site i, assume it is Mj). In the end, the GCN is defined as: GCN ¼

Nj X

Mj CN max j¼1

(5)

In which CNmax equals 12, 18, 22 for atop, bridge and three-fold hollow sites respectively, indicating the maximum number of second neighbor atoms in the bulk. GCN differentiates the surface sites that bear different environments when counting the second neighbors of the adsorption site,38 and improves the prediction of adsorption energy from errors of level 0.090 eV (result of conventional CN) to the level of 0.056 eV. Since the surface sites of nanoclusters contain many sites that can only be distinguished by considering the second neighbors, GCN is suitable in studying adsorption on nanoclusters (Fig. 9). The GCN method can not be used with system with heteroatoms, such as alloys. To incorporate the different ligand effects from the heteroatoms, the electronic structure for each element must be included for constructing the descriptors. One example is demonstrated by Ma et al.39 named as orbital-wise coordination number (CNa). The CNa is defined by Eq. (6). 2 Prij (111) surface (2.02 eV) > (100) surface (1.69 eV) and (110) surface (1.67 eV).5 The narrowing of the band gap for surfaces with more coordinatively unsaturated atoms decreases the energy cost for electron transfer from O to Ce. This results in lower barriers for surface oxygen vacancy formation and thus in a higher reducibility of the surface.6 It explains why ceria-based materials exposing different facets exhibit different reactivities in heterogeneous catalytic reactions.7–11 The possible transformation of one type of facet into another one due to the surface reconstruction upon thermal treatment brings another level of complexity into the surface chemistry of cerium dioxide.12 When the goal is to understand facet-dependent surface chemistry of inorganic crystal phases, it is important to develop advanced synthetic approaches that allow preferential exposure of particular facets. Significant advances in our understanding of structure-activity relationships in catalysis by ceria-based materials have been achieved in the past decades, owing to the development of shape-controlled synthesis of CeO2.13 The main insights derived from these efforts will be discussed later in this chapter. The ceria crystal size is another important parameter that can significantly influence its catalytic properties.14,15 Smaller crystallites of ceria usually contain a higher concentration of oxygen defects.16,17 Moreover, it has been observed that the reducibility of ceria also changes as a function of the particle size, with smaller NPs being more reducible than larger ones.18 In a series of comprehensive DFT studies, Neyman and co-workers investigated the quantum size-effect on the reactivity of ceria.19–21 The energy cost for oxygen vacancy formation of ceria NPs with a size in the 1–4 nm range is substantially lower than ceria NPs with typical sizes larger than 10 nm (Fig. 2). Recently, the same group combined computational modeling and a surface science approach to shed light on oxygen transport phenomena at the Pt-CeO2 interface.22 A particularly important finding is that oxygen can efficiently migrate from the ceria support to Pt clusters on nanosized CeO2, while this process is energetically unfavorable for Pt on relatively large ceria

Metal-support interfaces in ceria-based catalysts (B)

(A)

1093K Bulk oxygen

760K Surface oxygen

Vacancy formation energy (eV)

246

30 nm 9.9 nm 6.0 nm 5.3 nm 4.4 nm

4 Ce21 Ce40Ce60

900 Temperature (K)

1200

Cef

3

2

1

0 600

Ce80 Ce140

0.5

1.0

1.5

2.0

2.5

Particle size (nm)

(C)

Oxygen release Evac

Surface:

Oxygen spillover

2.25 eV

Decreasing O vacancy formation energy

+Pt

2.46 eV Pt/CeO2–x (111)

+1.0 eV O/Pt8/CeO2–x (111)

Pt8/CeO2(111)

Nanoparticle:

0.80 eV +Pt

Ce40O79

Ce40O80

–0.51 eV

1.08 eV

Pt8/Ce40O79

Pt8/Ce40O80

O/Pt8/Ce40O79

Fig. 2 Size effects in CeO2. (A) Ceria reducibility probed by hydrogen temperature-programmed reduction (H2-TPR). (B) DFT-derived energetics of oxygen vacancy formation for a series of CenO2n nanoclusters differing in size. (C) Reverse oxygen spillover from CeO2 to the Pt cluster on nanosized ceria and extended CeO2 surfaces. (A) Adapted from Xu, J.; Harmer, J.; Li, G.; Chapman, T.; Collier, P.; Longworth, S.; Tsang, S. C. Size Dependent Oxygen Buffering Capacity of Ceria Nanocrystals. Chem. Commun. 2010, 46(11), 1887–1889, https://doi.org/10.1039/b923780a; (B) Adapted from Trovarelli, A.; Llorca, J. Ceria Catalysts at Nanoscale: How Do Crystal Shapes Shape Catalysis? ACS Catal. 2017, 7(7), 4716–4735, https://doi.org/10.1021/acscatal.7b01246; (C) Data from Vayssilov, G. N.; Lykhach, Y.; Migani, A.; Staudt, T.; Petrova, G. P.; Tsud, N.; Skála, T.; Bruix, A.; Illas, F.; Prince, K. C.; Matolín, V.; Neyman, K. M.; Libuda, J. Support Nanostructure Boosts Oxygen Transfer to Catalytically Active Platinum Nanoparticles. Nat. Mater. 2011, 10(4), 310–315, https://doi.org/10.1038/nmat2976.

particles. The reverse oxygen spillover process for nanosized ceria suggested by these experiments can explain the high activity observed by metal NPs supported on nanocrystalline ceria.23,24 It has also been observed by a combination of XRD, neutron pair-distribution function (nPDF) and Raman spectroscopy that nanosized CeO2 undergoes anomalous thermal expansion below 0  C.25 Such lattice expansion has not been observed for bulk

Metal-support interfaces in ceria-based catalysts

247

ceria. The conductivity measurements together with results of DFT simulations suggested that the counter-intuitive lattice expansion upon cooling is linked to the unusual charge-transfer dynamics between oxygen vacancies and Ce 4f orbitals in nanosized ceria. So far, we have summarized several of the main structural and chemical aspects of ceria relevant to catalysis without aiming to be comprehensive. Many other peculiarities of the ceria structure have been adequately reviewed by the groups of Trovarelli26 and Fornasiero.27 Mullins presented an extensive overview of the redox properties of ceria as investigated by a classical surface science approach.3 The importance of such aspects of the surface structure, redox dynamics and the size of ceria will be further discussed throughout this chapter in the context of preparation, characterization and performance of ceria-based catalytic systems.

6.11.3

Synthesis of ceria

The most widespread use of ceria in catalysis is in automotive emissions neutralizers. Ceria-supported catalysts also display promising performance for the water-gas shift (WGS) reaction28 and solar thermochemical water splitting,29 while solid-oxide fuel cells typically contain ceria as a solid electrolyte.1 As with all catalytic supports, industrial scale production of ceria requires straightforward and reproducible synthesis recipes. Most of the current syntheses are based on wet chemistry methods. Metal loading of ceriabased catalysts is usually achieved using impregnation, resulting in the formation of (transition) metal NPs or clusters. A high metal dispersion typically requires a high surface area of the support material. Ceria with high surface area can be prepared by precipitation.30,31 This approach has the advantage that other (transition) metals can be co-precipitated, resulting in high dispersion of the added transition metal, possibly in the ceria lattice as a solid solution.32–34 Ceria-zirconia (CexZr1-xO2-d), which is the most studied mixed oxide of ceria, is typically obtained by co-precipitation. Ceria-zirconia is widely used in the automotive industry35 and in heterogeneous catalysis in general.32,33 Facile preparation, excellent thermal stability, and enhanced oxygen storage capacity with respect to ceria are the main reasons for its widespread use as a catalyst support.32,33 Using templates in combination with precipitation allows increasing the surface area of ceria. The work of Terribile et al. demonstrated that the use of an amphiphilic surfactant such as cetyltrimethylammonium bromide (CTAB) can result in an ordered mesoporous structure of cerium dioxide.36 Due to the very strong but reversible adsorption of CO2 on ceria, such porous materials with large exposed surface are promising candidates to support CO2 conversion catalysts.37 Sol-gel synthesis is another wet chemistry method widely used for the preparation of ceria-based systems with high surface area.38 A recent report of Hyeon and coworkers demonstrates the unprecedented versatility of the sol-gel route for the size- and shape-controlled synthesis of very small ceria crystallites (Fig. 3).39 Datye and co-workers showed that the sol-gel method can be also used for the preparation of highsurface-area ceria that is homogeneously doped with transition metals such as Ni, Fe, Mn, and Cu.40 Apart from the large surface area, the sol-gel method brings the advantage of its versatility in preparing thin films and coatings. This aspect is exploited in solidoxide fuel cells, in photo- and electrochemistry and in corrosion protection. The development of spectroscopic techniques and especially of atomic-resolved microscopic imaging has led to notable progress in establishing structure-function relationships of ceria-based catalysts. Using state-of-the-art imaging and spectroscopy, we are now able to resolve the surface structure of powder catalysts at the atomic level.41 Before, such profound insights could only be gained with surface science model systems. While earlier a large surface area of the support was the main target in catalyst preparation, we now understand that crystallites shape and surface termination can have a substantial impact on the overall catalytic performance.8,11,42 Such a control over the surface termination is therefore an important element of current and future investigations aimed at optimizing ceria-based catalysts. In recent years, extensive research of hydrothermal synthesis of ceria has led to many examples of shape control in the synthesis of ceria nanostructures.12,43,44 The seminal work of Li and co-workers involved a fairly simple hydrothermal synthesis of ceria nanorods predominantly exposing (110) surfaces, while conventional precipitation method yielded CeO2 NPs with dominant (111) facets.45 Although the total surface area of the obtained materials was rather similar, the CO oxidation performance was markedly different. The authors explained the higher activity of their nanorod-shaped ceria by the higher reducibility of exposed (110) surfaces as compared to (111) ones. In later work, Mai et al. showed that by changing the concentration of the precipitation agent (NaOH) and temperature of the synthesis, other shapes of ceria can be obtained (Fig. 3).46 For instance, ceria nanocubes are preferentially (100) faceted and exhibit higher reducibility than polyhedra of ceria, featuring predominantly (111) surfaces. Detailed high-resolution transmission microscopy (HR-TEM) analysis revealed that CeO2 (100) and (110) surfaces can reconstruct into more stable ones, and usually the (111) one. The combination of HR-TEM with IR spectroscopy of low-temperature adsorbed adsorption was used to unambiguously conclude that ceria nanorods are predominantly enclosed by (111) facets (Fig. 3).44 The same conclusion was drawn by Datye and co-workers based on the catalytic activity of nanorods in the WGS reaction.10 More complex morphologies, such as core-shell structures, are also within reach for ceria. Cargnello et al. demonstrated that coating/encapsulation of Pd NPs in CeO2 remarkably enhances their thermal stability and activity towards methane oxidation (Fig. 4).47 Recently, it was shown that similar structures can be prepared by a facile and direct self-assembly strategy.48 Confinement in such core-shell structures has also been exploited for hydrogenation reactions. For instance, Ni NPs encapsulated in ceria(zirconia) showed better stability and catalytic performance in CO2 methanation.49 A modular design strategy paves the way to the development of catalysts with unconventional metal-support interactions. The increased stability and activity of such modular catalysts can outweigh their higher complexity and cost of preparation. Preparation of ceria-based materials by combustion and mechanochemical approaches appears to be a promising alternative to conventional wet chemistry methods. Some of the disadvantages of wet chemistry routes are the involvement of multiple steps such

248

Metal-support interfaces in ceria-based catalysts

Fig. 3 Nano-shaped ceria crystals: (A) nanowires prepared by sol-gel methods, (B) rod- and cube-shaped ceria prepared by hydrothermal synthesis. Elucidation of the predominant faceting of CeO2 nanorods: (C) HRTEM images showing (111) surfaces and (D) IR spectra of lowtemperature CO adsorption on ceria exhibiting bands at 2152 cm 1 and 2170 cm 1related to CO on Ce4þ sites of (111) and (110) facets, respectively. (A) Adapted from Yu, T.; Joo, J.; Park, Y. il; Hyeon, T. Large-Scale Nonhydrolytic Sol-Gel Synthesis of Uniform-Sized Ceria Nanocrystals with Spherical, Wire, and Tadpole Shapes. Angew. Chem. Int. Ed. 2005, 44(45), 7411–7414, https://doi.org/10.1002/anie.200500992; (B) Adapted from Mai, H. X.; Sun, L. D.; Zhang, Y. W.; Si, R.; Feng, W.; Zhang, H. P.; Liu, H. C.; Yan, C. H. Shape-Selective Synthesis and Oxygen Storage Behavior of Ceria Nanopolyhedra, Nanorods, and Nanocubes. J. Phys. Chem. B 2005, 109(51), 24380–24385, https://doi.org/10.1021/jp055584b; (C, D) Adapted from Yang, C.; Yu, X.; Heißler, S.; Nefedov, A.; Colussi, S.; Llorca, J.; Trovarelli, A.; Wang, Y.; Wöll, C. Surface Faceting and Reconstruction of Ceria Nanoparticles. Angew. Chem. Int. Ed. 2017, 56(1), 375–379, https://doi.org/10.1002/anie.201609179.

as precipitation, washing, drying, calcination, and impregnation, and the large amount of solvent used during washing steps. Recently, the group of Trovarelli demonstrated that ball-milling of commercially available low-surface-area precursors of Ce and Pd yields highly-active and stable catalysts for methane combustion.50 The cost-efficiency and flexibility, with respect to the choice of precursors, and the simplicity of this preparation method are appealing from the perspective of scale-up. The same group also used a solution-combustion synthesis method to prepare Pd-CeO2.51 The resulting materials are PdxCe1-xO2-d solid solutions obtained in a single step. Despite their low surface area (< 20 m2/g), these catalysts displayed higher methane oxidation activity than conventional impregnated Pd/CeO2 catalysts. Another combustion method for the single-step preparation of ceria and its derivatives is flame spray pyrolysis (FSP). In contrast to the majority of solution-combustion methods, FSP allows obtaining ceria with a very large surface area ( 150– 200 m2/g). FSP is a versatile preparation method fit for scale up.52 The basic principle of FSP synthesis lies in the injection of liquid precursors containing the metals of interest into a methane-oxygen flame (Fig. 5). The temperature in the flame (typically, 1500–2000  C) can be tuned by using solvents and precursors with different enthalpies of combustion. The high temperature results in fast evaporation of the solution droplets, which shortens the residence time in the flame and, thus, limits the sintering of the resulting metal oxide and/or metal nanosized materials. The high temperature of FSP synthesis also ensures full decomposition of the precursors, which has the advantage that the resulting materials can be

Metal-support interfaces in ceria-based catalysts

249

Fig. 4 Core-shell ceria structures. (A) TEM images of Pd@CeO2 composites, (B) element analysis line profile displaying the Pd-core and CeO2-shell, (C) beneficial effect of core-shell structure on activity of composites in methane combustion. (D) Dual-confinement concept for stabilization of metaloxide interfaces. (A–C) Adapted from Cargnello, M.; Jaen, J. J. D.; Garrido, J. C. H.; Bakhmutsky, K.; Montini, T.; Gamez, J. J. C.; Gorte, R. J.; Fornasiero, P. Exceptional Activity for Methane Combustion over Modular Pd@CeO2 Subunits on Functionalized Al2O3. Science 2012, 337(6095), 713–717, https://doi.org/10.1126/science.1222887; (D) Reproduced from Peng, H.; Rao, C.; Zhang, N.; Wang, X.; Liu, W.; Mao, W.; Han, L.; Zhang, P.; Dai, S. Confined Ultrathin Pd-Ce Nanowires with Outstanding Moisture and SO2 Tolerance in Methane Combustion. Angew. Chem. Int. Ed. 2018, 57(29), 8953–8957, https://doi.org/10.1002/anie.201803393.

directly used without further drying or calcination. FSP is also ideally suited to prepare mixed metal oxides and solid solutions of two or more metals.53,54 Wet chemical methods, such as coprecipitation, often lead to inhomogeneous doping of ceria because of the different pH regions of precipitation of cerium and transition metal dopants. In FSP synthesis, salts of several transition metals can be solubilized in suitable organic solvents, contributing to a homogenous mixing of the metal ions in the resulting material. Another advantage of FSP applied to ceria preparation is that the products are highly crystalline. It is typically observed that FSP of ceria precursors leads to particles with a well-defined octahedral shape exposing predominantly the thermodynamically most stable (111) surface (Fig. 5).55 The group of Sayle has successfully developed computational tools to simulate synthesis, generating atomistic models in quantitative agreement with experimentally observed crystallite morphologies. 56 In brief, an amorphous precursor structure is melted at very high temperature ( 2000 K), followed by rapid cooling to facilitate the crystallization of the framework. The appeal of these simulations is that they allow to predict the faceting of relatively large nanostructures, for instance ceria nanorods, exposing primarily (111) facets with a small fraction of (100) ones. Simulating crystallization of ceria doped with Ti atoms led to core-shell structure of the nanocomposites, which was validated by experiment. Coupling molecular dynamics simulations with experiments provides unique insights into the mechanism of crystal growth and will stimulate the knowledge-driven design of ceria nanomaterials with controlled morphology and nanostructure.

250

Metal-support interfaces in ceria-based catalysts

Fig. 5 Flame spray pyrolysis (FSP) synthesis. (A) Schematic representation of FSP synthesis. (B)–(E) Highly crystalline (111) faceted CeO2 prepared by FSP; (F) CeO2 nanoparticle morphology predicted by molecular dynamics simulations of the annealing process. Adapted from Feng, X.; Sayle, D. C.; Wang, Z. L.; Paras, M. S.; Santora, B.; Sutorik, A. C.; Sayle, T. X. T.; Yang, Y.; Ding, Y.; Wang, X.; Her, Y.-S. Converting Ceria Polyhedral Nanoparticles into Single-Crystal Nanospheres. Science 2006, 312 (5779), 1504–1508, https://doi.org/10.1126/science.1125767.

6.11.4

Characterization of ceria materials

In recent decades, the fast development of operando spectroscopy tools has led to unprecedented insights into the chemistry of ceria. Many of the older literature is focused on understanding the role of ceria in automotive exhaust emissions.57 Thus, such reactions as CO oxidation, NO reduction, oxidation of hydrocarbons and the water-gas shift reaction are typically used to evaluate the catalytic performance of ceria-based catalysts. The surface science approach, which depends on the use of model systems such as thin films and single crystals with a well-defined surface structure, substantially contributed to understanding of metal-support interactions (MSI) in ceria-based catalysts.58 However, this research is typically limited by “pressure” and “material” gaps. The former refers to the use of characterization techniques that require high-vacuum conditions where the pressure can be up to  10 orders of magnitude lower than the pressure under real reaction conditions. Moreover, the chemistry of model systems may differ significantly from the chemistry of nano-sized powder catalysts containing defects and other imperfections, which is referred to as the “material gap”. Recent developments of spectroscopy techniques and the advancement of synchrotron radiation facilities allowed studying of catalysts in the operando regime. As such, it has already become possible to follow the dynamics of catalysts during the actual chemical reaction under close-to-practical conditions. A key element of this approach is to measure the catalytic activity and acquire spectroscopic data simultaneously, allowing the establishment of firm structure-performance relationships. Detailed accounts of the operando methodology in catalysis can be found in excellent reviews by Lukashuk and Foettinger59 and Weckhuysen.60 Traditionally, the use of operando spectroscopy is aimed at unveiling the speciation of the active components in catalysts, which are typically metal NPs

Metal-support interfaces in ceria-based catalysts

251

or clusters on a support. For ceria-based catalysts, the support can directly participate in the catalytic cycle and, thus, it is often considered a part of the active phase. Therefore, in this section we will particularly focus on the spectroscopy tools that allow following the chemistry of Ce. The most important techniques to characterize ceria-based catalysts are X-ray diffraction (XRD), Raman spectroscopy, Fouriertransformed infrared spectroscopy (FTIR), X-ray photoelectron spectroscopy (XPS), X-ray absorption spectroscopy (XAS) and transmission electron microscopy (TEM). Although almost all of these techniques are now available for in situ investigations, the majority of previous studies were carried out in an ex situ manner, implying the inherent limitation that the structure and state of catalysts were determined before and/or after the catalytic reaction. XRD is typically used to study the structure and phase purity of ceria. By using the Scherrer equation, it is possible to estimate the average crystallite size, while Rietveld refinement of XRD patterns can provide information about lattice parameters and strain. Such data are especially relevant to understand the influence of dopants on the structure of mixed oxides and solid solutions of ceria.61 In situ XRD can be used to unveil the phase transitions of the supported transition metals and also provide insight into the redox chemistry of CeO2 (Fig. 6).62 Senanayake and co-workers showed that lattice expansion of ceria under reducing conditions is related to reduction of Ce4þ to Ce3þ. The larger ionic radius of Ce3þ together with the electrostatic repulsion between oxygen vacancies and the surrounding cations lead to an expansion of the lattice, which can be followed by XRD. In this way, we can probe Ce3þ/Ce4þ redox processes under operando conditions. Moreover, using synchrotron radiation, it is possible to follow such processes with high temporal resolution. A disadvantage of this methodology, however, is the bulk sensitivity of XRD. In this respect, the use of ceria supports with relatively small particle size to maximize the number of atoms exposed to reactants seems to be a promising direction for studying redox dynamics in these systems by diffraction-based techniques. Raman spectroscopy is another scattering technique widely used to interrogate the structural and redox properties of ceria. As this technique can probe vibrational modes, it is much more sensitive to short-range order than XRD. Spectral changes related to vibrational features can be used to investigate disorder in defective nanomaterials such as ceria. As it relatively easy to simulate vibrational spectra, Raman spectroscopy is a good choice for studying the structural properties of ceria.63 The red-shift of the main F2g band ( 460 cm 1) in the Raman spectrum of ceria can be correlated to lattice strains and the presence of structural defects. In turn, the width of the band was shown to be inversely proportional to the crystallite size.64 The broad band in the range of 550-560 cm 1 is a signature of nanocrystalline ceria. This D band is related to point defects in the ceria lattice. The formation of oxygen vacancies and reduction of Ce4þ to Ce3þ are some of the reasons for the local lattice disorder, allowing the observation of vibrational modes that are normally symmetry-forbidden. Using the intensity of the D band as a proxy for oxygen vacancies concentration, it is possible to follow the redox behavior of the support in an operando manner. As Raman is a vibrational spectroscopy, it is also very sensitive to isotope exchange, which opens up possibilities for studying the oxygen mobility in ceria lattice using labelled oxygen atoms. Recently, Schilling et al. elegantly proved a pivotal role of oxygen of the support in the WGS reaction over gold-ceria catalysts using isotopic labelling and operando Raman spectroscopy.65 An important benefit of Raman spectroscopy for

Fig. 6 In situ XRD of Co-CeO2 catalyst. (A) The XRD patterns resolve cobalt oxide and ceria reduction features; (B) ceria lattice expansion due to Ce3þ formation upon reduction. Adapted from Zhang, F.; Liu, Z.; Zhang, S.; Akter, N.; Palomino, R. M.; Vovchok, D.; Orozco, I.; Salazar, D.; Rodriguez, J. A.; Llorca, J.; Lee, J.; Kim, D. H.; Xu, W.; Frenkel, A. I.; Li, Y.; Kim, T.; Senanayake, S. D. In Situ Elucidation of the Active State of CoCeOx Catalysts in the Dry Reforming of Methane: The Important Role of the Reducible Oxide Support and Interactions with Cobalt. ACS Catal. 2018, 8(4), 3550–3560, https://doi.org/10.1021/acscatal.7b03640.

252

Metal-support interfaces in ceria-based catalysts

ceria characterization is the ability to detect peroxo (O22, 820–890 cm 1) and superoxo (O2, 1120-1140 cm 1) species, which are vital intermediates in various oxidation reactions.66 One of the disadvantages is its strong dependence on the optical absorption properties of the material, which can often change as a function of the reaction conditions. Irradiation damage by the laser source is another issue that can complicate data interpretation in Raman spectroscopy. Further information on the application of Raman spectroscopy for investigation of ceria-based catalysts can be found in a recent review by Loridant.67 FTIR spectroscopy is one of the key techniques to characterize catalytic materials. In the context of heterogeneous catalysts, it is mostly applied in combination with probe molecules. For example, IR analysis of chemisorbed CO is a widely used method for studying catalytic surfaces. The vibrational frequency of adsorbed CO is sensitive to the geometric and electronic structure of the adsorption site. CO adsorption FTIR spectroscopy can be considered an operando technique, if CO is one of the reactants or products of the reaction. In some cases, it is preferred to avoid reactions of CO, which can be achieved by cooling to very low temperatures. The low CO adsorption strength on surface sites of metal-oxides, such as ceria, is another reason for performing low-temperature CO adsorption. The group of Wöll studied CO adsorption on different surfaces of ceria single crystals and shaped nanocrystallites.44 They concluded that IR spectroscopy of low-temperature CO adsorption is a suitable method to distinguish between the most relevant facets of ceria, i.e. (111), (110) and (110), based on the frequency of the observed carbonyl IR bands. The position of these IR bands also depends on the oxidation state of Ce atoms. This approach led to unambiguous determination of the primary faceting of nano-shaped ceria crystallites. IR spectroscopy of methanol adsorption is a less common method, but it is also sensitive to ceria surface termination.68 The C-H and C-O stretching frequencies of adsorbed methanol molecules can also be used to determine the surface structure of CeO2 (Fig. 7). It should be noted that the ceria surface can be reduced by methanol, even at room temperature. IR spectroscopy of adsorbed probe molecules is a powerful methodology which, however, does not directly probe the electronic structure and redox properties of ceria. The O-H region of the ceria IR spectrum can also be used to follow surface reduction/ oxidation phenomena, although the interpretation is not very straightforward.69 A direct probe for Ce3þ states is the broad band at  2120 cm 1, which is due to 2F5/2 / 2F7/2 electronic transition. Recently, Parastaev et al. successfully used this IR signature to study the Ce3þ/Ce4þ redox dynamics of cobalt-ceria-zirconia interface.70 Yet, the weak intensity of the adsorption band and its appearance only at high ceria reduction degrees significantly limits the use of this spectroscopic method. Further details into the characterization of ceria-based materials by IR spectroscopy can be found elsewhere.69 XPS has become a very valuable technique for studying heterogenous catalysts. The importance of XPS to heterogenous catalysis arises from its surface sensitivity (typically 2-10 nm). Using synchrotron radiation, it is now possible to vary the sample probing depth (from tens to fractions of nm) by tuning the incident photon energy. The photoelectron lines in XPS spectra are specific to the elements and can thus be employed for elemental analysis, while the use of appropriate relative sensitivity factors allows quantifying the surface composition of the sample. Analysis of the binding energy of the core-line electrons can be used to identify the valence state of the surface atoms. This is especially relevant for ceria-based systems, where the redox Ce3þ/Ce4þ dynamics are of particular interest. Conventional XPS analysis is done in ultra-high vacuum (UHV), required by the technical considerations and to minimize the attenuation of electrons escaping the solid. Under such conditions the samples can be characterized ex situ. An often employed approach is to characterize samples quasi in situ, referring to a particular pre-treatment of a sample followed by transfer to

Fig. 7 IR probes for the surface termination and electronic state of ceria. (A) C-H and C-O regions of methoxy groups IR spectrum over nanoshaped ceria. (B) Ce3þ formation, as manifested by the 2115 cm 1 band (2F5/2 / 2F7/2) during the reduction of cobalt-ceria-zirconia catalyst. (A) Reproduced from Wu, Z.; Li, M.; Mullins, D. R.; Overbury, S. H. Probing the Surface Sites of CeO2 Nanocrystals with Well-Defined Surface Planes Via Methanol Adsorption and Desorption. ACS Catal. 2012, 2(11), 2224–2234, https://doi.org/10.1021/cs300467p; (B) Adapted from Parastaev, A.; Muravev, V.; Huertas Osta, E.; van Hoof, A. J. F.; Kimpel, T. F.; Kosinov, N.; Hensen, E. J. M. Boosting CO2 Hydrogenation Via Size-Dependent Metal–Support Interactions in Cobalt/Ceria-Based Catalysts. Nat. Catal. 2020, 1–8, https://doi.org/10.1038/s41929-020-0459-4.

Metal-support interfaces in ceria-based catalysts

253

the UHV analysis chamber of the spectrometer without exposure to the ambient atmosphere. Although such approaches are an improvement over conventional ex situ analysis, the characterization of ceria in this way remains problematic, especially for the study of Ce redox dynamics. A main reason for this is that Ce3þ can be readily oxidized at relatively low partial pressure of oxygen. Another issue relates to the reduction of Ce4þ to Ce3þ upon prolonged exposure to UHV conditions and especially under X-ray irradiation, resulting in erroneous determination of the surface valence state.71 Recent advances in the development of in situ nearambient pressure XPS (NAP-XPS) or ambient pressure XPS (AP-XPS) resolved such challenges to a large extent.72 Modern NAP-XPS systems allow studying the catalysts at pressures of up to 100 mbar, which is very close to real conditions of many important catalytic reactions. The combination of high flux and tunable energy of the excitation radiation available at synchrotrons makes NAP-XPS an ultimate tool for studying surface redox dynamics in ceria-based materials. Bluhm and co-workers73 have recently analyzed the Ce 3d core-line spectra with the aim of following the Ce valence as a function of applied potential in a working solid-oxide electrochemical cell (Fig. 8). This example is very illustrative of how dramatic the changes in surface Ce3þ/Ce4þ ratio can be due to the applied (electro)chemical conditions. By means of lab-based NAP-XPS study, the group of Hensen demonstrated oxygen spillover effects from and to the support at cobalt-ceria-zirconia interface.70 Kato et al. obtained another important insight into the chemistry of Ce3þ and oxygen vacancies.74 Using a tunable photon energy, the authors showed that catalytically relevant Ce3þ ions and oxygen vacancies reside in the top-most surface layers of Pt-CeO2 catalysts. This study underpins the vital importance of (i) an in situ approach to follow the Ce3þ/Ce4þ dynamics and (ii) the surface sensitivity for accurate probing of the catalytically relevant charge states in ceria. With respect to the latter, the use of resonance photoelectron spectroscopy (RPES) of Ce 4d / 4f transitions to quantify the Ce3þ states at the very surface, appears to be very promising.75 Yet, this approach has not been used at near-ambient pressure and on realistic samples. Hard X-ray absorption spectroscopy (XAS) is a synchrotron-based technique, which provides insights into element-specific electronic and atomic structure of the materials. XAS has become an important analytical tool in materials chemistry and, in particular, in catalysis studies. High sensitivity (with some fluorescence detectors possibly down to sub-ppm concentrations) and flexibility with respect to pressure and temperature are the key reasons for its popularity in heterogeneous catalysis. Although the principles of XAS remain the same regardless of the incident photon energy, it is convenient to separate two types of X-ray absorption using soft (< 5 keV) and hard (> 10 keV) X-rays. With high-energy X-rays, it is feasible to perform operando spectroscopic studies (almost) without any compromises with regards to reaction conditions. X-ray absorption near edge structure (XANES) spectroscopy allows following the electronic structure of the element (e.g., valence state) as function of time, reaction conditions and other stimuli. Another benefit of hard X-rays is the possibility to obtain information about the local structure around the absorbing atoms, resulting from the photoelectron scattering by the neighboring atoms. This spectroscopy method is called extended X-ray absorption fine structure (EXAFS). After proper data reduction, local structural parameters of the absorbers, including the nature of neighboring

Fig. 8 In situ NAP-XPS data. (A) Ce 3d core-line spectra exhibiting different Ce3þ fractions at the surface of the operating solid-oxide electrochemical cell as a function of applied bias. (B) Spatial distribution of Ce3þ as a function of probing depth – majority of the oxygen vacancies reside on the very surface. (A) Adapted from Zhang, C.; Grass, M. E.; McDaniel, A. H.; Decaluwe, S. C.; El Gabaly, F.; Liu, Z.; McCarty, K. F.; Farrow, R. L.; Linne, M. A.; Hussain, Z.; Jackson, G. S.; Bluhm, H.; Eichhorn, B. W. Measuring Fundamental Properties in Operating Solid Oxide Electrochemical Cells by Using In Situ X-Ray Photoelectron Spectroscopy. Nat. Mater. 2010, 9(11), 944–949, https://doi.org/10.1038/nmat2851; (B) Adapted from Kato, S.; Ammann, M.; Huthwelker, T.; Paun, C.; Lampimäki, M.; Lee, M.-T.; Rothensteiner, M.; van Bokhoven, J. A. Quantitative Depth Profiling of Ce3þ in Pt/CeO2 by In Situ High-Energy XPS in a Hydrogen Atmosphere. Phys. Chem. Chem. Phys. 2015, 17(7), 5078–5083, https://doi. org/10.1039/C4CP05643D.

254

Metal-support interfaces in ceria-based catalysts

atoms, their number and interatomic distances, can be resolved. For structural studies of ceria-based materials, the Ce K-edge ( 40 keV) is commonly used. It is also very useful to use XANES to directly probe Ce3þ/Ce4þ redox transformations and interrogate the electronic structure of Ce. Typically, this is done at the Ce L3-edge ( 5.7 keV), which is sensitive to the valence state of Ce due to dipole allowed 2p-5d transitions. The fine structure of the XANES region can be obtained by using high-energy resolution fluorescence detection (HERFD). This technique allows reaching an energy resolution below the one dictated by the core-hole lifetime broadening. Such high spectral resolution is especially valuable for studying CeO2 NPs, whose electronic structure is markedly different from their bulk counterparts.76 Better signal quality comes, however, at a cost, because HERFD-XAS requires energydispersive detectors or crystal analyzers, complicated in use and not so often available at XAS beamlines. Therefore, conventional Ce L3 XANES still remains a workhorse for studying oxidation-reduction processes in ceria. Using this approach, Ganzler et al.77 demonstrated that nano-sized ceria in Pt-CeO2-Al2O3 composites undergoes reduction by CO already at temperatures as low as 100  C. Such low-temperature lattice oxygen removal provides strong evidence for the high oxygen mobility in nanosized ceria (Fig. 9). The small size of CeO2 NPs used in this study ( 5 nm) is particularly suitable due to the high surface-to-bulk atom ratio, since XAS employed in normal incidence geometry is a bulk-sensitive technique. The size-dependent reducibility of ceria NPs was also validated by Ce L3 XANES.78 By following the formation of Ce3þ as a function of temperature, the authors showed that small NPs of ceria can be reduced by CO at a lower temperature than their bulkier counterparts. Another effective technique to follow minute changes in the population of Ce3þ states is resonant X-ray emission spectroscopy (RXES). Under optimized conditions, Safonova and co-workers pushed the sensitivity limit down to 0.3% of Ce3þ while operating at sub-second time resolution (Fig. 10).79 X-ray emission techniques require rather complicated analyzers and a high-flux of incident photons, in return providing exceptional sensitivity to the changes in valence states with high temporal resolution. Electron microscopy has evolved into a very powerful tool for imaging the nanoscale properties and the combination with other spectroscopy tools provides very valuable insights into local structure including its physico-chemical aspects. With the development of aberration-corrected electron microscopes, exact atomic arrangements at the surface of materials can be visualized. This kind of knowledge is indispensable for developing structure-activity relationships for ceria-based catalysts. Using atomically resolved transmission electron microscopy, Wöll and colleagues demonstrated that calcination of ceria nano-cubes leads to significant surface reconstructions12 (Fig. 11). Re-faceting of (100) and (110) surfaces towards thermodynamically more stable (111) surfaces upon calcination caused the experimentally observed increase of CO oxidation activity of annealed samples. Impressive possibilities in atomic scale imaging with elemental and valence state analysis can be derived by combining EM with electron energy loss spectroscopy (EELS). Current aberration-corrected TEM/STEM techniques allow imaging of isolated single atoms as well as acquiring spectroscopic information from a single atomic column. The valence states of Ce atoms in ceria nano-structures can be spatially resolved with close-to-atomic precision. Van Tendeloo and co-workers successfully employed STEM-EELS for mapping the Ce3þ and Ce4þ states in ceria NPs (Fig. 12), using the Ce M4,5 edges as a fingerprint.80 They found that (100) surfaces are substantially richer in oxygen vacancies (up to  6 atomic layers deep) as reflected by a higher Ce3þ/Ce4þ ratio than (111) facets where only 1-2 underlying atomic layers are of mixed valence state. These findings are in line with the DFT-predicted lower barriers for oxygen vacancy formation and the higher reducibility of the (100) surface in comparison to the (111) one. Pioneering EM work of Takeda and colleagues81 revealed the dynamic reconstruction of ceria-supported Au NPs in the presence of CO and O2 already at room temperature (Fig. 13). By adapting the energy of incident electron beam, these researchers could observe CO adsorbed on the reconstructed surface of gold nanoparticles. Even such level of detail, however, did not allow to unambiguously answer the important issue of identifying the metal-support interface or metallic sites as the active sites for lowtemperature CO oxidation. For further details on the latest advancements in microscopy imaging of ceria-based materials, the reader is referred to an excellent review by the group of Van Tendeloo.41

Fig. 9 Following the reduction of Ce4þ to Ce3þ by CO using in situ XANES at Ce L3-edge. Adapted from Gänzler, A. M.; Casapu, M.; Maurer, F.; Störmer, H.; Gerthsen, D.; Ferré, G.; Vernoux, P.; Bornmann, B.; Frahm, R.; Murzin, V.; Nachtegaal, M.; Votsmeier, M.; Grunwaldt, J. D. Tuning the Pt/CeO2 Interface by In Situ Variation of the Pt Particle Size. ACS Catal. 2018, 8(6), 4800–4811, https://doi.org/10.1021/acscatal.8b00330.

Metal-support interfaces in ceria-based catalysts

255

Fig. 10 Schematic representation of the approach by Safonova and co-workers for time-resolved Ce3þ detection using RXES. Reproduced from Kopelent, R.; van Bokhoven, J. A.; Nachtegaal, M.; Szlachetko, J.; Safonova, O. v. X-Ray Emission Spectroscopy: Highly Sensitive Techniques for Time-Resolved Probing of Cerium Reactivity under Catalytic Conditions. Phys. Chem. Chem. Phys. 2016, 18(47), 32486–32493, https://doi.org/10. 1039/c6cp05830b.

Fig. 11 Surface refaceting of ceria cubic nanostructures induced by the thermal treatment. Reproduced from Yang, C.; Capdevila-Cortada, M.; Dong, C.; Zhou, Y.; Wang, J.; Yu, X.; Nefedov, A.; Heißler, S.; López, N.; Shen, W.; Wöll, C.; Wang, Y. Surface Refaceting Mechanism on Cubic Ceria. J. Phys. Chem. Lett. 2020, 11(18), 7925–7931, https://doi.org/10.1021/acs.jpclett.0c02409.

For completeness, we mention some of the other frequently used tools to characterize the electronic and structural properties of ceria. UV–Vis spectroscopy can be applied to determine the band gap structure and the Ce3þ to Ce4þ charge transfer transitions ( 570 nm) in the operando regime.82 Yet, direct quantification of Ce3þ states cannot be performed in this manner. In electron paramagnetic resonance (EPR), a characteristic line with g  1.97 has been used to quantify Ce3þ states (or oxygen vacancies). However,

256

Metal-support interfaces in ceria-based catalysts

Fig. 12 STEM-EELS color map of CeO2 NPs with Ce3þ (red) and Ce4þ (green). (A) Overview of the entire nanoparticle. (B) HAADF-STEM image showing the (100) surface truncation and the scan region indicated. (C) Averaged Ce M4,5 spectra from regions A-H together with reference spectra for Ce3þ and Ce4þ. The fitted weight of Ce3þ is given for each spectrum. High resolution HAADF-STEM image showing the (D) (111) surface and the EELS scan region. (E) Ce4þ map. (F) Ce3þ map. (G) Color map with Ce3þ (red) and Ce4þ (green). Reproduced from Zhang, Y.; Bals, S.; van Tendeloo, G. Understanding CeO2-Based Nanostructures through Advanced Electron Microscopy in 2D and 3D. Part. Part. Syst. Charact. 2019, 36(1), 1800287, https://doi.org/10.1002/ppsc.201800287.

Fig. 13 TEM images of Au supported on CeO2 in (A) vacuum and (B) reaction environment (1 vol% CO in air gas mixture at 45 Pa at room temperature). The (100) facets are indicated by areas I and II. Enlarged images of these regions in vacuum and in the CO in air gas mixture are shown at the bottom. Adapted from Yoshida, H.; Kuwauchi, Y.; Jinschek, J. R.; Sun, K.; Tanaka, S.; Kohyama, M.; Shimada, S.; Haruta, M.; Takeda, S. Visualizing Gas Molecules Interacting with Supported Nanoparticulate Catalysts at Reaction Conditions. Science 2012, 335(6066), 317–319, https:// doi.org/10.1126/science.1213194.

a recent report of Rakhmatullin et al.83 suggests that this assignment is a misinterpretation, i.e., the EPR signal is rather due to an electron trapped near the surface Ce3þ/Ce4þ redox pair. Solid-state NMR using 17O as a probe can be considered a potent method for studying the surface structure and defect chemistry of ceria NPs.84 The high cost associated with enriching ceria with 17O in

Metal-support interfaces in ceria-based catalysts

257

gaseous 17O2 limits its wide application for studying catalytic systems, although isotopic exchange with H217O appears to be a cheaper alternative. Despite the significant advances in the characterization of ceria, there are many remaining challenges in understanding the chemistry of these unique materials. In the next sections we will discuss how the toolbox of advanced synthesis techniques together with the state-of-the-art spectroscopy and imaging can help us to understand and improve the catalytic functions of ceria-based materials.

6.11.5

Metal-support interfaces in ceria catalysts

Nanoparticles of transition metals, mostly of the platinum group, dispersed over ceria-containing support materials are widely used as catalysts in car exhaust neutralizers. The strong interactions between the metal and the ceria support can result in a high metal dispersion during preparation and the use of the catalyst. In some cases, the metal-support interaction (MSI) is so strong that the supported metals can be atomically dispersed, giving rise to a new class of materials – single-atom catalysts, which will be the subject of the next section. Decades after the first reports about strong MSI,85 there remains a lack of detailed atomistic understanding about the origin and nature of these interactions. The challenge to comprehend MSI arises from the difficulty to disentangle its chemical origins from the many manifestations. Such ambiguities can be overcome by establishing firm structure-performance relationships. The latest developments in spectroscopy and imaging hold promise to resolve many of the underlying physico-chemical aspects of MSI. In a recent review, De Jong and colleagues discussed the main concepts underlying the chemistry of MSI.86 Applied to ceria-based catalysts, the morphology of the support and metal NPs, charge transfer and spillover effects are identified as the most important consequences of MSI. These interactions at the metal-support interface cause the interfacial sites (metal atoms in close proximity to the supporting oxide) to be distinctly different from other surface atoms of the supported NPs. The charge transfer between metal and support (known as electronic MSI or EMSI) causes the perturbations of the electronic structure of atoms at the metal–oxide interface.87 Such interactions can give rise to a new type of sites, whose catalytic properties are markedly different from conventional metallic sites. This was for instance shown for the WGS reaction by Bruix et al.88 Following the classical surface science approach, the authors engineered model Pt-CeO2 interfaces, which are substantially more active than extended Pt surfaces or conventional Cu/ ZnO catalysts. Detailed photoemission spectroscopy study revealed that the valence state of ceria-supported Pt clusters is different from bulk Pt metal, which is caused by the charge transfer from Pt to Ce. In a seminal work, Lykhach and coworkers quantified these charge-transfer phenomena using Pt-CeO2 as a model system.89 They first prepared an (111) oriented thin film of CeO2 on which the Pt atoms were deposited at a controlled rate. Tracking the valence states of Ce and Pt atoms as a function of Pt coverage using surface-sensitive RPES and XPS, it was possible to count the number of electrons transferred from Pt0 to Ce4þ atoms. As follows from Fig. 14, the size of the deposited Pt clusters strongly influences the extent of the EMSI, the smallest ones (< 2 nm) being the most sensitive to charge redistribution. Similar charge-transfer phenomena were also observed for Cu clusters deposited on ceria.90 These results pave the way to understanding of the size-induced effects in structure-property relationships of metal NPs supported on ceria. Spillover effects are another class of phenomena occurring at the interface between metals and the reducible ceria support. Conceptually, the term spillover implies the process in which certain species are activated on the metal surface and then transferred to the support, which is typically a metal oxide. Despite being already observed in the 1960s,91 the actual understanding of hydrogen spillover is far from complete.92 Van Bokhoven and colleagues used an elegant state-of-the-art in situ X-ray photoemission electron microscopy (X-PEEM) approach to follow the hydrogen spillover from metal to metal oxides.93 The authors first prepared a suitable model system, employing electron beam lithography to produce well-defined NPs of platinum and iron oxide on typical reducible (titania) and non-reducible (alumina) supports (Fig. 15). By varying with nanometer precision the distance between nanoparticles of Pt and FeOx, hydrogen spillover could be monitored by studying the reduction of iron oxide. When supported on alumina, only the reduction of iron oxide NPs in close proximity (< 15 nm) to Pt was observed. On the contrary, when supported on titanium oxide, the reduction of iron oxide occurred to the same extent regardless of the distance from Pt. This is a clear illustration of the prominent hydrogen spillover effect on reducible oxides. Lykhach et al. demonstrated the relevance of hydrogen spillover on ceria as well, detecting Ce3þ formation at the surface of Pt-CeO2 model system when exposed to hydrogen.94 Facile hydrogen spillover is an important property of ceria-based catalysts, particularly relevant for hydrogenation reactions, e.g., CO2 methanation. In principle, not only hydrogen but also oxygen atoms can be spilled over to the support, as was demonstrated recently for cobalt-ceria-zirconia catalysts.70 This is likely caused by the participation of the cobalt-ceria interface in CO2 and CO dissociation. The reverse spillover process can also occur, entailing that active species are transferred from the support (reducible metal oxide) to the interfacial sites of the deposited metal. Typically, this concerns oxygen atoms leaving the ceria lattice positions with the formation of oxygen vacancies and adjunct Ce3þ cations. Key examples of reverse oxygen spillover were captured in a surface science study of Gorte and Zafiris,95 who demonstrated the transfer of oxygen atoms from ceria to Rh NPs. Using CO adsorption and temperature-programmed desorption (TPD) experiments, they found that a substantial fraction of adsorbed CO molecules leaves the surface as CO2, due to oxidation by interfacial oxygen atoms. Later, Smirnov and Graham96 showed by XPS that heating in UHV conditions leads to oxidation of Pd clusters on the surface of ceria due to facile oxygen transport from the support to metal species. Finally, Neyman and colleagues22 directly observed reverse oxygen spillover in model Pt-CeO2 nanocomposites (Fig. 16). The authors first showed that in the absence of a metal-support interface (i.e., without deposited Pt NPs) reduction of Ce4þ to Ce3þ (as a proxy for oxygen vacancies formation) did not occur upon heating the model catalyst in vacuum. However, by depositing Pt clusters on a thin film of ceria, the redox properties were dramatically changed. Thermal treatment of the

258

Metal-support interfaces in ceria-based catalysts

Fig. 14 Charge-transfer phenomena at Pt-CeO2 interface. (A) The number of electrons transferred per Pt particle to the ceria support increases with increasing particle size (green squares). The partial charge per Pt atom reaches a maximum for particles with 30–70 atoms (yellow circles). (B) At higher Pt coverage the total amount of transferred charge approaches a limit – “charge transfer limit” (red squares). The atomic models show schematically the average particle sizes in the different regions. Reproduced from Lykhach, Y.; Kozlov, S. M.; Skála, T.; Tovt, A.; Stetsovych, V.; Tsud, N.; Dvorák, F.; Johánek, V.; Neitzel, A.; Myslivecek, J.; Fabris, S.; Matolín, V.; Neyman, K. M.; Libuda, J. Counting Electrons on Supported Nanoparticles. Nat. Mater. 2016, 15(3), 284–288, https://doi.org/10.1038/nmat4500.

Pt-CeO2 (111) composite led to Ce3þ formation, as monitored using high surface and valence state sensitivity of RPES. Their experimental findings were rationalized by DFT calculations, pointing at nanosize-dependent oxygen mobility (via reverse spillover). These spillover effects significantly broaden the scope of mechanistic steps that has to be considered in catalytic cycles over ceria-based materials. Understanding spillover effects and other phenomena underlying the MSI is pivotal for knowledge-driven design of novel interfaces with tailored catalytic properties.

Metal-support interfaces in ceria-based catalysts

259

Fig. 15 Hydrogen spillover. (A) Scheme of hydrogen spillover from platinum to an iron oxide particle over a titanium oxide or aluminum oxide support. (B) SEM image (left) and X-PEEM (right) of the nanofabricated model system on an aluminum oxides support, with 15 such pairs and a single iron oxide particle over an area of 3  3 mm2 (scale bar, 500 nm). The iron oxide particles have a diameter of 60 nm, while the platinum particles have a diameter of 30 nm. The distance between the pairs is 1 mm. Higher-magnification SEM images (scale bars, 25 nm) of pairs ‘d1’ and ‘a1,’ encircled in green and red, respectively. Adapted from Karim, W.; Spreafico, C.; Kleibert, A.; Gobrecht, J.; VandeVondele, J.; Ekinci, Y.; van Bokhoven, J. A. Catalyst Support Effects on Hydrogen Spillover. Nature 2017, 541(7635), 68–71, https://doi.org/10.1038/nature20782.

Next, we will discuss some illustrative examples of catalytic reactions, where the chemistry is dominated by the MSI phenomena at the interface between metal NPs and ceria. A prominent example is found in the work of Cargnello et al., who employed colloidal synthesis to prepare a set of well-defined ceria-supported Pd, Pt and Ni NPs of different size for CO oxidation (Fig. 18).97 The advantage of this modular synthesis approach is the uniformity of the deposited particles, which usually cannot be achieved by conventional methods such as impregnation. A specific feature of metal NPs supported on ceria is that they are already active in CO oxidation at much lower temperature than metal NPs supported on non-reducible supports. Extensive HR-TEM analysis and careful kinetic studies provided insight into the relation between the size of the metal NPs and the CO oxidation activity. Specifically, the authors concluded that interfacial metal atoms – despite not necessarily being metallic – are the most active sites for lowtemperature CO oxidation, which can be explained by the active participation of oxygen of the ceria support via a Mars-van Krevelen-type of mechanism. The normalized reaction rates and apparent activation energies for ceria-supported Pt, Pd and Ni NPs were very similar, yet markedly different from those of metal NPs supported on alumina. This study revealed the crucial role of metal-support interface in catalysis by ceria systems. Another good example of interfacial synergy between metal NPs and ceria was provided by Chen et al.98 By depositing well-defined Pd NPs on the surface of ceria nanorods, they found that the catalytic activity in CO oxidation is substantially higher as compared to the activities of the separate components. The synergetic effect was explained by electron charge transfer at the metal-support interface, which was demonstrated by Ce L3-edge XAS. Kopelent et al. used time-resolved RXES to directly prove the involvement of ceria in CO oxidation at low temperatures.99 RXES was used to follow the Ce3þ/Ce4þ dynamics as a function of transient switches between feeds of CO þ O2 and CO over impregnated catalysts comprising small Pt NPs on CeO2. The initial rate of Ce3þ formation (upon a switch to CO) is markedly slower than the subsequent reoxidation to Ce4þ, when O2 is co-fed with CO again. The working hypothesis was that CO oxidation occurs at the Pt-CeO2 interface, in which Pt atoms serve as CO adsorption sites, while O2 molecules are activated at the oxygen vacancies transiently formed in the reaction cycle. The Ce4þ-to-Ce3þ reduction due to the reverse oxygen spillover to Pt sites was concluded to be a rate-limiting step in the overall process. This mechanistic proposal was further corroborated by the similarity between the steady-state apparent activation energy and the one derived from the transient experiments using Ce reduction kinetics. As such, this work forms an excellent example of how advanced spectroscopy with temporal resolution, comparable to the rate of elementary reaction steps, can resolve key mechanistic details governing the overall kinetics. The group of Shen demonstrated the relevance of interfacial sites in copperceria catalysts for the low-temperature WGS reaction.100 Using state-of-the-art microscopy and IR spectroscopy, they concluded that

260

Metal-support interfaces in ceria-based catalysts

Fig. 16 Reverse oxygen spillover effect. (A) Schematics of Pt–CeO2/Cu(111) model catalyst (left). Reference measurements are made on a pure CeO2 film on Cu(111) (right). (B) Changes in the ceria oxidation state monitored by RPES. Three selected situations are shown: (i) the pristine CeO2 film; (ii) after deposition of Pt at 300 K; (iii) after subsequent annealing to 700 K. The insets show representative scanning tunneling microscopy images of the model systems. (C) The resonant enhancement ratio from RPES, reflecting the Ce3þ concentration, reveals two individual processes: Spontaneous Ce3þ formation on Pt deposition at 300 Kdattributed to a purely electronic MO interactiondand an abundant further formation of Ce3þ on annealing above 500 Kdattributed to the activated process of oxygen reverse spillover. The right axis shows the surface stoichiometry of CeO2  x as determined from the RER. In the absence of Pt, no reduction is observed. Adapted from Vayssilov, G. N.; Lykhach, Y.; Migani, A.; Staudt, T.; Petrova, G. P.; Tsud, N.; Skála, T.; Bruix, A.; Illas, F.; Prince, K. C.; Matolín, V.; Neyman, K. M.; Libuda, J. Support Nanostructure Boosts Oxygen Transfer to Catalytically Active Platinum Nanoparticles. Nat. Mater. 2011, 10(4), 310–315, https://doi.org/10.1038/nmat2976.

the reaction takes place at the perimeter of the Cu clusters (Fig. 17). An important finding was that the actual interfacial copper sites are not metallic, but positively charged Cuþ atoms. These atoms provide adsorption sites for CO molecules that can further react with OH groups situated on ceria to form CO2 and hydrogen. Adsorbed water then dissociates at the oxygen vacancies, which are formed as a result of hydrogen spillover, and regenerates the OH groups closing the catalytic cycle. The ability of ceria to maintain a high dispersion of metal NPs in comparison to irreducible (stoichiometric) oxides is a wellknown manifestation of strong MSI. Despite its prominence, it has only recently been explained in a quantitative manner. Campbell and Farmer101 used calorimetric measurements to demonstrate that the sintering resistance of silver NPs is strongly dependent on their size with the smallest ones (< 2 nm) being identified as the most stable on a ceria surface rich in oxygen vacancies. The large adhesion energy of smaller silver NPs to ceria was explained by lattice strain effects caused by strong metal-oxide bonding. Later, this explanation was generalized for other oxide supports and the observed trend suggests that higher adhesion energies of a metal are also linked to a higher heat of formation of the corresponding metal oxide.102 The ability of ceria to stabilize highly dispersed metal NPs under transiently changing reaction conditions is especially important for the catalysis in automotive industry. Using state-ofthe-art time-resolved operando XAS, Nagai et al.103 demonstrated that the size of the Pt NPs supported on ceria-zirconia can change as a function of reaction conditions. They observed that agglomeration of Pt under reducing conditions can be reversed through redispersion during the oxidizing cycle. Such self-regenerative behavior can explain the higher stability and activity of ceria-supported catalysts. When Pt is supported on alumina, it sinters during reduction-oxidation treatments, leading to catalyst deactivation. Cargnello and co-workers104 used mild oxidative treatments of ceria-supported Ru NPs to tune the selectivity in CO2 hydrogenation. Ru NPs can be dispersed into single-atom species through such an oxidative pretreatment. This transformation

Metal-support interfaces in ceria-based catalysts

261

Fig. 17 Catalysis by metal-support interfaces. (A) HR-TEM CeO2 supported Pd NPs; (B) intrinsic CO oxidation activity as a function of metal particle size and relation to the number of interfacial sites. (C) The atom-resolved HAADF-STEM images of copper clusters on ceria rods show the atomic configurations of copper domains, monolayers and bilayers; (D) a schematic illustration of the Cu bilayer on ceria. (A, B) Adapted from Cargnello, M.; Doan-Nguyen, V. V. T.; Gordon, T. R.; Diaz, R. E.; Stach, E. A.; Gorte, R. J.; Fornasiero, P.; Murray, C. B. Control of Metal Nanocrystal Size Reveals Metal-Support Interface Role for Ceria Catalysts. Science 2013, 341(6147), 771–773, https://doi.org/10.1126/science.1240148; (C, D) Adapted from Chen, A.; Yu, X.; Zhou, Y.; Miao, S.; Li, Y.; Kuld, S.; Sehested, J.; Liu, J.; Aoki, T.; Hong, S.; Camellone, M. F.; Fabris, S.; Ning, J.; Jin, C.; Yang, C.; Nefedov, A.; Wöll, C.; Wang, Y.; Shen, W. Structure of the Catalytically Active Copper–Ceria Interfacial Perimeter. Nat. Catal. 2019, 2(4), 334–341, https://doi.org/10.1038/s41929-019-0226-6.

in turn led to dramatic changes in the catalytic performance. Atomically dispersed Ru species showed  90% selectivity towards CO in CO2 hydrogenation, while the same catalyst without oxidative pretreatment showed > 90% selectivity towards methane under the same reaction conditions. Interestingly, substantially lower selectivity towards CO after the mild oxidative pretreatment was found for TiO2- and Al2O3-supported catalysts. This finding was explained by incomplete redispersion of initially deposited Ru NPs on these supports, which points at the exceptional ability of CeO2 to strongly anchor noble metal atoms. The redispersion of metallic species on ceria does not always require the actual reoxidation of the supported NPs. Newton et al. in a combined time-resolved EXAFS/DRIFTS operando study revealed a non-oxidative route for Pd redispersion upon switching CO and NO reactants over a ceria-based catalyst.105 The strong MSI in ceria catalysts seems to be the driving force for these peculiar transformations of metal speciation. Facet-dependent catalysis by ceria-based materials is an explicit manifestation of how the support structure can influence the catalytic behavior of the system. Wang et al. prepared nanostructured ceria supports with predominant (111) and (110) faceting to study the crystal plane effect in CO oxidation catalyzed by supported copper oxide clusters.106 When deposited on (111) ceria facets, these copper oxide clusters are more prone to reduction (from Cu2þ to Cuþ) as compared to those on (110) facets. This in turn led to substantially different CO oxidation performance, copper oxide clusters on (111) ceria surfaces being the most active at low temperature. For CO2 hydrogenation, on the contrary, (111) faceting of ceria seems to be unfavorable.107 It was shown that small ruthenium clusters supported on (100) ceria facets were the most active in CO2 methanation. Higher activity of Ru-CeO2 (100) catalyst stems from easier oxygen vacancy formation in (100) surfaces than in (110) and (111) ones, as was concluded from Raman and TPR data. On the other hand, the role of support faceting on the catalytic performance appeared to be relatively small for CO2 hydrogenation to methanol over Cu NPs supported on nanoshaped CeO2.108 It is clear that many aspects of the facetdependent chemistry related to metal-ceria interfaces are yet to be systematized and mechanistically understood. Seminal work of Corma and colleagues23 demonstrated how the size of support particles can impact the catalytic activity of supported metal NPs via MSI. Gold NPs deposited on large ceria particles were found to be nearly inactive in low-temperature CO

262

Metal-support interfaces in ceria-based catalysts

oxidation. On the contrary, small ( 4 nm) Au NP supported on small CeO2 nanocrystallites ( 5 nm) exhibited exceptional activity in CO oxidation already at 10  C. In a follow-up publication, it was shown that the ability of nanosized ceria to accommodate peroxide and superoxide species on its surface plays a decisive role in efficient CO oxidation.109 Parastaev et al. have recently shown that the optimal size of support NPs is key to strong MSI in cobalt-ceria-zirconia (CoCZ) catalysts for CO2 methanation.70 Specifically, the Co dispersion depended strongly on the particle size of the support. While relatively small and large CZ support particles resulted in low dispersion of Co NPs, medium-sized CZ support gave rise to the highest methanation activity due to a high dispersion of Co NPs. The enlarged metal-support interface in optimal catalyst was found to be beneficial for C-O bond dissociation at interfacial sites. To maximize metal-support interface, Tsang and colleagues encapsulated Pt NPs within a thin layer of CeO2.110 Such nanoarchitecture was completely inactive for CO methanation, while it was exceptionally active and selective towards the WGS reaction. This work is an elegant example of how knowledge-driven engineering of the metal-support interfaces can be used to fine-tune catalytic performance. In this section we discussed the role of the metal-support interface in catalysis by ceria-based materials. Synthesis of well-defined interfaces and extensive characterization including operando spectroscopy techniques can play a key role in identifying the reasons and effects of the MSI. The physical and chemical phenomena underlying the metal-support interactions are complex and difficult to unravel. Nonetheless, current knowledge already can be applied for fine-tuning the catalyst architecture for specific reactions. Truly inspiring example was recently demonstrated by groups of Somorjai and Yang, who used different functions of metalsupport interfaces to direct two chemical reactions into a controlled tandem.111 Combing colloidal synthesis and self-assembly, they engineered a catalyst, where each platinum nanoparticle had both Pt-SiO2 and Pt-CeO2 interfaces (Fig. 18). Each interfacial region serves its purpose: platinum-ceria catalyzes methanol decomposition into CO and H2, while platinum-silica stimulates hydroformylation of ethylene. Altogether this unique structure exhibits very high selectivity towards propanal, which is not accessible with separate metal-oxide interfaces. We believe that many more examples of such multifunctional nanostructured catalysts are yet to be explored once fundamental understanding of the metal-support interface chemistry is developed.

6.11.6

Catalysis by single-atom ceria-based catalysts

Recently, single-atom catalysis (SAC) has been actively investigated by the heterogeneous catalysis community.112–115 Single-atom catalysis refers to supported catalysts in which the active metal phase is present as isolated atoms. In such catalysts, strong MSI play a crucial role in order to achieve and maintain atomic dispersion of the active metal. The approach to atomically disperse oftenexpensive transition metals (e.g., PGMs) on an oxide support also provides a strategy to tune the activity and selectivity of catalytic reactions. An obvious application area is in automotive exhaust systems, where there remains a need for cheap, robust, and efficient catalysts for the foreseeable future.116 Other catalytic processes involving the use of expensive metals can also benefit from the maximized atom efficiency and the unusual reactivity offered by the interface between the single metal atom and the support. The SACs are also interesting model systems for studying MSI, because all metal atoms are part of the interface; thus, the influence of MSI is maximized. The uniformity in metal speciation renders SACs suitable for studying the mechanistic steps of catalytic reactions.117 In the following section, we will discuss several examples that highlight how the use of isolated metal atoms on ceria can contribute to our understanding of ceria chemistry in heterogeneous catalysts. One of the first examples of ceria-based SACs was provided by the group of Flytzani-Stephanopolous, who studied the WGS performance of nanoparticles of Au and Pt on ceria.118 While it was well accepted that metal sites in Pt/CeO2 catalyze the WGS reaction, the elegant approach to remove a large part of metallic species using cyanide (e.g., nearly 90% of all Au atoms can be removed in this way from Au/CeO2) indicated that the WGS activity is caused by a small fraction of the active metal in a highly dispersed form. The importance of a small fraction of highly dispersed metal was supported by the unaltered kinetic parameters after leaching. The strong MSI in ceria resulted in the dominance of cationic Au and Pt species in cyanide-leached samples as probed

Fig. 18 Interfacial tandem catalysis. (A) TEM image of CeO2 nanocube monolayer on a Pt monolayer forming a double interface system. (B) Propanal produced as a function of reaction time over CeO2–Pt–SiO2 and Pt-CeO2-SiO2 bilayers from ethylene and MeOH. (C) Illustration of the CeO2–Pt–SiO2 tandem catalyst. Adapted from Yamada, Y.; Tsung, C.-K.; Huang, W.; Huo, Z.; Habas, S. E.; Soejima, T.; Aliaga, C. E.; Somorjai, G. A.; Yang, P. Nanocrystal Bilayer for Tandem Catalysis. Nat. Chem. 2011, 3(5), 372–376. https://doi.org/10.1038/nchem.1018.

Metal-support interfaces in ceria-based catalysts

263

by XPS. Another important observation was that the cationic metal species activate the ceria surface oxygen atoms. Such ceria surface defects and highly dispersed Pt and Au cations, possibly embedded in the ceria surface, were considered to be pivotal to the WGS activity. This seminal work showed the profound effect of MSI not only on the dispersion but also on the electronic state of the supported metal. Datye and coworkers provided another evidence for the exceptional ability of ceria to anchor atomically dispersed noble metals.119 High-temperature calcination typically leads to sintering of the support and the active metal phase in supported noble-metal catalysts. As CeO2 strongly interacts with most metals, it can maintain a high metal dispersion. During aging at 800  C of a physical mixture of La-modified alumina-supported Pt and CeO2, volatile PtO2 species can be trapped on the surface of CeO2 (Fig. 20). The resulting Pt phase on CeO2 is atomically dispersed, which resulted in sinter-resistant catalysts even under CO oxidation conditions. Another interesting finding is the dependence of the trapping of single Pt atoms on the CeO2 facet. While nanocubes enclosed by (100) surfaces were shown to lessen the extent of Pt sintering, nanorods and polyhedra, exposing (111) surfaces, completely suppressed Pt agglomeration. As discussed earlier, the extent of MSI in CeO2-based catalysts is highlydependent on the exact ceria surface structure. Dvorak et al. used a classical surface science approach to show that Pt single atoms preferentially occupy the step-edge sites on the (111) ceria surfaces.120 XPS evidenced that Pt is present in the 2 þ oxidation state (Fig. 19). The strong interaction with the support strongly stabilizes dispersed Pt2þ ions via formation of Pt-O-Ce bonds. In a subsequent work, Datye and colleagues used aberration corrected-scanning transmission electron microscopy (AC-STEM) to confirm the isolated nature of Pt trapped on ceria steps after high-temperature annealing of real powder catalysts.121 Although challenging, advanced microscopy techniques such as HAADF-STEM allow distinguishing between Pt and Ce atoms. Compared to Pt, lighter transition metals of interest, for example Pd and Rh are more difficult to be discerned from Ce by techniques such as HAADF-STEM.

Fig. 19 Pt-CeO2 single-atom catalysts. (A) Atom-trapping of volatile Pt species transferred from the alumina supported Pt NPs; (B) in situ DRIFTS during CO oxidation over the Pt-CeO2 SAC. One major band at 2095 cm 1 and absence of bridged carbonyls manifest the atomic dispersion of Pt under reaction conditions. (C) Atomically dispersed Pt on step-edges of CeO2 (111). No Pt NP discerned from scanning tunneling microscopy image (left); Pt single-atoms are trapped at steps as (2þ) cations, as derived from XPS spectrum (right). (d) Pt single atoms localize at step-edges of ceria in powder catalysts. (A, B) Adapted from Jones, J.; Xiong, H.; DeLaRiva, A. T.; Peterson, E. J.; Pham, H.; Challa, S. R.; Qi, G.; Oh, S.; Wiebenga, M. H.; Pereira Hernandez, X. I.; Wang, Y.; Datye, A. K. Thermally Stable Single-Atom Platinum-on-Ceria Catalysts Via Atom Trapping. Science 2016, 353(6295), 150–154, https://doi.org/10.1126/science.aaf8800. (C) Adapted from Dvorák, F.; Camellone, M. F.; Tovt, A.; Tran, N.; Negreiros, F. R.; Vorokhta, M.; Skála, T.; Matolínová, I.; Myslivecek, J.; Matolín, V.; Fabris, S. Creating Single-Atom Pt-Ceria Catalysts by Surface Step Decoration. Nat. Commun. 2016, 7(10801), 1–8, https://doi.org/10.1038/ncomms10801. (D) Adapted from Kunwar, D.; Zhou, S.; Delariva, A.; Peterson, E. J.; Xiong, H.; Pereira-Hernández, X. I.; Purdy, S. C.; ter Veen, R.; Brongersma, H. H.; Miller, J. T.; Hashiguchi, H.; Kovarik, L.; Lin, S.; Guo, H.; Wang, Y.; Datye, A. K. Stabilizing High Metal Loadings of Thermally Stable Platinum Single Atoms on an Industrial Catalyst Support. ACS Catal. 2019, 9(5), 3978– 3990. https://doi.org/10.1021/acscatal.8b04885.

264

Metal-support interfaces in ceria-based catalysts

A systematic spectroscopic study by Neitzel et al. provides unique insight into the stability and redox chemistry of ceriasupported single atoms of Pt, Pd, and Ni.122 By preparing thin films of model catalysts using simultaneous physical vapor deposition of ceria and respective metals in the form of oxides, a set of materials was obtained with the active phase speciating from single atoms to clusters. Low-loaded model catalysts (up to  10 mol%) showed distinct features in XPS spectra related to highly-dispersed Pt2þ, Pd2þ, and Ni2þ on CeO2. To probe their reactivity, the catalysts were exposed to hydrogen ( 10 7 mbar) during stepwise heating. Simultaneously, the valence state of the metals under study was monitored by surface-sensitive XPS, while the Ce electronic state was probed by RPES. Atomically dispersed Pt, Pd, and Ni are practically inert towards reduction by hydrogen at temperatures as high as 400  C. However, the use of higher loadings of Pt and Pd led to the formation of some metal oxide clusters on the surface and a dramatic change in the redox properties of the system. Exposure of such high-loaded samples to hydrogen led to a substantial reduction of Ce4þ to Ce3þ already at 125  C, likely via hydrogen spillover from the agglomerated metallic species, followed by reduction of single atoms of supported transition metals. From these data, the authors concluded that in contrast to metallic species, cationic single atom species cannot activate H2. It should be noted, however, that the hydrogen partial pressure, used in this work is orders of magnitude lower than in actual catalytic hydrogenation reactions. Such detailed spectroscopic insights into the redox chemistry of ceria-supported SACs, gained under more relevant conditions (at least in a mbar range), can further improve our understanding of these complex catalysts. Despite the high atom efficiency, the activity offered by ceria-based SACs is not always on par with conventional catalysts based on metal nanoparticles supported by CeO2. For example, it has been reported that Pt single atoms on CeO2 exhibit a low CO oxidation activity at temperature below 150  C.123 The low activity has been linked to the very strong CO adsorption on Pt single atoms. Using CO reduction at 275  C, Datye and colleagues demonstrated that Pt-CeO2 SAC can be transformed into a very active and stable catalyst. This was explained by agglomeration of initially present Pt single atoms into small Pt clusters, which strongly interact with the CeO2 support. The enhanced CO oxidation activity at low temperature can be associated with the lability of ceria lattice oxygen atoms at the metal-support interface. An operando NAP-XPS study of the Pt and Ce valence states upon switches between CO/ O2 and CO gas feeds demonstrated that CeO2 reduction with creation of oxygen vacancies can occur at temperatures as low as 50  C. These findings together with TPR-CO and DRIFTS characterization, allowed the authors to conclude that CO is oxidized by mobile oxygen associated with the Pt-CeO2 interface. Wang et al. suggested that the formation of distinct Pt NPs is not necessary and, instead, very small clusters with Pt-O-Pt structural fragments obtained by a mild reductive pretreatment of a Pt/CeO2 SAC, are responsible for the higher CO oxidation activity than SAC parent.124 A recent report by the group of Grunwaldt emphasizes the importance of using operando characterization for studying the dynamic Pt-CeO2 interface.125 Although ex situ characterization suggested that the active sites remain virtually unaltered, this work revealed that the Pt speciation changes during the reaction. By employing operando HERFD-XANES, it was shown that reduction of atomically dispersed Pt2þ species to Ptd þ results in formation of few-atom Pt clusters during heating in a CO and O2 mixture (Fig. 20). Their presence was linked to a rapid increase in CO oxidation activity, pointing at participation of the semi-reduced Pt species in CO oxidation at elevated temperatures (> 200  C). Importantly, such small Pt clusters are prone to redispersion into Pt single atoms upon cooling in the same reaction mixture. This finding underpins the essential role of operando methodology in attempts to understand the metal-support interfaces in dynamic metal-ceria catalysts. The activity of Pt single atoms supported on ceria in oxidation reactions can be efficiently tuned by changing the oxidation state of Pt atoms.126 The authors prepared a system in which Pt atoms are trapped on highly defective nanosized CeO2 particles dispersed over an alumina support. Next, by straightforward H2 reduction of as-prepared SACs at different temperatures, the oxidation state of Pt single atoms could be controlled, circumventing their agglomeration into clusters or NPs as evidenced by XAS. The different oxidation states of Pt species caused dramatic differences in the activity for oxidation of CO, CH4 and NO. The optimal catalyst, reduced at 300  C, contained  70% of metallic Pt species, which remained intact even after prolonged catalytic testing. Temperature-programmed desorption experiments shed light on the origin of the performance differences as a function of Pt oxidation state. The adsorption strength of reactants was shown to be strongly dependent on the oxidation state of Pt centers, which can be linked to the activity trends. Another interesting approach to control the electronic structure of the metal and thus the performance of Pt-CeO2 SACs was illustrated by Yanfu et al.127 Doping the surface of ceria nanorods with phosphorous altered the valence state of the supported Pt atoms due to the charge-transfer from Pt to P. This led to an almost one order of magnitude higher activity of the P-doped catalyst upon in the hydrogenation of styrene, cyclohexane, phenylacetylene and nitrobenzene. Such enhancement in hydrogenation was linked to more pronounced hydrogen spillover for the P-doped system. The work of Cargnello et al. demonstrated that Pd and Pt atoms located at the interface between metallic NPs and ceria display similar catalytic behavior in low-temperature CO oxidation.97 However, when dispersed as single atoms, Pt and Pd catalysts exhibit distinctly different catalytic activity. In contrast to Pt, atomically dispersed Pd-O species on the surface of ceria are highly active in low-temperature CO oxidation.128 Spezzati et al. found that Pd single atoms on the surface of ceria nanorods, exposing (111) surfaces, are stabilized exclusively in the oxidized form according to XAS. However, exposure to the reaction feed at 50  C led to the formation of semi-reduced Pd clusters, as follows from the presence of bridged carbonyls in CO IR spectra. Later work of the same authors8 unveiled that the stabilizing effect of ceria on atomically dispersed Pd depended on the facet exposed by ceria. When Pd was deposited on ceria nanorods and nanocubes, exposing predominantly (111) and (100) surfaces respectively, PdO NPs were accompanying Pd single atoms in the nanocubes case. Pd supported on CeO2 nanocubes was found to be substantially less active in low-temperature CO oxidation than nanorod-supported Pd-CeO2 SACs. These differences in reactivity were linked to intrinsically different redox properties of (111) and (100) facets, with (100) ones being less favorable for the stabilization of oxidized single-atom Pd-O moieties responsible for CO oxidation at low temperatures. Another example of the influence of ceria

Metal-support interfaces in ceria-based catalysts

265

Fig. 20 Tracking the interface dynamics in Pt-CeO2 SACs. (A) Operando HERFD-XAS of Pt L3-edge upon heating in reaction mixture. (B) Catalytic activity during the HERFD-XAS and Pt speciation during the CO oxidation tests. (C) Proposed scheme for the reversible formation of the catalytically active PtXd þ cluster based on operando HERFD-XANES analysis (grey, Pt; yellow, Ce; red, O; dark grey, C). Adapted from Maurer, F.; Jelic, J.; Wang, J.; Gänzler, A.; Dolcet, P.; Wöll, C.; Wang, Y.; Studt, F.; Casapu, M.; Grunwaldt, J.-D. Tracking the Formation, Fate and Consequence for Catalytic Activity of Pt Single Sites on CeO2. Nat. Catal. 2020,3(10), 824–833, https://doi.org/10.1038/s41929-020-00508-7.

surface termination on the activity of single-atom Pd species was evident for electrocatalytic oxidation of formic acid.129 Here, however, polar (100) surfaces of ceria cubes were catalytically favored over (111) and (110) surfaces exposed by nanorod and spherical ceria. On the basis of DFT calculations and in situ NAP-XPS results, the authors concluded that on (100) facets oxygen vacancies can be generated easier through a reverse oxygen spillover to Pd sites. CO stripping voltammetry indicated a key role of spilled-over oxygen in circumventing poisoning of Pd by CO, which is the product of the unwanted process of formic acid dehydration. These examples demonstrate the pronounced effects of MSI on the redox properties, the reactivity, and the stability of ceria-supported SACs. Finally, a recent report by the group of Lee demonstrated a viable strategy to stabilize noble metals as single atoms and clusters of low nuclearity on nanosized ceria under harsh reaction conditions.130 They developed a nanoarchitecture in which Pt, Pd and Rh were dispersed over defect-rich ceria NPs ( 5 nm) stabilized on alumina (Fig. 21). Depending on the impregnation temperature, reduction of these nanocomposites in hydrogen led to either single atoms or clusters of noble metals, as demonstrated by CO IR and EXAFS analysis. Catalytic testing with a model exhaust feed revealed that clustered metal catalysts significantly outperform their SAC counterparts during the simultaneous conversion of CO, C3H6, C3H8 and NO at low temperature. This was explained by competitive adsorption of CO and NO with unreacted hydrocarbons on single atoms of noble metals. Surprisingly, the clustered catalysts maintained a high activity, even after harsh hydrothermal aging at 900  C, making them comparable in performance with the most durable and efficient catalysts for exhaust after treatment reported so far.131 It still remains unclear what are the intrinsic reasons for such exceptional activity and stability of these complex nanoarchitectures. An explanation can be sought in the peculiar properties of nanosized ceria. As we discussed earlier, using model systems as an example, nanostructured ceria displays facile oxygen transfer at the metal-support interface via spillover.22 This oxygen mobility can play an important role in reduction of, e.g., CO2 or H2O as well as in oxidation of, e.g., CO or H2. In addition, lattice strain effects and the presence of defects characteristic for small ceria NPs lead to repulsive interactions between small metal clusters, preventing them from sintering.101 It is clear that novel catalysts based on nanosized ceria such as developed by Lee and coworkers require further in-depth mechanistic studies in order to explain and fully utilize their peculiar properties. In this section, we discussed various aspects of metal-ceria interfaces where the metal phase is limited to a single metal atom. This represents a burgeoning field of modern heterogeneous catalysis. The exceptional ability of ceria to maintain noble metals in highly dispersed form, even under harsh reaction conditions, is one of the key reasons for the widespread application of ceria as a support for SACs. The research on ceria-based SACs is mostly related to PGMs due to the significant interest in developing novel atomefficient and stable catalysts for automotive exhaust neutralizers. The presence of PGMs in such catalysts seems to be inevitable. Therefore, innovations in the design of catalysts containing atomically dispersed PGM and ceria need to be sought. From

266

Metal-support interfaces in ceria-based catalysts

Fig. 21 Highly durable metal ensemble catalysts. (A) Synthetic procedure of metal ensemble catalysts (ESC) with full dispersion and reduced metallic states. (B, C) HAADF-STEM images of Pt ESC. (D) CO oxidation performance during simultaneous CO, C3H6, C3H8 and NO conversion. Adapted from Jeong, H.; Kwon, O.; Kim, B. S.; Bae, J.; Shin, S.; Kim, H. E.; Kim, J.; Lee, H. Highly Durable Metal Ensemble Catalysts with Full Dispersion for Automotive Applications beyond Single-Atom Catalysts. Nat. Catal. 2020, 1–8, https://doi.org/10.1038/s41929-020-0427-z.

a perspective of fundamental studies, the undercoordinated nature of supported single atoms maximizes the influence of MSI, which in turn can be studied in greater detail using well-defined SACs. One should be cautious, however, with propagating the knowledge of the properties of single atoms to interfacial sites in NPs and vice versa – the contrasting cases of Pt and Pd for low-temperature CO oxidation illustrates this limitation. Nevertheless, ceria-based materials with atomically dispersed metal species appear to be a promising class of catalysts, for which many aspects remain to be discovered.

6.11.7

Summary

The unique properties of ceria have made it a widely used support material in heterogeneous catalysts. The tunable redox chemistry, which is intimately related to the surface structure and the crystallite size of ceria, allows optimizing the surface reactivity for specific catalytic applications. Ceria is a well-suited support for transition metals because of the relatively strong ceria-metal interactions. Ceria is particularly known for its rich interfacial chemistry. The reactivity and exact properties of the interfacial sites are yet to be understood, but the rapidly growing body of knowledge points to a key role of the metal-support interface in a large number of reactions catalyzed by ceria-supported catalysts. Aiming at better understanding of the individual steps in the underlying reaction mechanisms, it is critical to prepare well-defined interfaces between ceria and metal phases. Novel synthetic approaches together with advances in operando spectroscopy and microscopic imaging can push the boundaries in the development of structurefunction relationships in these catalyst systems. Single atoms of noble metals dispersed on ceria hold a promise of maximizing the use of such PGMs, which are key in combatting harmful automotive emissions. When prepared in a uniform manner, such single-atom catalysts can serve as a bridge between classical surface science models and technical powder catalysts in different types of studies in which materials and pressure gaps are overcome. Despite notable progress in understanding ceria-based catalysis, many aspects of CeO2 chemistry are yet to be scrutinized. Simple and controlled synthesis of well-defined CeO2 nanostructures as well as that of the metal phase supported by ceria are pivotal. Uniformity of metal species ranging from nanoparticles to few-atom clusters or even single atoms is an essential prerequisite for the preparation of catalysts suitable for fundamental studies. To date, the majority of spectroscopic approaches are based on following the structure and electronic state of the supported metals, traditionally assumed to be the active component of the catalysts. For ceria-based catalysts, the nature of the metal-support interface often governs the overall activity. Thus, the speciation and electronic structure of both components needs to be studied with equal attention. X-ray based in situ spectroscopy has seen an immense progress in the past decade. Surface science techniques, such as RPES, can provide extreme sensitivity to even minute changes in the valence state of elements. Yet, these studies have only been performed under close-to high-vacuum conditions, leaving the pressure gap open. Using hard X-ray based spectroscopy tools, one can study catalysts in a wide range of pressures

Metal-support interfaces in ceria-based catalysts

267

and temperature, including the very conditions realized in practical applications. This comes, however, at the expense of surface sensitivity, as compared to the soft X-ray regime. However, this downside of bulk-sensitive XAS and XRD can be alleviated using nanosized materials, in which the surface-to-bulk ratio is maximized. The properties of such nanosized ceria supports appear to be very different from their bulkier counterparts and further in-depth fundamental studies of nanosized CeO2 are necessary. A promising approach involves time-resolved XAS to follow supported metal and Ce valence states, while the catalytic performance is tracked simultaneously as a function of transiently changing reaction conditions. This combination allows to unambiguously distinguish the real active sites from spectator species. We believe that basic concepts behind the chemistry of ceria materials, which were introduced here in the context of heterogeneous catalysis, will be an asset for a broad chemical science community. It is thrilling to see how rapidly the applications of ceria expand, attracting the interest of more and more researchers across many disciplines. Despite the substantial insights already acquired, we can only imagine the exciting chemical phenomena that will be uncovered in the years to come.

References 1. Jaiswal, N.; Tanwar, K.; Suman, R.; Kumar, D.; Uppadhya, S.; Parkash, O. A Brief Review on Ceria Based Solid Electrolytes for Solid Oxide Fuel Cells. J. Alloys Compd. 2019, 984–1005. https://doi.org/10.1016/j.jallcom.2018.12.015. 2. Kim, H. J.; Lee, G.; Jang, M. G.; Noh, K.; Han, J. W. Rational Design of Transition Metal Co-Doped Ceria Catalysts for Low-Temperature CO Oxidation. ChemCatChem 2019, 11 (9), 2288–2296. https://doi.org/10.1002/cctc.201900178. 3. Mullins, D. R. The Surface Chemistry of Cerium Oxide. Surf. Sci. Rep. 2015, 70 (1), 42–85. https://doi.org/10.1016/j.surfrep.2014.12.001. 4. Nolan, M.; Grigoleit, S.; Sayle, D. C.; Parker, S. C.; Watson, G. W. Density Functional Theory Studies of the Structure and Electronic Structure of Pure and Defective Low Index Surfaces of Ceria. Surf. Sci. 2005, 576 (1–3), 217–229. https://doi.org/10.1016/j.susc.2004.12.016. 5. van Santen, R. A.; Tranca, I.; Hensen, E. J. M. Theory of Surface Chemistry and Reactivity of Reducible Oxides. Catal. Today 2015, 244, 63–84. https://doi.org/10.1016/ j.cattod.2014.07.009. 6. Duchon, T.; Hackl, J.; Mueller, D. N.; Kullgren, J.; Du, D.; Senanayake, S. D.; Mouls, C.; Gottlob, D. M.; Khan, M. I.; Cramm, S.; Veltruská, K.; Matolín, V.; Nemsák, S.; Schneider, C. M. Establishing Structure-Sensitivity of Ceria Reducibility: Real-Time Observations of Surface-Hydrogen Interactions. J. Mater. Chem. A 2020, 8 (11), 5501– 5507. https://doi.org/10.1039/c9ta11784a. 7. Wu, Z.; Li, M.; Overbury, S. H. On the Structure Dependence of CO Oxidation over CeO2 Nanocrystals with Well-Defined Surface Planes. J. Catal. 2012, 285 (1), 61–73. https://doi.org/10.1016/j.jcat.2011.09.011. 8. Spezzati, G.; Benavidez, A. D.; DeLaRiva, A. T.; Su, Y.; Hofmann, J. P.; Asahina, S.; Olivier, E. J.; Neethling, J. H.; Miller, J. T.; Datye, A. K.; Hensen, E. J. M. CO Oxidation by Pd Supported on CeO2(100) and CeO2(111) Facets. Appl. Catal. B Environ. 2019, 243, 36–46. https://doi.org/10.1016/J.APCATB.2018.10.015. 9. Paier, J.; Penschke, C.; Sauer, J. Oxygen Defects and Surface Chemistry of Ceria: Quantum Chemical Studies Compared to Experiment. Chem. Rev. 2013, 12, 3949–3985. https://doi.org/10.1021/cr3004949. 10. Agarwal, S.; Lefferts, L.; Mojet, B. L.; Ligthart, D. A. J. M.; Hensen, E. J. M.; Mitchell, D. R. G.; Erasmus, W. J.; Anderson, B. G.; Olivier, E. J.; Neethling, J. H.; Datye, A. K. Exposed Surfaces on Shape-Controlled Ceria Nanoparticles Revealed through AC-TEM and Water-Gas Shift Reactivity. ChemSusChem 2013, 6 (10), 1898–1906. https:// doi.org/10.1002/cssc.201300651. 11. Zhao, E. W.; Zheng, H.; Zhou, R.; Hagelin-Weaver, H. E.; Bowers, C. R. Shaped Ceria Nanocrystals Catalyze Efficient and Selective Para-Hydrogen-Enhanced Polarization. Angew. Chem. Int. Ed. 2015, 54 (48), 14270–14275. https://doi.org/10.1002/anie.201506045. 12. Yang, C.; Capdevila-Cortada, M.; Dong, C.; Zhou, Y.; Wang, J.; Yu, X.; Nefedov, A.; Heißler, S.; López, N.; Shen, W.; Wöll, C.; Wang, Y. Surface Refaceting Mechanism on Cubic Ceria. J. Phys. Chem. Lett. 2020, 11 (18), 7925–7931. https://doi.org/10.1021/acs.jpclett.0c02409. 13. Wu, Q.; Zhang, F.; Xiao, P.; Tao, H.; Wang, X.; Hu, Z.; Lü, Y. Great Influence of Anions for Controllable Synthesis of CeO2 Nanostruetures: From Nanorods to Nanocubes. J. Phys. Chem. C 2008, 112 (44), 17076–17080. https://doi.org/10.1021/jp804140e. 14. Tsunekawa, S.; Wang, J. T.; Kawazoe, Y. Lattice Constants and Electron Gap Energies of Nano- and Subnano-Sized Cerium Oxides from the Experiments and First-Principles Calculations. J. Alloys Compd. 2006, 408–412, 1145–1148. https://doi.org/10.1016/j.jallcom.2004.12.140. Elsevier. 15. Tsunekawa, S.; Sahara, R.; Kawazoe, Y.; Ishikawa, K. Lattice Relaxation of Monosize CeO2-x Nanocrystalline Particles. Appl. Surf. Sci. 1999, 152 (1), 53–56. https://doi.org/ 10.1016/S0169-4332(99)00298-6. 16. Zhou, X. D.; Huebner, W. Size-Induced Lattice Relaxation in CeO2 Nanoparticles. Appl. Phys. Lett. 2001, 79 (21), 3512–3514. https://doi.org/10.1063/1.1419235. 17. Chen, L.; Fleming, P.; Morris, V.; Holmes, J. D.; Morris, M. A. Size-Related Lattice Parameter Changes and Surface Defects in Ceria Nanocrystals. J. Phys. Chem. C 2010. https://doi.org/10.1021/jp1031465. 18. Xu, J.; Harmer, J.; Li, G.; Chapman, T.; Collier, P.; Longworth, S.; Tsang, S. C. Size Dependent Oxygen Buffering Capacity of Ceria Nanocrystals. Chem. Commun. 2010, 46 (11), 1887–1889. https://doi.org/10.1039/b923780a. 19. Migani, A.; Vayssilov, G. N.; Bromley, S. T.; Illas, F.; Neyman, K. M. Dramatic Reduction of the Oxygen Vacancy Formation Energy in Ceria Particles: A Possible Key to their Remarkable Reactivity at the Nanoscale. J. Mater. Chem. 2010, 20 (46), 10535–10546. https://doi.org/10.1039/c0jm01908a. 20. Sk, M. A.; Kozlov, S. M.; Lim, K. H.; Migani, A.; Neyman, K. M. Oxygen Vacancies in Self-Assemblies of Ceria Nanoparticles. J. Mater. Chem. A 2014, 2 (43), 18329–18338. https://doi.org/10.1039/c4ta02200a. 21. Bruix, A.; Neyman, K. M. Modeling Ceria-Based Nanomaterials for Catalysis and Related Applications. Catal. Lett. 2016, 146 (10), 2053–2080. https://doi.org/10.1007/ s10562-016-1799-1. 22. Vayssilov, G. N.; Lykhach, Y.; Migani, A.; Staudt, T.; Petrova, G. P.; Tsud, N.; Skála, T.; Bruix, A.; Illas, F.; Prince, K. C.; Matolín, V.; Neyman, K. M.; Libuda, J. Support Nanostructure Boosts Oxygen Transfer to Catalytically Active Platinum Nanoparticles. Nat. Mater. 2011, 10 (4), 310–315. https://doi.org/10.1038/nmat2976. 23. Carrettin, S.; Concepción, P.; Corma, A.; López Nieto, J. M.; Puntes, V. F. Nanocrystalline CeO2 Increases the Activity of Au for CO Oxidation by Two Orders of Magnitude. Angew. Chem. Int. Ed. 2004, 43 (19), 2538–2540. https://doi.org/10.1002/anie.200353570. 24. Puigdollers, A. R.; Schlexer, P.; Tosoni, S.; Pacchioni, G. Increasing Oxide Reducibility: The Role of Metal/Oxide Interfaces in the Formation of Oxygen Vacancies. ACS Catal. 2017, 6493–6513. https://doi.org/10.1021/acscatal.7b01913. American Chemical Society. 25. Zhu, H.; Yang, C.; Li, Q.; Ren, Y.; Neuefeind, J. C.; Gu, L.; Liu, H.; Fan, L.; Chen, J.; Deng, J.; Wang, N.; Hong, J.; Xing, X. Charge Transfer Drives Anomalous Phase Transition in Ceria. Nat. Commun. 2018, 9 (1), 1–8. https://doi.org/10.1038/s41467-018-07526-x. 26. Trovarelli, A. Catalytic Properties of Ceria and CeO2-Containing Materials. Catal. Rev. 1996, 38 (4), 439–520. https://doi.org/10.1080/01614949608006464. 27. Montini, T.; Melchionna, M.; Monai, M.; Fornasiero, P. Fundamentals and Catalytic Applications of CeO2-Based Materials. Chem. Rev. 2016, 116 (10), 5987–6041. https:// doi.org/10.1021/acs.chemrev.5b00603. 28. Gorte, R. J. Ceria in Catalysis: From Automotive Applications to the Water-Gas Shift Reaction. AICHE J. 2010, 56 (5). https://doi.org/10.1002/aic.12234.

268

Metal-support interfaces in ceria-based catalysts

29. le Gal, A.; Abanades, S. Dopant Incorporation in Ceria for Enhanced Water-Splitting Activity during Solar Thermochemical Hydrogen Generation. J. Phys. Chem. C 2012, 116 (25), 13516–13523. https://doi.org/10.1021/jp302146c. 30. Chen, P.-L.; Chen, I.-W. Reactive Cerium(IV) Oxide Powders by the Homogeneous Precipitation Method. J. Am. Ceram. Soc. 1993, 76 (6), 1577–1583. https://doi.org/ 10.1111/j.1151-2916.1993.tb03942.x. 31. Zhou, X. D.; Huebner, W.; Anderson, H. U. Room-Temperature Homogeneous Nucleation Synthesis and Thermal Stability of Nanometer Single Crystal CeO2. Appl. Phys. Lett. 2002, 80 (20), 3814–3816. https://doi.org/10.1063/1.1481244. 32. di Monte, R.; Kaspar, J. Heterogeneous Environmental Catalysis – A Gentle Art: CeO2-ZrO2 Mixed Oxides as a Case History. Catal. Today 2005, 100, 27–35. https://doi.org/ 10.1016/j.cattod.2004.11.005. Elsevier. 33. Devaiah, D.; Reddy, L. H.; Park, S. E.; Reddy, B. M. Ceria–Zirconia Mixed Oxides: Synthetic Methods and Applications. Catal. Rev. Sci. Eng. 2018, 60 (2), 177–277. https:// doi.org/10.1080/01614940.2017.1415058. 34. Mukherjee, D.; Reddy, B. M. Noble Metal-Free CeO2-Based Mixed Oxides for CO and Soot Oxidation. Catal. Today 2018, 309, 227–235. https://doi.org/10.1016/ j.cattod.2017.06.017. 35. Wang, J.; Chen, H.; Hu, Z.; Yao, M.; Li, Y. A Review on the Pd-Based Three-Way Catalyst. Catal. Rev. 2015, 57 (1), 79–144. https://doi.org/10.1080/ 01614940.2014.977059. 36. Terribile, D.; Trovarelli, A.; Llorca, J.; de Leitenburg, C.; Dolcetti, G. The Synthesis and Characterization of Mesoporous High-Surface Area Ceria Prepared Using a Hybrid Organic/Inorganic Route. J. Catal. 1998, 178 (1), 299–308. https://doi.org/10.1006/JCAT.1998.2152. 37. Boaro, M.; Colussi, S.; Trovarelli, A. Ceria-Based Materials in Hydrogenation and Reforming Reactions for CO2 Valorization. Front. Chem. 2019, 28. https://doi.org/10.3389/ fchem.2019.00028. Frontiers Media S.A. 38. Laberty-Robert, C.; Long, J. W.; Lucas, E. M.; Pettigrew, K. A.; Stroud, R. M.; Doescher, M. S.; Rolison, D. R. Sol-Gel-Derived Ceria Nanoarchitectures: Synthesis, Characterization, and Electrical Properties. Chem. Mater. 2006, 18 (1), 50–58. https://doi.org/10.1021/cm051385t. 39. Yu, T.; Joo, J.; Park, Y.i.; Hyeon, T. Large-Scale Nonhydrolytic Sol-Gel Synthesis of Uniform-Sized Ceria Nanocrystals with Spherical, Wire, and Tadpole Shapes. Angew. Chem. Int. Ed. 2005, 44 (45), 7411–7414. https://doi.org/10.1002/anie.200500992. 40. Riley, C.; Canning, G.; de La Riva, A.; Zhou, S.; Peterson, E.; Boubnov, A.; Hoffman, A.; Tran, M.; Bare, S. R.; Lin, S.; Guo, H.; Datye, A. Environmentally Benign Synthesis of a PGM-Free Catalyst for Low Temperature CO Oxidation. Appl. Catal. B Environ. 2020, 264, 118547. https://doi.org/10.1016/j.apcatb.2019.118547. 41. Zhang, Y.; Bals, S.; van Tendeloo, G. Understanding CeO2-Based Nanostructures through Advanced Electron Microscopy in 2D and 3D. Part. Part. Syst. Charact. 2019, 36 (1), 1800287. https://doi.org/10.1002/ppsc.201800287. 42. Aneggi, E.; Wiater, D.; de Leitenburg, C.; Llorca, J.; Trovarelli, A. Shape-Dependent Activity of Ceria in Soot Combustion. ACS Catal. 2014, 4 (1), 172–181. https://doi.org/ 10.1021/cs400850r. 43. Trovarelli, A.; Llorca, J. Ceria Catalysts at Nanoscale: How Do Crystal Shapes Shape Catalysis? ACS Catal. 2017, 7 (7), 4716–4735 https://doi.org/10.1021/ acscatal.7b01246. 44. Yang, C.; Yu, X.; Heißler, S.; Nefedov, A.; Colussi, S.; Llorca, J.; Trovarelli, A.; Wang, Y.; Wöll, C. Surface Faceting and Reconstruction of Ceria Nanoparticles. Angew. Chem. Int. Ed. 2017, 56 (1), 375–379. https://doi.org/10.1002/anie.201609179. 45. Zhou, K.; Wang, X.; Sun, X.; Peng, Q.; Li, Y. Enhanced Catalytic Activity of Ceria Nanorods from Well-Defined Reactive Crystal Planes. J. Catal. 2005, 229 (1), 206–212. https://doi.org/10.1016/j.jcat.2004.11.004. 46. Mai, H. X.; Sun, L. D.; Zhang, Y. W.; Si, R.; Feng, W.; Zhang, H. P.; Liu, H. C.; Yan, C. H. Shape-Selective Synthesis and Oxygen Storage Behavior of Ceria Nanopolyhedra, Nanorods, and Nanocubes. J. Phys. Chem. B 2005, 109 (51), 24380–24385. https://doi.org/10.1021/jp055584b. 47. Cargnello, M.; Jaen, J. J. D.; Garrido, J. C. H.; Bakhmutsky, K.; Montini, T.; Gamez, J. J. C.; Gorte, R. J.; Fornasiero, P. Exceptional Activity for Methane Combustion over Modular Pd@CeO2 Subunits on Functionalized Al2O3. Science 2012, 337 (6095), 713–717. https://doi.org/10.1126/science.1222887. 48. Peng, H.; Rao, C.; Zhang, N.; Wang, X.; Liu, W.; Mao, W.; Han, L.; Zhang, P.; Dai, S. Confined Ultrathin Pd-Ce Nanowires with Outstanding Moisture and SO2 Tolerance in Methane Combustion. Angew. Chem. Int. Ed. 2018, 57 (29), 8953–8957. https://doi.org/10.1002/anie.201803393. 49. Vrijburg, W. L.; van Helden, J. W. A.; Parastaev, A.; Groeneveld, E.; Pidko, E. A.; Hensen, E. J. M. Ceria-Zirconia Encapsulated Ni Nanoparticles for CO2 Methanation. Catal. Sci. Technol. 2019, 9 (18), 5001–5010. https://doi.org/10.1039/c9cy01428d. 50. Danielis, M.; Colussi, S.; de Leitenburg, C.; Soler, L.; Llorca, J.; Trovarelli, A. Outstanding Methane Oxidation Performance of Palladium-Embedded Ceria Catalysts Prepared by a One-Step Dry Ball-Milling Method. Angew. Chem. Int. Ed. 2018, 57 (32), 10212–10216. https://doi.org/10.1002/anie.201805929. 51. Colussi, S.; Gayen, A.; Camellone, M. F.; Boaro, M.; Llorca, J.; Fabris, S.; Trovarelli, A. Nanofaceted Pd-O Sites in Pd-Ce Surface Superstructures: Enhanced Activity in Catalytic Combustion of Methane. Angew. Chem. Int. Ed. 2009, 48 (45), 8481–8484. https://doi.org/10.1002/anie.200903581. 52. Gröhn, A. J.; Pratsinis, S. E.; Sánchez-Ferrer, A.; Mezzenga, R.; Wegner, K. Scale-up of Nanoparticle Synthesis by Flame Spray Pyrolysis: The High-Temperature Particle Residence Time. Ind. Eng. Chem. Res. 2014, 53 (26), 10734–10742. https://doi.org/10.1021/ie501709s. 53. Seo, D. J.; Ryu, K. O.; Park, S. B.; Kim, K. Y.; Song, R. H. Synthesis and Properties of Ce1-XGdxO2-x/2 Solid Solution Prepared by Flame Spray Pyrolysis. Mater. Res. Bull. 2006, 41 (2), 359–366. https://doi.org/10.1016/j.materresbull.2005.08.012. 54. Strobel, R.; Baiker, A.; Pratsinis, S. E. Aerosol Flame Synthesis of Catalysts. Adv. Powder Technol. 2006, 457–480. https://doi.org/10.1163/156855206778440525. VSP BV. 55. Feng, X.; Sayle, D. C.; Wang, Z. L.; Paras, M. S.; Santora, B.; Sutorik, A. C.; Sayle, T. X. T.; Yang, Y.; Ding, Y.; Wang, X.; Her, Y.-S. Converting Ceria Polyhedral Nanoparticles into Single-Crystal Nanospheres. Science 2006, 312 (5779), 1504–1508. https://doi.org/10.1126/science.1125767. 56. Sayle, D. C.; Feng, X.; Ding, Y.; Zhong, L. W.; Sayle, T. X. T. “Simulating Synthesis”: Ceria Nanosphere Self-Assembly into Nanorods and Framework Architectures. J. Am. Chem. Soc. 2007, 129 (25), 7924–7935. https://doi.org/10.1021/ja070893w. 57. Kaspar, J.; Fornasiero, P.; Hickey, N. Automotive Catalytic Converters: Current Status and Some Perspectives. Catal. Today 2003, 77, 419–449. https://doi.org/10.1016/ S0920-5861(02)00384-X. Elsevier. 58. Rodriguez, J. A.; Grinter, D. C.; Liu, Z.; Palomino, R. M.; Senanayake, S. D. Ceria-Based Model Catalysts: Fundamental Studies on the Importance of the Metal-Ceria Interface in CO Oxidation, the Water-Gas Shift, CO2 Hydrogenation, and Methane and Alcohol Reforming. Chem. Soc. Rev. 2017, 1824–1841. https://doi.org/10.1039/c6cs00863a. 59. Lukashuk, L.; Foettinger, K. In Situ and Operando Spectroscopy: A Powerful Approach towards Understanding Catalysts. Johnson Matthey Technol. Rev. 2018, 62 (3), 316– 331. https://doi.org/10.1595/205651318X15234323420569. 60. Weckhuysen, B. M. Determining the Active Site in a Catalytic Process: Operando Spectroscopy is More than a Buzzword. Phys. Chem. Chem. Phys. 2003, 5, 4351–4360. https://doi.org/10.1039/b309650p. Royal Society of Chemistry. 61. Anirban, S.; Paul, T.; Dutta, A. Vacancy Mediated Ionic Conduction in Dy Substituted Nanoceria: A Structure-Property Correlation Study. RSC Adv. 2015, 5 (62), 50186– 50195. https://doi.org/10.1039/c5ra06730h. 62. Zhang, F.; Liu, Z.; Zhang, S.; Akter, N.; Palomino, R. M.; Vovchok, D.; Orozco, I.; Salazar, D.; Rodriguez, J. A.; Llorca, J.; Lee, J.; Kim, D. H.; Xu, W.; Frenkel, A. I.; Li, Y.; Kim, T.; Senanayake, S. D. In Situ Elucidation of the Active State of Co-CeOx Catalysts in the Dry Reforming of Methane: The Important Role of the Reducible Oxide Support and Interactions with Cobalt. ACS Catal. 2018, 8 (4), 3550–3560. https://doi.org/10.1021/acscatal.7b03640. 63. Schilling, C.; Hofmann, A.; Hess, C.; Ganduglia-Pirovano, M. V. Raman Spectra of Polycrystalline CeO2: A Density Functional Theory Study. J. Phys. Chem. C 2017, 121 (38), 20834–20849. https://doi.org/10.1021/acs.jpcc.7b06643. 64. Kosacki, I.; Suzuki, T.; Anderson, H. U.; Colomban, P. Raman Scattering and Lattice Defects in Nanocrystalline CeO2 Thin Films. Solid State Ionics 2002, 149 (1–2), 99–105. https://doi.org/10.1016/S0167-2738(02)00104-2.

Metal-support interfaces in ceria-based catalysts

269

65. Schilling, C.; Hess, C. Elucidating the Role of Support Oxygen in the Water-Gas Shift Reaction over Ceria-Supported Gold Catalysts Using Operando Spectroscopy. ACS Catal. 2019, 9 (2), 1159–1171. https://doi.org/10.1021/acscatal.8b04536. 66. Choi, Y. M.; Abernathy, H.; Chen, H. T.; Lin, M. C.; Liu, M. Characterization of O2-CeO2 Interactions Using In Situ Raman Spectroscopy and First-Principle Calculations. ChemPhysChem 2006, 7 (9), 1957–1963. https://doi.org/10.1002/cphc.200600190. 67. Loridant, S. Raman Spectroscopy as a Powerful Tool to Characterize Ceria-Based Catalysts. Catal. Today 2020. https://doi.org/10.1016/j.cattod.2020.03.044. 68. Wu, Z.; Li, M.; Mullins, D. R.; Overbury, S. H. Probing the Surface Sites of CeO2 Nanocrystals with Well-Defined Surface Planes Via Methanol Adsorption and Desorption. ACS Catal. 2012, 2 (11), 2224–2234. https://doi.org/10.1021/cs300467p. 69. Binet, C.; Daturi, M.; Lavalley, J. C. IR Study of Polycrystalline Ceria Properties in Oxidised and Reduced States. Catal. Today 1999, 50 (2), 207–225. https://doi.org/10.1016/ S0920-5861(98)00504-5. 70. Parastaev, A.; Muravev, V.; Huertas Osta, E.; van Hoof, A. J. F.; Kimpel, T. F.; Kosinov, N.; Hensen, E. J. M. Boosting CO2 Hydrogenation Via Size-Dependent Metal–Support Interactions in Cobalt/Ceria-Based Catalysts. Nat. Catal. 2020, 1–8. https://doi.org/10.1038/s41929-020-0459-4. 71. Zhang, F.; Wang, P.; Koberstein, J.; Khalid, S.; Chan, S. W. Cerium Oxidation State in Ceria Nanoparticles Studied with X-Ray Photoelectron Spectroscopy and Absorption near Edge Spectroscopy. Surf. Sci. 2004, 563 (1–3), 74–82. https://doi.org/10.1016/j.susc.2004.05.138. 72. Salmeron, M.; Schlögl, R. Ambient Pressure Photoelectron Spectroscopy: A New Tool for Surface Science and Nanotechnology. Surf. Sci. Rep. 2008, 169–199. https:// doi.org/10.1016/j.surfrep.2008.01.001. North-Holland. 73. Zhang, C.; Grass, M. E.; McDaniel, A. H.; Decaluwe, S. C.; El Gabaly, F.; Liu, Z.; McCarty, K. F.; Farrow, R. L.; Linne, M. A.; Hussain, Z.; Jackson, G. S.; Bluhm, H.; Eichhorn, B. W. Measuring Fundamental Properties in Operating Solid Oxide Electrochemical Cells by Using In Situ X-Ray Photoelectron Spectroscopy. Nat. Mater. 2010, 9 (11), 944–949. https://doi.org/10.1038/nmat2851. 74. Kato, S.; Ammann, M.; Huthwelker, T.; Paun, C.; Lampimäki, M.; Lee, M.-T.; Rothensteiner, M.; van Bokhoven, J. A. Quantitative Depth Profiling of Ce3þ in Pt/CeO2 by In Situ High-Energy XPS in a Hydrogen Atmosphere. Phys. Chem. Chem. Phys. 2015, 17 (7), 5078–5083. https://doi.org/10.1039/C4CP05643D. 75. Skála, T.; Sutara, F.; Prince, K. C.; Matolín, V. Cerium Oxide Stoichiometry Alteration Via Sn Deposition: Influence of Temperature. J. Electron Spectrosc. Relat. Phenom. 2009, 169 (1), 20–25. https://doi.org/10.1016/j.elspec.2008.10.003. 76. Cafun, J.-D.; Kvashnina, K. O.; Casals, E.; Puntes, V. F.; Glatzel, P. Absence of Ce3þ Sites in Chemically Active Colloidal Ceria Nanoparticles. ACS Nano 2013, 7 (12), 10726–10732. https://doi.org/10.1021/nn403542p. 77. Gänzler, A. M.; Casapu, M.; Maurer, F.; Störmer, H.; Gerthsen, D.; Ferré, G.; Vernoux, P.; Bornmann, B.; Frahm, R.; Murzin, V.; Nachtegaal, M.; Votsmeier, M.; Grunwaldt, J. D. Tuning the Pt/CeO2 Interface by In Situ Variation of the Pt Particle Size. ACS Catal. 2018, 8 (6), 4800–4811. https://doi.org/10.1021/acscatal.8b00330. 78. della Mea, G. B.; Matte, L. P.; Thill, A. S.; Lobato, F. O.; Benvenutti, E. V.; Arenas, L. T.; Jürgensen, A.; Hergenröder, R.; Poletto, F.; Bernardi, F. Tuning the Oxygen Vacancy Population of Cerium Oxide (CeO2  x, 0 < X < 0.5) Nanoparticles. Appl. Surf. Sci. 2017, 422, 1102–1112. https://doi.org/10.1016/j.apsusc.2017.06.101. 79. Kopelent, R.; van Bokhoven, J. A.; Nachtegaal, M.; Szlachetko, J.; Safonova, O. V. X-Ray Emission Spectroscopy: Highly Sensitive Techniques for Time-Resolved Probing of Cerium Reactivity under Catalytic Conditions. Phys. Chem. Chem. Phys. 2016, 18 (47), 32486–32493. https://doi.org/10.1039/c6cp05830b. 80. Turner, S.; Lazar, S.; Freitag, B.; Egoavil, R.; Verbeeck, J.; Put, S.; Strauven, Y.; van Tendeloo, G. High Resolution Mapping of Surface Reduction in Ceria Nanoparticles. Nanoscale 2011, 3 (8), 3385–3390. https://doi.org/10.1039/c1nr10510h. 81. Yoshida, H.; Kuwauchi, Y.; Jinschek, J. R.; Sun, K.; Tanaka, S.; Kohyama, M.; Shimada, S.; Haruta, M.; Takeda, S. Visualizing Gas Molecules Interacting with Supported Nanoparticulate Catalysts at Reaction Conditions. Science 2012, 335 (6066), 317–319. https://doi.org/10.1126/science.1213194. 82. Schilling, C.; Hess, C. Real-Time Observation of the Defect Dynamics in Working Au/CeO2 Catalysts by Combined Operando Raman/UV-Vis Spectroscopy. J. Phys. Chem. C 2018, 122 (5), 2909–2917. https://doi.org/10.1021/acs.jpcc.8b00027. 83. Rakhmatullin, R. M.; Semashko, V. V.; Korableva, S. L.; Kiiamov, A. G.; Rodionov, A. A.; Tschaggelar, R.; van Bokhoven, J. A.; Paun, C. EPR Study of Ceria Nanoparticles Containing Different Concentration of Ce3þ Ions. Mater. Chem. Phys. 2018, 219, 251–257. https://doi.org/10.1016/j.matchemphys.2018.08.028. 84. Wang, M.; Wu, X.-P.; Zheng, S.; Zhao, L.; Li, L.; Shen, L.; Gao, Y.; Xue, N.; Guo, X.; Huang, W.; Gan, Z.; Blanc, F.; Yu, Z.; Ke, X.; Ding, W.; Gong, X.-Q.; Grey, C. P.; Peng, L. Identification of Different Oxygen Species in Oxide Nanostructures with 17O Solid-State NMR Spectroscopy. Sci. Adv. 2015. https://doi.org/10.1126/sciadv.1400133. 85. Tauster, S. J.; Fung, S. C.; Garten, R. L. Strong Metal-Support Interactions. Group 8 Noble Metals Supported on TiO2. J. Am. Chem. Soc. 1978, 100 (1), 170–175. https:// doi.org/10.1021/ja00469a029. 86. van Deelen, T. W.; Hernández Mejía, C.; de Jong, K. P. Control of Metal-Support Interactions in Heterogeneous Catalysts to Enhance Activity and Selectivity. Nat. Catal. 2019, 955–970. https://doi.org/10.1038/s41929-019-0364-x. Nature Publishing Group. 87. Campbell, C. T. Catalyst–Support Interactions: Electronic Perturbations. Nat. Chem. 2012. https://doi.org/10.1038/nchem.1412. 88. Bruix, A.; Rodriguez, J. A.; Ramírez, P. J.; Senanayake, S. D.; Evans, J.; Park, J. B.; Stacchiola, D.; Liu, P.; Hrbek, J.; Illas, F. A New Type of Strong Metal-Support Interaction and the Production of H2 through the Transformation of Water on Pt/CeO2(111) and Pt/CeOx/TiO2(110) Catalysts. J. Am. Chem. Soc. 2012, 134 (21), 8968–8974. https:// doi.org/10.1021/ja302070k. 89. Lykhach, Y.; Kozlov, S. M.; Skála, T.; Tovt, A.; Stetsovych, V.; Tsud, N.; Dvorák, F.; Johánek, V.; Neitzel, A.; Myslivecek, J.; Fabris, S.; Matolín, V.; Neyman, K. M.; Libuda, J. Counting Electrons on Supported Nanoparticles. Nat. Mater. 2016, 15 (3), 284–288. https://doi.org/10.1038/nmat4500. 90. Szabová, L.; Skála, T.; Matolínová, I.; Fabris, S.; Farnesi Camellone, M.; Matolín, V. Copper-Ceria Interaction: A Combined Photoemission and DFT Study. Appl. Surf. Sci. 2013, 267, 12–16. https://doi.org/10.1016/j.apsusc.2012.04.098. 91. Khoobiar, S. Particle to Particle Migration of Hydrogen Atoms on Platinum-Alumina Catalysts from Particle to Neighboring Particles. J. Phys. Chem. 1964, 411–412. https:// doi.org/10.1021/j100784a503. McGraw-Hill Book Co., Inc. 92. Prins, R. Hydrogen Spillover. Facts and Fiction. Chem. Rev. 2012, 2714–2738. https://doi.org/10.1021/cr200346z. American Chemical Society. 93. Karim, W.; Spreafico, C.; Kleibert, A.; Gobrecht, J.; VandeVondele, J.; Ekinci, Y.; van Bokhoven, J. A. Catalyst Support Effects on Hydrogen Spillover. Nature 2017, 541 (7635), 68–71. https://doi.org/10.1038/nature20782. 94. Lykhach, Y.; Staudt, T.; Vorokhta, M.; Skála, T.; Johánek, V.; Prince, K. C.; Matolín, V.; Libuda, J. Hydrogen Spillover Monitored by Resonant Photoemission Spectroscopy. J. Catal. 2012, 285 (1), 6–9. https://doi.org/10.1016/j.jcat.2011.09.002. 95. Zafiris, G. S.; Gorte, R. J. Evidence for Low-Temperature Oxygen Migration from Ceria to Rh. J. Catal. 1993, 139 (2), 561–567. https://doi.org/10.1006/jcat.1993.1049. 96. Smirnov, M. Y.; Graham, G. W. Pd Oxidation under UHV in a Model Pd/Ceria-Zirconia Catalyst. Catal. Lett. 2001, 72 (1–2), 39–44. https://doi.org/10.1023/ A:1009092005849. 97. Cargnello, M.; Doan-Nguyen, V. V. T.; Gordon, T. R.; Diaz, R. E.; Stach, E. A.; Gorte, R. J.; Fornasiero, P.; Murray, C. B. Control of Metal Nanocrystal Size Reveals MetalSupport Interface Role for Ceria Catalysts. Science 2013, 341 (6147), 771–773. https://doi.org/10.1126/science.1240148. 98. Chen, Y.; Chen, J.; Qu, W.; George, C.; Aouine, M.; Vernoux, P.; Tang, X. Well-Defined Palladium-Ceria Interfacial Electronic Effects Trigger CO Oxidation. Chem. Commun. 2018, 54 (72), 10140–10143. https://doi.org/10.1039/C8CC04935A. 99. Kopelent, R.; van Bokhoven, J. A.; Szlachetko, J.; Edebeli, J.; Paun, C.; Nachtegaal, M.; Safonova, O.v. Catalytically Active and Spectator Ce3þ in Ceria-Supported Metal Catalysts. Angew. Chem. Int. Ed. 2015, 54 (30), 8728–8731. https://doi.org/10.1002/anie.201503022. 100. Chen, A.; Yu, X.; Zhou, Y.; Miao, S.; Li, Y.; Kuld, S.; Sehested, J.; Liu, J.; Aoki, T.; Hong, S.; Camellone, M. F.; Fabris, S.; Ning, J.; Jin, C.; Yang, C.; Nefedov, A.; Wöll, C.; Wang, Y.; Shen, W. Structure of the Catalytically Active Copper–Ceria Interfacial Perimeter. Nat. Catal. 2019, 2 (4), 334–341. https://doi.org/10.1038/s41929-019-0226-6. 101. Farmer, J. A.; Campbell, C. T. Ceria Maintains Smaller Metal Catalyst Particles by Strong Metal-Support Bonding. Science 2010, 329 (5994), 933–936. https://doi.org/ 10.1126/science.1191778.

270

Metal-support interfaces in ceria-based catalysts

102. Hemmingson, S. L.; Campbell, C. T. Trends in Adhesion Energies of Metal Nanoparticles on Oxide Surfaces: Understanding Support Effects in Catalysis and Nanotechnology. ACS Nano 2017, 11 (2), 1196–1203. https://doi.org/10.1021/acsnano.6b07502. 103. Nagai, Y.; Dohmae, K.; Ikeda, Y.; Takagi, N.; Tanabe, T.; Hara, N.; Guilera, G.; Pascarelli, S.; Newton, M. A.; Kuno, O.; Jiang, H.; Shinjoh, H.; Matsumoto, S. In Situ Redispersion of Platinum Autoexhaust Catalysts: An on-Line Approach to Increasing Catalyst Lifetimes? Angew. Chem. Int. Ed. 2008, 47 (48), 9303–9306 https://doi.org/ 10.1002/anie.200803126. 104. Aitbekova, A.; Wu, L.; Wrasman, C. J.; Boubnov, A.; Hoffman, A. S.; Goodman, E. D.; Bare, S. R.; Cargnello, M. Low-Temperature Restructuring of CeO2-Supported Ru Nanoparticles Determines Selectivity in CO2 Catalytic Reduction. J. Am. Chem. Soc. 2018, 140 (42), 13736–13745. https://doi.org/10.1021/jacs.8b07615. 105. Newton, M. A.; Belver-Coldeira, C.; Martínez-Arias, A.; Fernández-García, M. Dynamic In Situ Observation of Rapid Size and Shape Change of Supported Pd Nanoparticles during CO/NO Cycling. Nat. Mater. 2007, 6 (7), 528–532. https://doi.org/10.1038/nmat1924. 106. Wang, W. W.; Yu, W. Z.; Du, P. P.; Xu, H.; Jin, Z.; Si, R.; Ma, C.; Shi, S.; Jia, C. J.; Yan, C. H. Crystal Plane Effect of Ceria on Supported Copper Oxide Cluster Catalyst for CO Oxidation: Importance of Metal-Support Interaction. ACS Catal. 2017, 7 (2), 1313–1329. https://doi.org/10.1021/acscatal.6b03234. 107. Wang, F.; Li, C.; Zhang, X.; Wei, M.; Evans, D. G.; Duan, X. Catalytic Behavior of Supported Ru Nanoparticles on the {100}, {110}, and {111} Facet of CeO2. J. Catal. 2015, 329 (1), 177–186. https://doi.org/10.1016/j.jcat.2015.05.014. 108. Zhu, J.; Su, Y.; Chai, J.; Muravev, V.; Kosinov, N.; Hensen, E. J. M. Mechanism and Nature of Active Sites for Methanol Synthesis from CO/CO2 on Cu/CeO2. ACS Catal. 2020, 10 (19), 11532–11544. https://doi.org/10.1021/acscatal.0c02909. 109. Guzman, J.; Carrettin, S.; Corma, A. Spectroscopic Evidence for the Supply of Reactive Oxygen during CO Oxidation Catalyzed by Gold Supported on Nanocrystalline CeO2. J. Am. Chem. Soc. 2005, 127 (10), 3286–3287. https://doi.org/10.1021/ja043752s. 110. Yeung, C. M. Y.; Yu, K. M. K.; Fu, Q. J.; Thompsett, D.; Petch, M. I.; Tsang, S. C. Engineering Pt in Ceria for a Maximum Metal-Support Interaction in Catalysis. J. Am. Chem. Soc. 2005, 127 (51), 18010–18011. https://doi.org/10.1021/ja056102c. 111. Yamada, Y.; Tsung, C.-K.; Huang, W.; Huo, Z.; Habas, S. E.; Soejima, T.; Aliaga, C. E.; Somorjai, G. A.; Yang, P. Nanocrystal Bilayer for Tandem Catalysis. Nat. Chem. 2011, 3 (5), 372–376. https://doi.org/10.1038/nchem.1018. 112. Wang, A.; Li, J.; Zhang, T. Heterogeneous Single-Atom Catalysis. Nat. Rev. Chem. 2018, 1. https://doi.org/10.1038/s41570-018-0010-1. 113. Qiao, B.; Wang, A.; Yang, X.; Allard, L. F.; Jiang, Z.; Cui, Y.; Liu, J.; Li, J.; Zhang, T. Single-Atom Catalysis of CO Oxidation Using Pt1/FeOx. Nat. Chem. 2011, 3 (8), 634–641. https://doi.org/10.1038/nchem.1095. 114. Gates, B. C. Supported Metal Clusters: Synthesis, Structure, and Catalysis. Chem. Rev. 1995, 95 (3), 511–522. https://doi.org/10.1021/cr00035a003. 115. Flytzani-Stephanopoulos, M.; Gates, B. C. Atomically Dispersed Supported Metal Catalysts. Annu. Rev. Chem. Biomol. Eng. 2012, 3 (1), 545–574. https://doi.org/10.1146/ annurev-chembioeng-062011-080939. 116. Beniya, A.; Higashi, S. Towards Dense Single-Atom Catalysts for Future Automotive Applications. Nat. Catal. 2019, 2 (7), 590–602. https://doi.org/10.1038/s41929-0190282-y. 117. Resasco, J.; Derita, L.; Dai, S.; Chada, J. P.; Xu, M.; Yan, X.; Finzel, J.; Hanukovich, S.; Hoffman, A. S.; Graham, G. W.; Bare, S. R.; Pan, X.; Christopher, P. Uniformity is Key in Defining Structure-Function Relationships for Atomically Dispersed Metal Catalysts: The Case of Pt/CeO2. J. Am. Chem. Soc. 2020, 142 (1), 169–184. https://doi.org/ 10.1021/jacs.9b09156. 118. Fu, Q. Active Nonmetallic Au and Pt Species on Ceria-Based Water-Gas Shift Catalysts. Science 2003, 301 (5635), 935–938. https://doi.org/10.1126/science.1085721. 119. Jones, J.; Xiong, H.; DeLaRiva, A. T.; Peterson, E. J.; Pham, H.; Challa, S. R.; Qi, G.; Oh, S.; Wiebenga, M. H.; Pereira Hernandez, X. I.; Wang, Y.; Datye, A. K. Thermally Stable Single-Atom Platinum-on-Ceria Catalysts Via Atom Trapping. Science 2016, 353 (6295), 150–154. https://doi.org/10.1126/science.aaf8800. 120. Dvorák, F.; Camellone, M. F.; Tovt, A.; Tran, N.; Negreiros, F. R.; Vorokhta, M.; Skála, T.; Matolínová, I.; Myslivecek, J.; Matolín, V.; Fabris, S. Creating Single-Atom Pt-Ceria Catalysts by Surface Step Decoration. Nat. Commun. 2016, 7 (10801), 1–8. https://doi.org/10.1038/ncomms10801. 121. Kunwar, D.; Zhou, S.; Delariva, A.; Peterson, E. J.; Xiong, H.; Pereira-Hernández, X. I.; Purdy, S. C.; ter Veen, R.; Brongersma, H. H.; Miller, J. T.; Hashiguchi, H.; Kovarik, L.; Lin, S.; Guo, H.; Wang, Y.; Datye, A. K. Stabilizing High Metal Loadings of Thermally Stable Platinum Single Atoms on an Industrial Catalyst Support. ACS Catal. 2019, 9 (5), 3978–3990. https://doi.org/10.1021/acscatal.8b04885. 122. Neitzel, A.; Figueroba, A.; Lykhach, Y.; Skala, T.; Vorokhta, M.; Tsud, N.; Mehl, S.; Sevcikova, K.; Prince, K. C.; Neyman, K. M.; Matolin, V.; Libuda, J. Atomically Dispersed Pd, Ni, and Pt Species in Ceria-Based Catalysts: Principal Differences in Stability and Reactivity. J. Phys. Chem. C 2016, 120 (18), 9852–9862. https://doi.org/10.1021/ acs.jpcc.6b02264. 123. Pereira-Hernández, X. I.; DeLaRiva, A.; Muravev, V.; Kunwar, D.; Xiong, H.; Sudduth, B.; Engelhard, M.; Kovarik, L.; Hensen, E. J. M.; Wang, Y.; Datye, A. K. Tuning Pt-CeO2 Interactions by High-Temperature Vapor-Phase Synthesis for Improved Reducibility of Lattice Oxygen. Nat. Commun. 2019, 10 (1), 1358. https://doi.org/10.1038/s41467019-09308-5. 124. Wang, H.; Liu, J. X.; Allard, L. F.; Lee, S.; Liu, J.; Li, H.; Wang, J.; Wang, J.; Oh, S. H.; Li, W.; Flytzani-Stephanopoulos, M.; Shen, M.; Goldsmith, B. R.; Yang, M. Surpassing the Single-Atom Catalytic Activity Limit through Paired Pt-O-Pt Ensemble Built from Isolated Pt1 Atoms. Nat. Commun. 2019, 10 (1), 1–12. https://doi.org/10.1038/s41467019-11856-9. 125. Maurer, F.; Jelic, J.; Wang, J.; Gänzler, A.; Dolcet, P.; Wöll, C.; Wang, Y.; Studt, F.; Casapu, M.; Grunwaldt, J.-D. Tracking the Formation, Fate and Consequence for Catalytic Activity of Pt Single Sites on CeO2. Nat. Catal. 2020, 3 (10), 824–833. https://doi.org/10.1038/s41929-020-00508-7. 126. Jeong, H.; Shin, D.; Kim, B.-S.; Bae, J.; Shin, S.; Choe, C.; Han, J. W.; Lee, H. Controlling Oxidation State of Pt Single-Atoms for Maximizing Catalytic Activity. Angew. Chem. Int. Ed. 2020. https://doi.org/10.1002/anie.202009776. anie.202009776. 127. Ma, Y.; Chi, B.; Liu, W.; Cao, L.; Lin, Y.; Zhang, X.; Ye, X.; Wei, S.; Lu, J. Tailoring of the Proximity of Platinum Single Atoms on CeO2 Using Phosphorus Boosts the Hydrogenation Activity. ACS Catal. 2019, 9 (9), 8404–8412. https://doi.org/10.1021/acscatal.9b01536. 128. Spezzati, G.; Su, Y.; Hofmann, J. P.; Benavidez, A. D.; DeLaRiva, A. T.; McCabe, J.; Datye, A. K.; Hensen, E. J. M. Atomically Dispersed Pd-O Species on CeO2(111) as Highly Active Sites for Low-Temperature CO Oxidation. ACS Catal. 2017, 7 (10), 6887–6891. https://doi.org/10.1021/acscatal.7b02001. 129. Ye, L.; Mahadi, A. H.; Saengruengrit, C.; Qu, J.; Xu, F.; Fairclough, S. M.; Young, N.; Ho, P.-L.; Shan, J.; Nguyen, L.; Tao, F. F.; Tedsree, K.; Tsang, S. C. E. Ceria Nanocrystals Supporting Pd for Formic Acid Electrocatalytic Oxidation: Prominent Polar Surface Metal Support Interactions. ACS Catal. 2019, 9 (6), 5171–5177. https:// doi.org/10.1021/acscatal.9b00421. 130. Jeong, H.; Kwon, O.; Kim, B. S.; Bae, J.; Shin, S.; Kim, H. E.; Kim, J.; Lee, H. Highly Durable Metal Ensemble Catalysts with Full Dispersion for Automotive Applications beyond Single-Atom Catalysts. Nat. Catal. 2020, 1–8. https://doi.org/10.1038/s41929-020-0427-z. 131. Datye, A. K.; Votsmeier, M. Opportunities and Challenges in the Development of Advanced Materials for Emission Control Catalysts. Nat. Mater. 2020. https://doi.org/10.1038/ s41563-020-00805-3.

6.12

Solid acid catalysis; Part I, the zeolite protonic site

Rutger A. van Santen, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, MB Eindhoven, The Netherlands © 2023 Elsevier Ltd. All rights reserved.

6.12.1 6.12.2 6.12.2.1 6.12.2.1.1 6.12.2.1.2 6.12.2.2 6.12.2.2.1 6.12.2.2.2 6.12.2.2.3 6.12.2.3 6.12.2.3.1 6.12.2.3.2 6.12.2.4 References

Introduction The proton strength of the zeolite Brønsted acid The physical chemistry of the protonic bond Vibrational spectroscopy of zeolite hydroxyls The proton bond as a function of zeolite lattice Al/Si concentration ratio The OH chemical bond The deprotonation energy Flexibility of zeolite lattice Site dependence Proton activity of other than Al/Si framework materials Al substitution by Fe3þ and Ga3þ Non framework substituted systems In summary

271 274 274 274 277 279 279 281 282 285 285 286 289 290

Abstract Physical chemical properties of the zeolite proton are reviewed as deduced from spectroscopy, quantum-chemical calculation and kinetic studies. Emphasis is on relation of zeolite proton affinity with zeolite composition and structure. As long as Al/ (Al þ Si) zeolite composition ratio is less than 10% and protonic zeolites are not promoted by exchange with inorganic cations proton affinity is not composition or structure dependent. However for higher Al/(Al þ Si) ratio’s the dependence of proton affinity on structure can be significant. Proton reactivity may strongly depend on zeolite structure when in contact with a reactant molecule. Then match of molecular dimensions and zeolite cavity becomes important. This is the confinement effect.

6.12.1

Introduction

This chapter is part I of two chapters, where we will discuss the catalytic reactivity of zeolite solid acid catalysts. In this chapter we will focus on the inorganic chemistry of the zeolite proton. In part II, proton reactivity and catalysis of proton catalyzed reactions will be discussed. In this introduction a general background on protonic zeolites and their catalytic use is given. These solid acid catalysts introduced and discovered 70 years ago are currently widely applied in refinery processes where crude oil or natural gas is converted into gasoline or base chemicals. Over the past 50 years the relation between the inorganic chemistry of zeolites and their catalytic functionality has become increasingly better understood. In the case of the zeolites understanding of the catalyst structure and catalytic performance relation is facilitated because of their well-defined internal surface and the single site nature of the catalytic reactive site. Zeolite crystallites have a well-defined microporous structure, that is accessible to reactant molecules. Dimensions of the zeolite micro pore are comparable to that of organic molecules, but there can be a mismatch of molecular size and form and zeolite cavity. This makes catalyst reactivity micro pore structure dependent and is the cause of shape selectivity. We will investigate the validity of two approximate reactivity rules:

• •

The reactivity of protons as characterized by their proton affinity (PA) does not depend on zeolite structure and is primarily defined by zeolite composition. The proton affinity can be calculated from the deprotonation energy (DPE). The dependence of catalysis on zeolite structure is dominated by match of molecular size and shape of reaction intermediates with zeolite microporous structure. This is called the confinement effect. It does not relate directly to proton reactivity but rather to physical adsorption of molecules in the zeolite microporous cavities.

In this chapter the conditions of validity of the first rule will be extensively discussed. In Part II the confinement effect is major topic of discussion.

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00009-1

271

272

Solid acid catalysis; Part I, the zeolite protonic site

In mineralogy zeolites are known as alumina-silicates. They consist of a negatively charged alumino-silicate network that is stabilized by positively charged alkali or earth alkali cations. The alumino-silicate network creates a microstructure of micro channels and cavities in which the cations are located.1,2 There is also a rich variety of synthetic inorganic materials with a structure that is related to the zeolite framework topologies.3 These materials can have variable composition, different from the alumino-silicates. By chemical treatment zeolites and the corresponding zeolite materials can be converted into Brønsted acidic materials. As such they are extensively used in refinery processes as solid acid catalysts. They are robust materials and useful since they can be used at temperatures needed for many reactions where liquid acids cannot be applied.3–6 Since their early application in refinery processes zeolite catalysts also found additional use in a wide variety of other chemical processes, where they are often promoted with a cation or transition metal cluster. The zeolite network is constructed according to an elementary connection principle: it is a network connected through the tetrahedral vertices. In the center of the tetrahedron a four valent cation as Si4þ or Al3þ is located. The vertices of the tetrahedron are four oxygen atom anions. Currently there are 145 materials with different zeolite frameworks known and many more theoretical structures, based on their tetrahedral network have been proposed.7–9 In case the tetrahedron is composed of a Si cation surrounded by four negatively charged oxygen atoms, the zeolite network will be electrostatically neutral. It will have the SiO2 stoichiometry similar as that of quartz of which it can be considered a polymorph. Such siliceous materials are catalytically unreactive. When in a zeolite tetrahedron the formally charged Si4þ cation is replaced by a cation of lower charge as for instance the often used Al3þ cation, the zeolite framework gets a negative charge. In a stable structure this is to be compensated by a cation, as for instance an alkali or earth alkali cation. The material is converted into the solid acid form when by an ion exchange process the inorganic cation is exchanged with an ammonium cation. When in a subsequent step by heating the material ammonia desorbs, it will leave behind a proton attached to the zeolite lattice. Two representative zeolite structures, FAU and MFI, that are the network structures of Zeolite X or Y and ZSM-5 respectively of use in several important catalytic reactions are shown in Fig. 1. The structures of these zeolites can be understood as a network of respectively four, five and six membered tetrahedral rings. The Faujasite (FAU) structure contains large cavities, the super cages, that are connected through 12 membered ring openings. The MFI structure has channels with a 10 membered tetrahedral ring diameter. Other zeolite structures as TON or MOR structure8 have one dimensional channels with respectively 10 or 12 ring openings. Zeolite catalysts became useful as solid acid catalysts due to the discovery that protons attached to the crystalline zeolite framework are catalytically highly reactive.10,11 The first major application of zeolites as solid acid catalyst, discovered by Plank and Rosinsky,12 is the catalytic cracking process, that converts heavy oil molecules into lighter molecules. The zeolites used in this process have the FAU structure of zeolite Y or X. It had a large economic impact because due to reduced coke coproduction the yield of crude oil into gasoline becomes largely increased. The zeolite microporous channel structure suppresses formation of the large aromatic molecules (shape selectivity), that are coke formation precursors. Later followed the invention of other important refinery processes as catalytic hydrocracking and isomerization reactions.13,14 There is a rich variation in the composition of the zeolite that affects its performance as solid acid catalyst. Not only can the Si/Al atom zeolite framework ratio be varied over a large concentration interval, also the tetrahedral [AlO4] unit can be replaced by a tetrahedral [GaO4] or [FeO4] units or [SiO4]by a [PO4]unit. In the latter case, since charge neutrality has to be maintained two neighboring [SiO4] units are to be replaced by an [AlO4]unit and [PO4] unit. Such synthetic silico-alumino- phosphate catalysts (SAPO4)15 are used for the conversion of methanol into short alkenes or aromatics.10,16 This reaction is important reaction step in the overall reaction of natural gas to base chemicals or liquid fuels.17,18 The current availability of zeolite materials with a very rich variety of structures and a large variation in composition owns importantly to the use of organic template molecules as base in low temperature zeolite synthesis.18 The large organic base molecules used as template molecule adsorb into the channels and cavities of the crystallizing zeolite. Their charge is low, but due to their size they can extend over a substantial part of the zeolite micro pore. Their interaction with the zeolite framework reduces the need for a large concentration micro-pore charge compensating cations and hence will reduce Al/Si ratio. Organic templates used in synthesis are also structure directing. Their use has led to the discovery of several new zeolite related structures. For instance the MFI zeolite structure, shown in Fig. 1D, has been discovered using tetra-propyl cations as organic base. As we will discuss below the proton strength in zeolites with a low Al/Si ratio is substantially larger than in zeolites with an Al/ Si ratio near one. The now generally accepted structure of the reactive zeolite protonic site has been first proposed in 1963. This was based mainly on infrared spectroscopic experimental data.19 As is illustrated in Fig. 1B the zeolite proton site consists of a proton that is attached to an oxygen atom that bridges two tetrahedra that contain respectively a formally four valent or three valent cation. This a unique configuration and very different from the one-fold coordination of a silanol group (see Fig. 1E) that would result from the hydrolysis of a [SiOSi] connection. The silanols dominate the external surface of a zeolite crystallite. The internal reactive zeolite proton can be considered as a Lewis acid promoted silanol. The relation between proton reactivity and catalyst reactivity we will discuss in Part II. Here we will study the reactivity of the isolated zeolite proton defined as its proton affinity (PA).

Solid acid catalysis; Part I, the zeolite protonic site

273

Fig. 1 Fig. A illustrates the FAU zeolite network that consists of double six rings and sodalite cavities that are formed from four and six tetrahedra connecting rings. Fig. B shows part of the FAU network where Si4þ framework cations are substituted by Al3þ cations Si/Al ¼47). The negative charge generated by this substitution is compensated for by protons that locate at an oxygen atom that connects a Si4þ and Al3þ cation. Fig. C is a space filling model that shows oxygen atoms lining the zeolite MFI network microporous wall. Fig. D is a three dimensional microchannel network model of MFI that shows how a micro channel is built from four, five or six tetrahedra connecting rings. The walls of the channels are formed by 10 tetrahedral rings. Fig. E is a tree-like model of a surface silanol. (A) Reproduced with permission Liu, C.; Li, G.; Hensen, E.J.M.; Pidko, E.A. Nature and Catalytic Role of Extraframework Aluminum in Faujasite Zeolite: A Theoretical Perspective. ACS Catal. 2015, 5(11), 7024–7033, doi: 10.1021/ acscatal.5b02268. (B) Reproduced with permission Liu, C.; Tranca, I. C.; Santen, V.; Hensen, R. A.; Liu, Pidko, C.; Tranca, I.; Van Santen, R. A.; Hensen, E. J. M.; Pidko, E. A. Scaling Relations for Acidity and Reactivity of Zeolites. J. Phys. Chem. C 2017, 121(42), 23520–23530, doi: 10.1021/ acs.jpcc.7b08176. (C) Reproduced with permission Plank, C. J. The Invention of Zeolite Cracking CatalystsdA Personal Viewpoint, pp. 253–271. ACS Publications, 1983. (D) Reproduced with permission Olson, D. H.; Kokotailo, G. T.; Lawton, S. L.; Meier, W. M. Crystal Structure and StructureRelated Properties of ZSM-5. J. Phys. Chem., 1981, 85(15), 2238–2243, doi: 10.1021/j150615a020. (E) Reproduced with permission Sauer, J.; Hill, J. R. The Acidity of Surface Silanol Groups. A Theoretical Estimate Based on Ab Initio Calculations on a Model Surface. Chem. Phys. Lett. 1994, 218(4), 333–337, doi: 10.1016/0009-2614(94)00002-6.

It is useful to compare the definition of acidity in a homogenous phase with that of the zeolite proton. The thermodynamic definition of acidity in the homogenous acidic solutions is described by the Hammett function Ho20:   ½B H0 ¼ pKBHþ þ log (1) ½BHþ  It depends on the equilibrium constant Kp of protonated base probe BHþ and free base B B molecules and their respective concentrations. In Eq. (1) pKBHþ is defined as  log [Kp]. The Hammett function Ho is a generalization of the pH ¼  log [Hþ]. From a measurement of the

½B ½BHþ 

concentration ratio and the proton concentration Kp can be determined.

In the zeolite this relation is to be replaced by the respective concentrations of protons attached to the zeolite [SiOHAl] framework, deprotonated site [SiOAl] 1 and that of the concentration of adsorbed protonated probe molecule BHþ respectively. From a known value of pKBHþ of the probe indicator molecules the pK of the zeolite proton can then be deduced. Ho of a reactive zeolite a highly siliceous protonic zeolite as ZSM-5 (with MFI structure) is typically  10, whereas that of a superacid is  12 to  15.21 The reactivity of the zeolite proton is seen to be considerably lower than that of the proton of the liquid acid. The difference in acidity of the solid acid zeolite and that in a polar liquid is due to the difference in respective dielectric constants. The zeolite is a medium with low dielectric constant of the order of 222 whereas polar liquids as water have a high

274

Solid acid catalysis; Part I, the zeolite protonic site

dielectric constant of the order of 80. The near vacuum condition of the zeolite gives charge separation of the OH bond a large energetic penalty. Whereas acid catalysis in water or liquid acid is a reaction of charged hydronium or protonated acid molecules (e.g. H2Fþ 1) with substrate, in the zeolite the protonating state is electrostatically neutral. This is the fundamental reason of the lower acid strength of the zeolite proton versus that of a liquid acid23 The deprotonation energy of the protonated charged molecule is lower than that in a neutral molecule, since in the former the OH bond cleavage event does not involve additional charge separation. Deprotonation energy (DPE) differences of protons located in different site positions of the zeolite relate to the differences in covalent bond energies of the zeolite OH bond and the relative stability of (SiOAl) 1, the negative charge left on the zeolite lattice. We will argue that this is determined primarily by zeolite composition and that structural differences play a minor role. This may initially appear contradictory to extensive experimental evidence that solid acid catalysis is strongly zeolite structure dependent. In this respect it is important to realize that in addition to reaction with proton the catalytic reaction requires also adsorption of reactant or desorption of product, since there is exchange between the external reaction medium and internal part of the zeolite particle. Therefore, as will be discussed in detail in Part II, catalytic acidity will not only depend on the reactivity of the proton, but also on physical properties as the adsorption free energy of reactants, reaction intermediates or products. The structure dependence of zeolites is too a large extend due the dependence of adsorption energies on size and shape of zeolite micro cavities. Proton affinity as discussed here is solely a zeolite internal property . Its properties can de deduced spectroscopically. Complemented by quantum-chemical calculations this provides the physical chemical understanding of proton affinity difference as a function of zeolite composition, that we will be present in the following sections.

6.12.2

The proton strength of the zeolite Brønsted acid

In Section 6.12.2.1 we will summarize experimental information on the zeolite proton strength. Section 6.12.2.2 provides the relations with chemical bonding properties of the zeolite OH bond. In these subsections we mainly focus on the relation between Al/Si framework concentration and PA. Then proton reactivity of other than Al/Si framework materials will be described. Often zeolite catalysts contain cations, that are added to stabilize the zeolite framework against harsh hydrolyzing conditions of catalyst. Also zeolites may be activated by steaming treatments that induce imperfections. Such promoted zeolite systems will be discussed in the later subsections.

6.12.2.1

The physical chemistry of the protonic bond

In Section 6.12.2.1.1 vibrational spectroscopy of the zeolite proton will be reviewed. It will be shown how this can be used to determine experimentally the proton deprotonation energy (DPE).

6.12.2.1.1

Vibrational spectroscopy of zeolite hydroxyls

Absorbance (abs. u.)

Vibrational spectra of dehydrated protonic zeolites have spectral features in four frequency domains. This is illustrated by Fig. 2.24 All systems show a high frequency band around 3750 cm 1. Only when Al is present a spectral features evolves between 3670 and 3500 cm 1. The high frequency band is from the silanol (see Fig. 1E) hydroxyl, that is mainly located at the external surface of the zeolite crystal. The features below 3670 cm 1 are to be assigned to the zeolite proton. The lowered frequency of the zeolite proton compared to the silanol frequency is indicative of its weaker OH bond strength.

silicalite-1

ZSM-5

supercage

3750

3700

Y

sodalite

3650 3600 3550 Wavenumber (cm-1)

3500

3450

Fig. 2 Infrared frequency regimes of zeolite protons and surface silanol. Comparison of Al-free silanol (MFI, (MFI, N), ZSM-5 (MFI,18), Y(FAU,3)). In parentheses the zeolite topology and Si/Al ratio. After Bordiga, S.; Lamberti, C.; Bonino, F.; Travert, A.; Thibault-Starzyk, F.; Probing Zeolites by Vibrational Spectroscopies. Chem. Soc. Rev. 2015, 44(20), 7262–7341, doi: 10.1039/c5cs00396b.

Solid acid catalysis; Part I, the zeolite protonic site

275

Three proton adsorption intensity regimes can be distinguished. Protons with a higher vibrational frequency at 3650 cm 1. A second regime in the 3650 till 3500 cm 1 region and a lower band regime at 3540 cm 1. The high frequency and low frequency regimes are related. They refer to protons attached to zeolite frame works with relatively high Al/Si ratio. In the high Al/Si zeolite Y (Al/Si ¼ 1/3) that has the FAU structure, the zeolite OH bond spectrum becomes split into two. The protons with the higher frequency of 3650 cm 1 have a stronger bond strength. It indicates that their acidity becomes weaker. A lower vibrational frequency band appears at 3540 cm 1. This weakened OH feature is due to internal through space hydrogen bonding of the zeolite proton with additional framework oxygen atoms that connect nearby Si or Al containing tetrahedra. This proton is not directed outwards of the zeolite wall. The two bands are respectively called the high frequency (HF) and low frequency (LF) bands. Internal hydrogen bonded OH are not unique to the faujasite lattice and are also present many zeolites. The Mordenite and Ferrierite materials are examples. Proton affinity is little affected by internal hydrogen bonding of the protons and remains comparable to that of protons free of additional interactions directed into the zeolite channel.25 For siliceous zeolites with low Al/Si framework concentration the zeolite OH frequency is the lower 3625–3600 cm 1. These vibrational frequencies indicate that isolated protons have lower DPE and hence are more reactive when they are part of a zeolite framework as in ZSM-5 with low Al/Si ratio. The broadening of the silanol absorption band at 3750 cm 1 is also due to internal hydrogen bonding interactions. In this case due to silanol nests created in the zeolite lattice by removal of an Al atom, that we will discuss in Section 6.12.2.3.2.26 As we will discuss in Section 6.12.2.2 the variation in OH frequencies of the siliceous zeolites around 3600 cm 1is due to variation in the [SiOHAl] bond angle, that varies with protons positioned on crystallographically different site positions. Change in bond angle alters hybridization of 2s and 2p atomic orbitals on the oxygen atom and decreases OH bond strength when the [SiOAl] bond angle widens. We will also discuss below the vibrational frequency of the zeolite proton does poorly correlate with differences in DPE as a function of proton siting on crystallographically distinguishable positions of the zeolite framework. This is because the vibrational frequency measures the strength of the OH bond when attached to the zeolite framework, whereas DPE depends on the energy difference of the protonated [SiOHAl] site and the deprotonated site [SiOAl] 1. Whereas the deprotonation energy DPE cannot be deduced from the free proton OH vibrational spectra, as we will discuss now that disturbance of proton vibrational frequencies by adsorbed probe molecules makes this possible.27 Changes in the proton infrared spectra of protonic Mordenite zeolite, that has one dimensional 12 ring diameter micro channels, induced by low temperature adsorption of CO and N2 at low temperature are shown in Figs. 3A and B. In both cases proton intensity of undisturbed surface silanol (3750 cm 1) and zeolite proton (3610 cm 1) is lost and new peaks of higher intensities appear. These changes are due to hydrogen bonding between adsorbed CO and N2 and the surface OH groups. The silanol bond is weakened and its frequency shifts downwards to 3620 cm 1 (CO) and 3710 cm 1 (N2) respectively. The weakened zeolite proton bond when hydrogen bonded to adsorbate shows increased intensity at 3300 cm 1(CO) and 3490 cm 1 (N2) respectively. The smaller downwards frequency shift of the silanol band compared to that of the zeolite proton is due to the stronger OH bond strength of the former. The larger shift upon interaction with CO compared to N2 indicates that the hydrogen bond interaction with CO is larger than that with N2. The electronic nature of hydrogen bonding is complex. Upon Initial contact the doubly electron occupied lone pair orbitals of CO and N2 respectively will have a repulsive interaction with the also doubly electron occupied OH orbital. This repulsive Pauli interaction is overcome by polarization of the electrons of the OH bond and those of the adsorbing molecule away from the hydrogen bond regime. This reduces the repulsive electron density in the hydrogen bond regime. CO binds more strongly than N2 because the polarizability of the CO molecule is the larger. The increase in infrared absorption intensities is due to the increased polarization of the OH bond. The intensity increases square with OH dipole moment. The increase in intensity implies that a positive charge develops on the hydrogen atom upon contact with the probe molecule. The negative charge accumulates on the zeolite lattice oxygen atoms and hence the decrease in bond frequency relates to the degree of stabilization of the negative charge on the zeolite lattice. The latter will depend on the composition of the zeolite lattice. Fig. 3C, that relates the silanol frequency shift with that of zeolite proton for different probe molecules illustrates that as long as proton is not transferred to the probe molecule (at this point the curves become nonlinear), that the ratio of silanol frequency shift and proton shift frequency is a constant.24 It depends slightly on structure, due to small differences in the adsorption energies of the probe molecules. The smaller frequency shift of the HF molecule, in contact with probe molecules in an Ar matrix indicates that the proton bond in the isolated HF molecule is stronger than that of the respective zeolites. This difference with that of superacidic liquid HF is that its acidity is based on reactivity of HFHþ 1. The HF molecule is less reactive that the zeolite proton because of the difference in polarizability of F compared to the zeolite proton environment. Paukhstis et al.27 deduced an empirical formula, based on deprotonation data of high surface area silica and alumino-silicates that correlates DPE or PA’s with the ratio of proton shift frequencies of silanol and zeolite proton site.   DvBronsted 1 kJ=mol (2) PABronsted ¼ PASilanols  Alog ;A ¼ 0:00226 DvSilanols

276

Solid acid catalysis; Part I, the zeolite protonic site

(A)

(B)

(C)

H-MOR

HF:Ar

Fig. 3 Infrared absorption measurements of zeolite OH groups disturbed by weakly interacting adsorbates. (A) O–H stretching region of H–MOR (spectrum 1) and the effect of increasing doses of adsorbed CO (spectra 2–9) at equilibrium pressure from ca. 10–2 to 20 Torr (1 Torr ¼ 133.3 Pa). Spectra were taken at 77. (B) Effect of adsorbed N2, other details as in (A). (C) Schematic plot of frequency shift of silanol (Dv(OH)Silanol) versus that of Brønsted site (Dv(OH)Brønsted) as deduced from probe molecules with increasing stronger interaction. Red line, ZSM-5, black line Y, blue line HF in Ar. Vertical broken line denotes silanol frequency shift where for zeolite proton hydrogen bonding switches to ionic bonding due to proton transfer. The Dv(OH)Brønsted frequencies then show an upwards directed deviation. (A) With permission Bordiga, S.; Lamberti, C.; Geobaldo, F.; Zecchina, A.; Palomino, G. T.; Areán, C. O. Fourier-Transform Infrared Study of CO Adsorbed at 77 K on H-Mordenite and Alkali-Metal-Exchanged Mordenites. Langmuir, 1955, 11(2), 527–533, 1995, doi: 10.1021/la00002a027. (B) With permission Geobaldo, F.; Lamberti, C.; Ricchiardi, G.; Bordiga, S.; Zecchina, A.; Palomino, G. T.; Areán, C. O. N2 Adsorption at 77 K on H-Mordenite and Alkali-Metal-Exchanged Mordenites: An IR Study. J. Phys. Chem., 1995, 99, 28, 11167–11177, doi: 10.1021/j100028a018. (C) After Bordiga, S.; Lamberti, C.; Bonino, F.; Travert, A.; Thibault-Starzyk, F.; Probing Zeolites by Vibrational Spectroscopies. Chem. Soc. Rev. 2015, 44(20), 7262–7341, doi: 10.1039/c5cs00396b.

One calculates from Eq. (2) for the PA of the ZSM-5 proton a value of 1190 kJ/mol. This is close to DPE values reported from quantum chemical calculations to be discussed in Section 6.12.2.2. This value is substantially less than the DPE of the surface silanol, that is 1390 kJ/mol.28 It is interesting to compare the proton affinity of the zeolite proton with protonation energy or the proton affinity of interacting base molecules. Whereas acetonitrile with a proton affinity of 788 kJ/mol is only hydrogen bonded and the zeolite OH bond remains intact, pyridine (PA ¼ 930 kJ/mol) or ammonia (PA ¼ 853 kJ/mol) protonate when adsorbed to the zeolite proton. Evidence for interaction through hydrogen bonding versus proton transfer derives from IR spectroscopy. When the zeolite OH becomes strongly perturbed the single peak OH spectrum splits into a three peaked very broad band feature. Part of the adsorption

Solid acid catalysis; Part I, the zeolite protonic site

277

intensity is stolen by resonance with dipole transition forbidden overtones of in plane and out of plane OH bending modes. Their disappearance is indicative for proton transfer.29 Proton transfer between adsorbed base molecule and zeolite OH happens when the bond energy HBþ of the protonated base molecule exceeds the DPE energy that has to be corrected with the stabilization energy due to the attraction of the positively charged protonated base molecule and negative charge left on the zeolite lattice. For instance Ammonia that has a protonation energy of 854 kJ/mol30 and pyridine that has a protonation energy of 930 kJ/ mol31 become readily protonated. Whereas these two molecules are basic enough to become protonated this is not the case for acetonitrile (PA ¼ 779 KJ/mol or H2O (PA ¼ 691 kJ/mol)32 or H2O.24 Interestingly whereas the zeolite proton is not reactive enough to protonate a single water molecule, a cluster of water molecules will become protonated.33 Hydrogen bonding within the water cluster stabilizes the hydronium ion formation energy. Comparison of base molecule proton affinities with zeolite OH proton affinity shows that for proton transfer an energy cost of the order of 300 kJ/mol has been overcome for ammonia or pyridine. This is due the electrostatic stabilization of the positively charged cation by the negative charge of the zeolite wall.

6.12.2.1.2

The proton bond as a function of zeolite lattice Al/Si concentration ratio

Here we will present the evidence that indicates that there is a wide low Al/Si concentration range where proton strength is independent of zeolite structure. Only when Al/(Al þ Si) ratio becomes of the order of 10% protons become less acidic. This transition regime depends on structure of the zeolite. An important reason why there is this wide Al/Si window for the PA to be invariant, is the Löwenstein rule.34–36 According to this rule no two [AlO2] 1 tetrahedra can share a connecting bridging O atom. At least one [SiO2] tetrahedron has to be located between two [AlO2] 1 tetrahedra. It explains why zeolite structures as zeolite A (LTA structure) or zeolite X (FAU structure) with the maximum zeolite lattice composition of Al/(Al þ Si) ¼ 1/2 can only contain even four or six rings of tetrahedra. Only zeolites with lower Al/(Al þ Si) ratio can also contain uneven rings as the then frequently present five ring of tetrahedra. Zeolite structures with smaller three tetrahedra containing rings are unstable since the low connecting T-O-T angle that is the geometric consequence gives an energy penalty. Zeolites with three ring structures have been synthesized by replacement of [SiO2] tetrahedra by Zn or Ge containing tetrahedra, that can be more readily strained.37–39 An early paper by Kubelkova et al.39 analyzed the shifts in vibrational frequencies of CO adsorbed to zeolite protons over a considerable Al/(Al þ Si) ratio regime with variable alkali cation content for a variety of zeolites. They used Eq. (2) to deduce PA values of the respective protons. These data are summarized in Fig. 4. In the figure also the small changes are indicated when part of the protons become exchanged with Naþ cations. This early result illustrates the sharp increase in Proton Affinity when Al/Si ratio increases. For the Faujasite structure this happens when Al/(Al þ Si) increases beyond a value of 1/7. Below this value Proton Affinity is less and independent of Al/(Al þ Si). This increase in PA and hence decrease of proton reactivity occurs when most of Al containing tetrahedra become their next nearest neighbors.

Fig. 4 Proton affinities of the bridging hydroxyls plotted against the molar ratio Al/(Al þ Si) of the zeolite framework. Arrows designate the increasing concentration of Naþ or non-skeletal Al (:) ZSM-5; (n, C) Y zeolites. Reproduced with permission Kubelková, L.; Beran, S.; Lercher, J. A. Determination of Proton Affinity of Zeolites and Zeolite-Like Solids by Low-Temperature Adsorption of Carbon Monoxide. Zeolites, 1989, 9(6), 539–543, doi: 10.1016/0144-2449(89)90052-3.

278

Solid acid catalysis; Part I, the zeolite protonic site

The lower DPE value deduced for MFI with low Al concentration is 1148 kJ/mol and the highest value for high Al concentration FAU is 1173 kJ/mol. The latter value increases to 1204 kJ/mol when most of the acidic protons are substituted by Na þ cations. In the low Al systems proton substitution by Na ions decreases DPE. Then a large cation or cation of high positive charge located near the protonic site can provide additional stabilization to the negative lattice charge generated when the OH bond polarizes.40,41 The PA increase by Naþ exchange for the high Al/Si systems possibly relates to the inhomogeneous distribution of proton sites with Al in next nearest lattice tetrahedral positions. The larger this local concentration the more the PA increases. Naþ ions will initially replace the protons of the higher reactivity. As a consequence the average PA of the proton concentration increases. In the Si-AlPO4 zeolite materials protons bind to oxygen atoms that connect the Si and P containing tetrahedra, [SiOHP]. Their DPE becomes 1172 kJ/mol,42 a value that is close to the average value of the PA of the high Al concentration alumino-silicate FAU structure. The high PA in the Si -AlPO4 material is because the Si cation next to the protonic site is surrounded by a high concentration of [AlO2] 1 tetrahedra. Consistent with the Kubelkova et al. experiments of Fig. 4 early catalytic experiments with Alumino-silicate zeolites of FAU structure, where the Al/Si ratio of the zeolite framework had been altered by Al leaching treatments show a catalytic activity proportional to Al concentration only as long as Al/(Al þ Si) ratio is less than 20%.43 This is illustrated in Fig. 5A for the proton catalyzed cracking reaction of isooctane that gives isobutane and isobutene as primary products. As we will discuss in part II this reaction demands protons of high acid strength. Whereas in the low Al/(Al þ Si) concentration regime the normalized rate of reaction increases linearly with Al/(Al þ Si) as long as this is less than 20%, beyond this value further increase of Al concentration does not change activity. It indicates an abrupt decrease in proton activity, from its high and initial independence of Al concentration to a linearly decreasing proton activity with increase of Al/(Al þ Si) ratio beyond a critical value. Barthomeuf expressed proton reactivity as a proton activity coefficient a0, that is plotted Fig. 5B as a function of Al/(Al þ Si) ratio. It is equal to 1 as long as Al/(Al þ Si) is less than its critical value and decreases linearly to zero at the maximum value of Al/(Al þ Si) ¼ 1/2. a0 is proposed to be a local property that depends on the number of Al containing tetrahedra at next nearest neighbor sites with respect to another Al containing zeolite framework tetrahedron. This probability depends on the structure of the zeolite. Beyond the critical value of Al/(Al þ Si) a0 decreases in proportion to the Al/(Al þ Si) ratio. This linear decrease in activity implies that the activation energy of reaction increases logarithmically with Al tetrahedral next nearest neighbors. The critical Al/(Al þ Si) ratio where a0 decreases is lower for structures that are topological more dense. When one denotes within brackets the respective critical Si/Al ratio’s one finds the order: MFI (9.5) < MOR(9.4) < OFF(8.3) < FAU (6.8). The dependence on structure is readily understood. For instance a high concentration of four rings instead of six rings around the protonic site will lead to more sharing of next nearest neighbor Al sites. The short range dependence of proton reactivity on the local nearest neighbor number lattice Al concentration can be deduced from quantum-chemical model calculations. For Ferrierite zeolite with different global Al/Si composition ratios this domination of DPE change on Al concentration in next nearest neighbor tetrahedral positions is schematically illustrated by Fig. 6, that is based on DFT calculated results of.44 Differences mainly arise when Al substitutes in next nearest neighbor positions with respect to proton occur. From Fig. 6 illustrates the increase in DPE when the next nearest neighbor tetrahedral concentration of Al3þ increases. It implies that a0 of Fig. 5 decreases non linearly and more steeply with Al/(Al þ Si) ratio than postulated by D. Barthomeuf. The increase in DPE can become close to 10% of the lower DPE value of a low Al/(Al þ Si) siliceous zeolite. An important lemma has been validated: The Al concentration dependence of proton affinity is independent of structure as long as the activity coefficient a0 remains 1, or in other words when next nearest neighbor tetrahedra with respect to the zeolite proton are only occupied by Si4þ.

Fig. 5 Proton activity as a function of Al/(Al þ Si)framework ratio. (A) Rate of isobutene formation in isooctane cracking at 573 K as a function of zeolite framework Al molar fraction m ¼ Al/(Al þ Si), in faujasite structure zeolites. Extracting agent: ,, B, EDTA; n, C. (B) Proton activity coefficient dependence on Al molar fraction m ¼ Al/(Al þ Si) or 2-m value for: 1. FER; 2. MOR; 3. MAZ; 4. LTL, OFF ERI; 5. RHO, LTA; 6. FAU. Reproduced with permission Barthomeuf, D. Zeolite Acidity Dependence on Structure and Chemical Environment. Correlations with Catalysis. Mater. Chem. Phys., 1987, 17(1–2), 49–71, doi: 10.1016/0254-0584(87)90048-4.

Solid acid catalysis; Part I, the zeolite protonic site

279

80

70

Si(nA1) = 3

Edeprot (kJ/mol)

60

50

40

30

Si(nA1) = 2

20

10

Si(nA1) = 1

0 100

Global Si/Al ratio Fig. 6 Schematic illustration of DPE dependence on global Si/Al ratio and on local Si(nAl) descriptor of the number of Al tetrahedra in next nearest neighbor sites. Results are from H-FER(8), H-FER(35) and H-FER(71) model DFT calculations, within brackets Si/Al ratios are denoted. After Grajciar, L.; Areán, C. O.; Pulido, A.; Nachtigall, P. Periodic DFT Investigation of the Effect of Aluminium Content on the Properties of the Acid Zeolite H-FER. Phys. Chem. Chem. Phys., 2010, 12(7), 1497–1506, doi: 10.1039/b917969k.

The local dependence of DPE on Al occupation of next nearest neighbor tetrahedral site positions is a consequence of dominantly covalent chemical bonding in the zeolite network of tetrahedra as we will discuss in the next subsection.

6.12.2.2

The OH chemical bond

Quantum-chemical calculations of the DPE and chemical bonding properties45–48 in combination with data from vibrational spectroscopy32,49 and solid state NMR50–53 have deepened our understanding of the proton chemical bond considerably. In the next three subsections for unpromoted alumino-silicate systems the nature of the proton chemical bond will be discussed and its relation to zeolite structure and site location. Promoted systems will be discussed in the following Section 6.12.2.3. For the highly siliceous zeolites we will see that OH frequency depends on structure and lattice position but not DPE. This independence of zeolite structure will be seen to be due to the flexibility of the zeolite lattice.

6.12.2.2.1

The deprotonation energy

The two types of chemical bonding one distinguishes are covalent versus ionic bonding interactions. This is a relevant distinction since the deprotonation energy, DPE, is the difference between the energy of the OH bond in its equilibrium state, where it is mainly covalently bonded to the zeolite lattice and its separated ionic state of Hþ and [SiOAl] 1. The covalent energy contribution to the chemical bond of Z-0H is approximately 600 kJ/mol. This is the same order of magnitude as that of homolytic OH bond dissociation of an acidic molecule in the gas phase into two neutral radicals.54 Homolytically the [SiO-Hal] bond is split into the [SiOAl] radical that is part of the zeolite framework and the H atom. Deprotonation implies additional electron transfer from the H atom to the zeolite lattice. When the H atom is separated to infinite distance, the electron transfer cost is equal to the energy difference of ionization potential of the H atom and the electron affinity of the zeolite radical site. Similar to the OH covalent bond energy this is also approximately 600 kJ/mol. The sum of the two agrees with the experimentally as well as computationally deduced values of DPE around 1200 kJ/mol.54 Differences in DPE as function of proton siting will relate to differences in the covalent OH bond energy, defined by the chemical bond in its equilibrium state and the unrelated stability of the negative charge left on the zeolite lattice upon deprotonation. We will see that the differences in the latter tend to dominate changes in DPE.

280

Solid acid catalysis; Part I, the zeolite protonic site

The charge of the proton when attached to the zeolite lattice is close to neutral.55 Therefore the electrostatic field near the proton site will be very low. The difference in the interaction energy between a hydrocarbon adsorbed in the siliceous part of the zeolite channel and partially hydrogen bonded with the zeolite proton is therefore small. In the siliceous part of the zeolite channel the van der Waals dispersive interaction between hydrocarbon atoms and polarizable zeolite wall oxygen atoms dominates their adsorption energy. Whereas the adsorption energy of hexane in a zeolite micropore is of the order of 100 kJ/mol its additional interaction with a proton is only of the order of 10 kJ/mol.56 Compared to the hydrogen bonding interaction of the zeolite proton the interaction energy is larger when an alkali cation substitutes for a proton. Near the inorganic cation there will be a strong electrostatic field. The presence of such strong electrostatic interaction leads to strong polarization of hydrocarbon bonds as measured by NMR spectroscopy57 or Infrared Spectroscopy.58,59 From the upwards shift of vibrational frequency of CO or N2 adsorbed to alkali cations Zecchina et al.60 have estimated the electrostatic field of the order of 6.3 V/nm-1 to Na-ZSM5 to decrease of 2.4 V/nm for Cs-ZSM-5. The strong electrostatic field near the inorganic cations gives unique solvation chemistry, that relates to solid electrolytic chemistry.61,62 The covalent character of the zeolite OH bond can be quantified using electronic structure calculations. DFT computed electron partial density of states (PDOS) and Crystal Orbital Hamiltonian Population Densities (COHP) are shown in Figs. 7A and B respectively for an OH site in FAU.64 The COHP is related to the overlap population density between atomic orbitals at a particular electron energy; a positive sign indicates a bonding bond contribution, a negative sign indicates antibonding character (Ref. 65, p. 174 a.f.). Bonding interactions of electron occupied molecular orbitals are observed between O 2s and O 2p and H 1s atomic orbitals respectively around  20 eV and -10 eV. Antibonding interactions of unoccupied molecular orbitals are seen between 5 and 10 eV. The strong mixing O 2s and O 2p atomic orbitals is a signature of hybridization of the OH chemical bond. The large amplitudes in the COHP spectrum indicate strong covalent interaction between O atomic orbitals and H atoms. In the antibonding region the dominating contribution of H 1s atomic orbital to electron density reflects the presence of a small positive charge on the H atom.

Fig. 7 (A) PDOS for H and O atoms of protonic site in FAU zeolite (Si/Al ¼47) zeolite. (B) Corresponding COHP ‘s The z-axis is directed along the OH bond. (C) shows bond orders and atom charges (bond orders and charges have been calculated using the DDEC method see,63). (D) the respective bond lengths (distances in Angstrom), red gives atom number in the solid. (A, B) Reproduced with permission Liu, C.; Tranca, I. C.; Santen, V.; Hensen, R. A.; Liu, Pidko, C.; Tranca, I.; Van Santen, R. A.; Hensen, E. J. M.; Pidko, E. A. Scaling Relations for Acidity and Reactivity of Zeolites. J. Phys. Chem. C 2017, 121(42), 23520–23530, doi: 10.1021/acs.jpcc.7b08176. (C) Courtesy I. Tranca.

Solid acid catalysis; Part I, the zeolite protonic site

281

Figs. 7C and 7D show atom charges, bond orders and calculated bond distances respectively. The atom charges of Fig. 7C show charges on the oxygen atoms that are substantially lower than the formal ionic charges. The AleO bond orders are lower than the SieO bond orders. This is in line with the larger AleO distances compared to that of SieO (Fig. 7D). Attachment of the H atom to the bridging O atom weakens the respective neighbor AleO and SieO bonds as expected for strong covalent bonding.

6.12.2.2.2

Flexibility of zeolite lattice

The dominantly covalent nature of the chemical bonds of the zeolite lattice cause flexibility of the system. Whereas distortion of the [TO4] tetrahedra has a large energy penalty, the bond bending energy of the [SiOSi] angle is less than 10 kJ/mol. This is the basis to zeolite lattice flexibility. It makes adaptation to slight local deformations possible with little energy cost. This relaxation effect will affect the reactivity of the proton at different zeolite lattice positions, since it will influence accommodation of the negative charge generated on the zeolite lattice upon deprotonation.66,67 The simulated structure of the protonic site in Fig. 8A embedded in a siliceous FAU zeolite lattices (67) illustrate the large distortion the voluminous [SiOHAl] proton site causes to the siliceous [SiOSi] lattice site. As Table 1 shows compared to the protonated [SiOHAl] site the non-protonated [SiOAl] 1 site contracts with nearly 15%. There is also a significant change in the T’OT angle. When protonated this angle can be as low as 130 , but when non-protonated it may become similar to the average SiOSi angle of 160 . The smaller [AleOHeSi] angle compared to the [AleOeSi] 1 angle implies a change in hybridization of the respective AlO and SiO bonds. When a site with a wide bond angle changes to one with smaller bond angle hybridization changes from approximately sp. to sp2. A wider bond angle decreases contributions of the O 2 s atomic orbitals to the OH chemical bond, that gets a weaker bond energy. This lowers also the vibrational OH frequency. Ot is the reason why the vibrational frequencies of zeolite protons relate to the [SieOHeAl] bond angle of the proton site.46,68 Lattice infrared vibrational spectra differences between the OH containing zeolite and the system titrated with a basic molecule that deprotonates the lattice,69,70 are the consequences of relaxation of bond distances and angels. This is illustrated by Fig. 8B. The

(A)

Al

OH

a b

absorbance

absorbance

(B)

a

b c

c d 300

500

700

900

wavenumber (cm-1)

1100

1300

d 300

500

700

900

1100

1300

wavenumber (cm-1)

Fig. 8 Local relaxation of zeolite lattice. (A) Local geometry change of AlOH substitution for SiO in FAU from embedded cluster calculations. The figure extends up to the third Si-coordination shell around the Al substitution. (B)(b1) Infrared spectra of zeolite Y, (a) the ammonium form; (b) the acidic proton form; (c) the deuterium exchanged acidic form; (d) the difference spectrum (c) - (b). (b2) Infrared spectra of zeolite ZK-5 (a) the ammonium form, (b) the acidic proton form, (c) the deuterium exchanged acidic form; (d) the difference spectrum (c) - (b). (A) Reproduced with permission Kramer, G. J.; De Man, A. J. M.; Van Santen, R. A. Zeolites versus Aluminosilicate Clusters: The Validity of a Local Description. J. Am. Chem. Soc., 1991, 113(17), 6435–6441, doi: 10.1021/ja00017a012. (B) Reproduced with permission Jacobs, W. P. J. H.; van Wolput, J. H. M. C.; van Santen, R. A.; Jobic, H. A Vibrational Study of the OH and OD Bending Modes of the Brönsted Acid Sites in Zeolites. Zeolites, 1994, 14(2), 117– 125, doi: 10.1016/0144-2449(94)90005-1.

282

Solid acid catalysis; Part I, the zeolite protonic site Comparison bond lengths and angles Naþ versus Hþ site in low AlFAU.

Table 1 Bond

Naþ

Hþ

Al-O (Å) Si-O-Si (o) Si-O-Al (o)

1.7 163 162

2.10 163 140

After Kramer, G. J.; De Man, A. J. M.; Van Santen, R. A. Zeolites versus Aluminosilicate Clusters: The Validity of a Local Description. J. Am. Chem. Soc., 1991, 113(17), 6435–6441, doi: 10.1021/ja00017a012.

FAU vibrational mode of 580 cm 1 weakens upon protonation of the lattice. This vibrational mode is due to a four ring lattice vibration that is an admixture of tetrahedral bending modes and T-O stretch modes.71 The [TO4] tetrahedron around the Al cation is symmetric when Naþ cations compensate for the negative charge. However when an O atom becomes protonated, tetrahedral symmetry around Al is broken. Experimental validation of this symmetry breaking by protonation of the zeolite lattice is deduced from 27Al NMR spectroscopy that provides information on the increase in electric field gradient through nuclear quadrupole coupling. The Al signal in solid state NMR is hardly measurable when protonated, but its chemical shift can be well measured for the hydrated or ammonium exchanged zeolite.72–75 Another experimental observation that supports lattice framework flexibility is the structural transition of the silicalite MFI framework from the ortho to the para structure that is observed when benzene or xylene isomers are adsorbed with high micro pore filling.76,77 Simulations on siliceous model zeolites have identified a zeolite lattice density relaxation window.78 The low density zeolites, with the larger wide [SiOSi] bond angle are least flexible. Flexibility increases when mass density increases.78 For instance the MFI lattice will show more flexibility than the less dense FAU lattice. It implies that in the FAU lattice the [SiOHAl] unit is more constrained than in the MFI lattice. Calculations on sites with comparable topology show that the OH bond frequency in FAU is higher than in MFI. This implies that the covalent bond strength of the OH bond is the stronger. However the DPE of FAU is found to be lower than of MFI.66 The difference in strain of the protonated and deprotonated site dominates the PDE difference.

6.12.2.2.3

Site dependence

The possibility that a zeolite OH group positions at a specific zeolite site, requires an Al cation to be present in a [TiO2] tetrahedron of the zeolite framework. Whereas differences in Al siting energy exist, the experimental distribution of Al over the tetrahedral sites appears highly zeolite synthesis condition dependent.79 Al positioning can be related to the preferred adsorption positions of base templates during synthesis in the micro pores of the zeolite.80 An experimental study of Wang et al. compared Al distribution in ZSM-5 and ZSM-1181 that finds preferential siting of Al in the zeolite channel walls instead of channel intersections that relates to template siting. Because of this synthesis dependence and also the difficulty to determine the Al siting this information is usually absent. Therefore modeling studies usually assume a statistical distribution of Al over different sites of the zeolite. In agreement with this

100

10 R 1.0

0.1

10 100,000

100 10,000

1000 1,000

p.p.m.Al 100

SiO2 10 Al2O3

Fig. 9 The hexane cracking activity denoted as R plotted against the aluminum framework content in H-ZSM-5. Reproduced with permission Haag, W. O.; Lago, R. M.; Weisz, P. B. The Active Site of Acidic Aluminosilicate Catalysts. Nature, 1984, 309, 589–591, doi: 10.1038/309589a0.

Solid acid catalysis; Part I, the zeolite protonic site

283

assumption an experimental study of Haag et al.82 finds that the proton normalized rate of hexane cracking is independent of Al Al < 10%). The ZSM-5 zeolite has 12 distinguishable tetrahedral crysconcentration in H-ZSM-5 of a 103 concentration interval, (AlþSi tallographic positions. It not only indicates that in this case Al siting indeed is statistical, but also that proton strength is independent of Al siting (Fig. 9). Computations on low Al/Si concentration zeolites that average DPE over a statistical distribution of Al located at crystal graphically different positions indeed find such an independence and confirm the independence of structure of proton reactivity. In practice the differences become especially insignificant since proton mobility is high at the temperatures of catalysis. Then their position can be also averaged over different bridging O atom positions around the Al center.48 This is illustrated in Fig. 10 that gives such averaged calculated DPE’s for different zeolite structures at different Al site locations. Note the systematic difference of the order of 50 kJ/mol between empirically determined DPE values of Fig. 4 and computed value of Fig. 10 25 years later. Experimentally for zeolites with a framework composition of high Al/Si ratio the OH vibrational frequency, DPE and Al/Si ratio correlate. This is illustrated in Fig. 11A for the OH frequencies.80 As we discussed in Section 6.12.2.1.2. in this Al/Si concentration regime the DPE is a strong function of Al/Si framework ratio. This is different In the low Al/(Al þ Si) systems. This is illustrated by Fig. 11b1 and b2 are based on the DFT quantum-chemical studies of low Al/(Al þ Si) systems.48 There is a correlation between OH vibrational frequency and T0 eOHeT bond angle in low Al/(Al þ Si) framework composition systems (Fig. 11b1). A different OH vibrational frequency denotes different proton siting. However different from the high Al/ (Al þ Si) ratio systems in this case there is no correlation between DPE and proton siting (see Fig. 11b2). In the low Al/(Al þ Si) systems stabilization of the negative charge left on the reactive site upon deprotonation counteracts the changes in covalent bond strength of the protonated site. The conclusion that the reactivity of the proton in zeolites with Al/(Al þ Si) framework composition ratio less than approximately 10% is independent of structure, implies that experimentally observed structure dependence of solid acid catalysis by siliceous zeolites has to have a different cause. We will see in part II that this is dominated by the structure sensitivity of the adsorption free energies of reactants or reaction intermediate. This is called the confinement effect of the zeolite micro pore.25,83 In Fig. 7 we illustrated that the nature of the chemical bond of SieO and AleO in the zeolite framework as well as that of the zeolite OH bond is dominantly covalent. Attachment of the H atom to the bridging O atom weakens the respective neighbor AleO and SieO bonds. This is an illustration of the Bond Order Conservation law (BOC,84).According to BOC the sum of the bond orders toward an atom is nearly constant and is independent of the atom coordination number. For the oxygen atom this value is close to

Fig. 10 Comparison of DFT calculated deprotonation energies for different zeolite structures. DPE’s are averaged at the different site locations of the low Al/(Al þ Si) concentration zeolite structures. They are seen to be independent of framework structure or proton location. Reproduced with permission Jones, A. J.; Iglesia, A. J. The Strength of Brønsted Acid Sites in Microporous Aluminosilicates. ACS Catal., 2015, 5(10), 5741–5755, doi: 10.1021/acscatal.5b01133.

284

Solid acid catalysis; Part I, the zeolite protonic site

3640 3600

ν/cm-1

(A)

0

2:1

4:1 Si:Al ratio

(B)

(D)

Fig. 11 Correlation of vibrational frequencies and DPE with proton site [SiOAl] bond angle. (A) Frequencies of the acidic hydroxy-groups as a function of the Si:Al ratio for zeolites with different zeolite structures. (B) (b1, b2) Comparison of computed OH frequency (c1) and DPE (c2) dependence on SieO eAl angle for low Al/(Al þ Si) MFI zeolite structure. Whereas DPE varies little, the OH frequency decreases with increasing bond angle. (A) After Barthomeuf, D. Description of a Characteristic Infrared Frequency for Acidic Hydroxy-Groups in Zeolites. J. Chem. Soc. Chem. Commun. 1977, 21, 743–744, doi: 10.1039/C39770000743. (B) After Jones, A. J.; Iglesia, A. J. The Strength of Brønsted Acid Sites in Microporous Aluminosilicates. ACS Catal., 2015, 5(10), 5741–5755, doi: 10.1021/acscatal.5b01133.

H

H

(A)

(B)

O

O

Si O

Si O

O

Al O

s w O s

O

O

Si O

O

Al

O

O

H

(C)

O

O

Al O

O

Si O

O

O

H

Si

Si O

O

Si

w' O w' s' O s' O

w O

O

O Al

O

O

Si O

O

O

Fig. 12 Schematic illustration of changes in OH bond strength as a function of Si by Al substitution in next nearest tetrahedral positions according to Bond Order Conservation. (A) The non-substituted system. (B) Bond strength changes by Al substitution (w weaker, s stronger). (C) Including proton attachment.

Solid acid catalysis; Part I, the zeolite protonic site

285

the classically expected value of 2. The bridging oxygen atom has three neighbor atoms compared to the two neighbors in the nonprotonated site. Since its valency or bonding power now has to be redistributed over three instead of two bonds, the individual bonds will weaken. BOC can also be used to explain the changes in OH bond strength as a function of Al concentration on the zeolite framework. Fig. 12 illustrates schematically the predicted changes in OH bond strength according to Bond Order Conservation (see,85 p. 322,323) as a function of Al substitution in next nearest neighbor tetrahedra. Fig. 12A shows schematically the local network with only Si containing tetrahedra as next nearest neighbors. In Fig. 12B the changes in bond order are indicated when Al substitution occurs. The AleO bond is weaker than the SieO bond (indicated as w). Since according to BOC total atom valency is a constant, the SieO bond with the O atom attached to Al becomes stronger (indicated with s), the next SiO bond then weakens and the OH bond strengthens. Al substitution weakens proton donation strength. Obviously when substitution of Al is in a tetrahedron with Si containing tetrahedra between the protonic site, bond weakening and strengthening effects will distribute over many bonds which dilutes its influence on the OH bond strength. Local disturbances decrease exponentially with distance.85 When a proton is attached to the O atom next to the substituting Al atom (Fig. 12C) Bond Order Conservation indicates an OH bond weakening correction to the OH bond strength in the absence of the additional proton. This is in agreement with a prediction that Naþ cation exchange of the proton should decrease proton donation power of the protonic site.

6.12.2.3 6.12.2.3.1

Proton activity of other than Al/Si framework materials Al substitution by Fe3þ and Ga3þ

Bond order conservation predicts that the strength of the zeolite OH bond will increase when Al3þ is replaced by a three valent cation with a weaker bond to oxygen atom. Experiments that compare proton strength of Al3þ substituted by Fe3þ and Ga3þ show that this is indeed the case. A convenient probe to proton strength is measurement of the heat of adsorption of ammonia or the rate of decompositions of ammonia. For zeolites with MFI structure Fig. 13A compares OH vibrational frequencies of zeolites with different cations in the tetrahedral site with the temperature of decomposition of ammonium to ammonia and the zeolite proton. This temperature is the temperatures where the decomposition rate is maximum. The higher this temperature the smaller proton DPE. Tmax correlates with measured vibrational frequencies, that increase in the order Al3þ, Fe3þ and Ga3þ and B3þ. This is the sequence expected from the corresponding increase in TeO differences, shown in Table 2, that indicate decreasing TeO bond strength and hence an increase in OH bond strength. The frequency of OH of the Boron cation similar to that of the surface silanol. Its inorganic chemistry is different from that of the zeolite proton.

Fig. 13 The effect of Al substation by different three valent cations. (A) Tmax of NH4þ decomposition rate versus OH stretching frequency from FTIR spectra of isomorphous substituted ZSM-5 zeolites. (B) Conversion of n-hexane over crystalline metal silicates (723 K, total pressure of 0.5 MPa (helium/n-hexane ¼ 4/1 mol/mol). Effect of trivalent metal content of MFI silicate on reaction rate constant. Comparison of Al3þ, Ga3þ and Fe3þ substituted systems. (A) With permission Chu, C. T. W.; Chang, C. D. Isomorphous Substitution in Zeolite Frameworks. 1. Acidity of Surface Hydroxyls in [B]-, [Fe]-, [Ga]-, and [Al]-ZSM-5. J. Phys. Chem., 1985, 89(9), 1569–1571, doi: 10.1021/j100255a005. (B) With permission Post, M. F. M.; Huizinga, T.; Emeis, C. A.; Nanne, J. M.; Stork, W. H. J. An Infrared and Catalytic Study of Isomorphous Substitution in Pentasil Zeolites. Stud. Surf. Sci. Catal., 1989, 46(C), 365–375, doi: 10.1016/S0167-2991(08)60993-3.

286

Solid acid catalysis; Part I, the zeolite protonic site Table 2 M M-2TH dM-O(av) M-8TH dM-O(av)

Calculated MO bond lengths of three valent cations substituted in two MFI zeolite models. Al

Ga

Fe

B

1.711

1.738

1.872

1.355

1.717

1.754

1.841

1.382

Reproduced with permission Yuan, S. P.; Wang, J. G.; Li, Y. W.; Jiao, H. Brønsted Acidity of Isomorphously Substituted ZSM-5 by B, Al, Ga, and Fe. Density Functional Investigations. J. Phys. Chem. A, 2002, 106(35), 8167–8172, doi: 10.1021/jp025792t.

The low conversion catalytic hexane cracking experiment of Fig. 13B illustrates that catalytic reaction rate is strongly affected by the differences in DPE of the respective cationic framework substitutions.86 This is because differences in DPE relate to differences in activation energies of proton activated elementary reactions. Related results have been published for catalytic methanol dehydration (with permission87).

6.12.2.3.2

Non framework substituted systems

Some of the wide pore zeolites, as zeolite X or Y, with the FAU structure of Fig. 1A, that are important to several catalytic processes are only commercially available with high Al content. Therefore their PA is high. Proton affinity can be improved by post synthesis methods that leach Al from the zeolite framework88 so that framework Al/(Al þ Si) ratio is decreased. As we will discuss shortly residual Aluminum cations located in site pockets of the micro channels will decrease PA and hence promote proton reactivity. Lattice imperfections will give internal silanol groups that we will discuss at the end of this section. In the hydrothermal steam treatment of high Al/(Al þ Si) concentration ratio zeolites initially hydroxylated monomeric Al cations are generated. In subsequent reactions inorganic self organization processes89 convert the monomeric Al cations to residual oligomeric AlxOy (OH)n þ clusters (see Fig. 14). These clusters become located in zeolite micro cavities, as the double six rings or the sodalite cages (see Fig. 1), that are not accessible to reactants. Promotion of dealuminated zeolite Y by these oligomeric Al cationic clusters is evidenced by model studies of the catalytic cracking reaction of propane and isobutane.89,90 The AlxOyHn þ clusters that are usually referred to as extra framework Alumina sites (EfAL) reduce proton content of the zeolite but decrease PA of the protons attached to the zeolite frame work.91 Severe degradation occurs in the fluid cracking process where the zeolite material is exposed to severe oxidative hydrothermal regeneration conditions. In order to stabilize zeolite Y against degradation zeolites partially exchanged with earth or rare earth alkali cations are often used.92 The earth alkali cations as well as the rare earth alkali cations become partially hydrolyzed.92 Similarly as for the AlxOyHn þ clusters the rare earth alkali cations become part of oxy-hydroxy-cationic clusters through related self assembly reactions.93 The

Fig. 14 Schematic representation of the selforganization reactions of hydroxylated Al3þ cations in the Faujasite structure. Al3þ oxy-hydroxy clusters locate in sodalite cages. Reproduced with permission Liu, C.; Li, G.; Hensen, E.J.M.; Pidko, E.A. Nature and Catalytic Role of Extraframework Aluminum in Faujasite Zeolite: A Theoretical Perspective. ACS Catal. 2015, 5(11), 7024–7033, doi: 10.1021/acscatal.5b02268.

Solid acid catalysis; Part I, the zeolite protonic site

287

stabilizing effect of La3þ containing cationic clusters is larger than that of the occluded Ca(OH)þ or Mg(OH)þ cations that are formed when earth alkali cations are used. Analogous to the residual Aluminum cations generated by steam treating Zeolite X and Y, the earth and rare earth alkali cations also promote proton reactivity.94,95 Promotion by La3þ ion exchange of zeolite Y and X has been extensively studied for the alkylation reaction of isobutane and propylene.96 The location and structure of the La3 oxy-hydroxy-cationic cluster91 is similar to that of the Al3-oxy-hydroxy cluster generated by steam treating high Al/(Al þ Si) zeolite.93 As Table 3 mentions reduction of the proton affinity by the Al clusters is the larger. Ca promoted Zeolite Y catalysts have also been extensively used as hydrocracking or hydroisomerization catalysts.97 Whereas the Al/Si ratio ¼ 1/2.5 is relatively high, these applications illustrate that proton reactivity of these Ca2þ promoted systems competes with that of low Al/Si ZSM-5 zeolites.98 The decreased proton affinity is due to the synergetic action of Lewis acidic cationic complexes and neighbor Brønsted acid protons. It can be illustrated by DFT calculations of the energies of adsorption of acetonitrile and ammonia of the FAU structure of different compositions, that are shown in Table 3 and Fig. 15. When it interacts with the zeolite proton acetonitrile adsorbs through hydrogen bonding60,83–85 without transfer of the proton, whereas ammonia becomes protonated.60,99 The acetonitrile molecule only interacts with a single proton, but the protonated ammonia molecule interacts with at least two of its hydrogen atoms to the negatively charged Lewis basic zeolite lattice oxygen atoms. Variation in the adsorption energy of acetonitrile is a measure of the covalent OH bond strength and that of the interaction energy of ammonia is a measure of DPE. In the comparison of composition induced changes the two relate. The quantum chemical calculations of Table 3 are the sum of the interaction energy contribution of adsorbate with the proton only and the additional contribution Dads due to the van der Waals dispersion interactions delta with the zeolite wall. The latter corresponds closely to the adsorption energy of the molecule with the siliceous part of the zeolite wall. The van der Waals dispersion interaction of the tetrahedral ammonium cation is substantially larger than that of linear acetonitrile. The dispersion interactions increases with increasing curvature of the zeolite cavity; MFI > CHA > FAU. Whereas the total adsorption energies depend strongly on zeolite cavity structure (also called confinement effect), the interaction with the proton only does not vary and is independent of zeolite cavity structure. It confirms the conclusions of the independence of DPE of zeolite structure mentioned in the previous section. In agreement with previous cluster calculations by Vassylev et al.41 also for the calculations64 with the full periodic structures proton replacement by a Naþ cation in the low Al/(Al þ Si) composition ratio zeolite decreases the DPE of the nearby located zeolite proton. These alkali cations are usually located in micro channels, where they neutralize for the zeolite lattice negative charge. These calculated results agree with the experiments of Fig. 4 on alkali cation exchanged zeolites of with low Al/(Al þ Si)ratio.

Table 3

DFT calculated interaction energies of CH3CN, NH3 with Fau zeolite models of varying composition.

Zeolite model

Fau-47 Fau-7 FAU-2.4 FAU-Fe FAU-Ga FAU-SAPO FAU-Na FAU-EFAl-mono FAU-EFAl-bi FAU-EFAl-tri MFI CHA

CH3CN

NH3

DEads

DEads(D)

DEads

DEads(D)

90 68 60 82 85 83 93 106 119 120 115 108

15 16 15 15 16 15 15 17 17 17 52 35

142 125 104 129 138 120 142 188 189 183 166 159

16 16 16 14 15 15 16 15 17 15 24 26

Total adsorption energies DEads and the respective Edisp dispersive energy contributions are given. In Fig. 15 the respective minimum energy configurations are shown that correspond to the calculated energies of this table. The computed ammonia adsorption energy of NH3 in La promoted FAU-7 is 156 kJ/mol94 This is to be compared with Fau-EF-Al ¼  188 kJ/mol. Reproduced with permission Liu, C.; Tranca, I. C.; Santen, V.; Hensen, R. A.; Liu, Pidko, C.; Tranca, I.; Van Santen, R. A.; Hensen, E. J. M.; Pidko, E. A. Scaling Relations for Acidity and Reactivity of Zeolites. J. Phys. Chem. C 2017, 121(42), 23520–23530, doi: 10.1021/ acs.jpcc.7b08176.

288

Solid acid catalysis; Part I, the zeolite protonic site

Fig. 15 Computational models of zeolites: faujasite (FAU) zeolites with different Si/AlF ratios [FAU-47 (Si/Al ¼ 47), FAU-7 (Si/Al ¼ 7), and FAU-2.4 (Si/Al ¼ 2.4)], heteroatom models [FAU-Fe (Si/Fe ¼ 47) and FAU-Ga (Si/Ga ¼ 47)], silico-alumino phosphate SAPO-37 [FAU-SAPO (Si:P:Al ¼ 1:23:24)], extra framework-cation-containing models [FAU-EFAl-mono (Si/AlF ¼ 7 and Si/Altotal ¼ 6), FAU-EFAl-bi (Si/AlF ¼ 7 and Si/ Altotal ¼ 5.3), FAU-EFAl-tri (Si/AlF ¼ 7 and Si/Altotal ¼ 4.7), and FAU-Na (Si/Al ¼ 7)], and high-silica CHA (Si/Al ¼ 35) and MFI (Si/Al ¼ 95). Reproduced with permission Liu, C.; Tranca, I. C.; Santen, V.; Hensen, R. A.; Liu, Pidko, C.; Tranca, I.; Van Santen, R. A.; Hensen, E. J. M.; Pidko, E. A. Scaling Relations for Acidity and Reactivity of Zeolites. J. Phys. Chem. C 2017, 121(42), 23520–23530, doi: 10.1021/acs.jpcc.7b08176.

Compared to activation by alkali cation located in the zeolite microchannels the decrease in DPE is significantly larger for the system promoted with AlxOHyn þ or LaxOHyn þ clusters.64 In Section 6.12.2.1.1 we mentioned the large difference in electrostatic field around a cation compared to that around a proton. The positively charged alkali cations act as strong Lewis acid sites. The interaction of base molecules with the alkali cations is stronger than with the zeolite protons.100 Calorimetric measurement of ammonia adsorption on zeolites that are partially exchanged by Naþ cations by Auroux101 (Fig. 16) as a function of ammonia coverage show at low coverage a strong heat of adsorption of 120 kJ/mol that decreases to 80 kJ/mol when coverage increases. The latter number compares well the computed value of

Solid acid catalysis; Part I, the zeolite protonic site

289

Fig. 16 Differential heats of adsorption of ammonia on zeolite Y exchanged with alkali cations. After Chester, A. W.; Derouane, E. G. Zeolite Chemistry and Catalysis. Dordrecht: Springer Netherlands, 2009.

90 kJ/mol of adsorption to the proton in Na exchanged Fau.64 The value of 120 kJ/mol of the heat of adsorption of ammonia corresponds to that of ammonia adsorbed to Naþ. The Fig. 16 also illustrates the strong decrease in electrostatic field as deduced from the decrease in heat of adsorption when alkali cations are exchanged with larger radius. Hydrolysis of lattice vacancies generated by Aluminum leaching from the zeolite lattices will generate internal silanol groups. They will usually consist of several silanols that partially interact through hydrogen bonds as sketched in Fig. 17. Three types of OH bonds can be distinguished: isolated, terminal and hydrogen bonded. The respective OH frequencies and related reactivities are well understood.100,102 The hydrogen bonded OH hydroxyls have substantially weakened bonds and shift downwards with a broadened higher intensity adsorption. For defective Silicalite (the Siliceous MFI structure) the terminal(II) OH frequency is lowered by 25 cm 1 compared the isolated OH vibrational frequency at 3725 cm 1. CO adsorption shows that the terminal (II) OH binds stronger than the isolated OH because, since it desorbs at higher temperature from the terminal (II) OH.26 Whereas isolated SieOH is not active in alkene internal double bond shift reactions, terminal OH will catalyzes this reaction.102 According to103 activated silanols as in silanol nests are able to catalyze also the Beckman isomerization reaction of cyclohexanolomine to the commercially important 3-caprolactam.

6.12.2.4

In summary

The surface chemistry of the oxide usually relates to the rupture of metal-oxide bonds due to surface formation. Surface hydroxyls, that can be Brønsted basic or acidic then are formed by dissociative adsorption of water. The external surface of the alumino-silicate zeolite will contain mainly weakly acidic silanol groups.

Si isolated

terminal (I)

terminal (II) O

H O

Si

H

H O

O Si

H O

H O

Si

Si

Si

O

H

H H

H O

O

Si Si

O

vicinal silanol

geminal silanol

silanol nest

Fig. 17 Schematic structures of different silanols. With permission Hattori, H.; Arudra, P.; Abdalla, A.; Aitani, A. M.; Al-Khattaf, S. S. Infrared Study of Silanol Groups on Dealuminated High Silica MFI Zeolite to Correlate Different Types of Silanol Groups with Activity for Conversion of 1-Butene to Propene. Catal. Lett., 2020, 150(3), 771–780, doi: 10.1007/s10562-019-02972-8.

290

Solid acid catalysis; Part I, the zeolite protonic site

Compared to these systems the zeolite proton site is unique, since this hydroxyl is part of the zeolite solid state structure. Its architecture consists of tetrahedral rings of different size from which the micro pores are build, that create an internal surface. The protons are attached to oxygen atoms that are part of the walls that surround the micro cavities. The question that we addressed in this chapter is the relation of zeolite proton reactivity with zeolite micro pore structure and zeolite composition. We will summarize the main conclusions here for the alumino-silicate zeolite. Analogous systems have been discussed where Al is replaced by other three valent cations or the alumino-silicate framework has been substituted for the silico-alumino-phosphate framework. Proton DPE (or PA) depends on framework Al/(Al þ Si) framework composition ratio and varies when Al/(Al þ Si) is smaller or larger than approximately 10%. For protonic alumino-silicate zeolites when Al/(Al þ Si) framework concentration ratio is less than 10% the DPE can be considered composition and site independent. When Al/(Al þ Si) concentration ratio is larger than 10% the DPE increases with increasing Al/(Al þ Si) concentration ratio. The dependence on framework concentration then is zeolite structure dependent. When alkali, earth alkali or rare earth metal cations are partially exchanged for protons DPE is strongly promotor concentration dependent. The promotion effect is structure dependent because site positions of the cations or cationic clusters vary with structure. PA generally decreases and is due to Lewis acid-Brønsted acid synergy. For zeolite protons three classes of proton affinity can be distinguished. The higher proton affinity and lower reactivity have protons of zeolites with an Al/(Al þ Si) framework concentration ratio larger than 10%. The second class of decreased proton affinity and smaller DPE is typical for siliceous zeolites with Al/(Al þ Si) ratio less than 10%. Earth or rare earth exchanged zeolites of high Al/(Al þ Si) framework composition ratio, will not only reduce proton concentration but will also decrease PA comparable to that of the low Al/(Al þ Si) systems. A third class of protons has an even lower proton affinity, as in dealuminated zeolites promoted by EfAl. The differences in DPE between the three classes are of the order of 30 kJ/mol. So far we discussed proton properties of the isolated proton or in contact with probe molecules. In part II we will discuss proton activation of organic reactants in solid acid catalytic reactions.

References 1. Coombs, D. S.; Alberti, A.; Armbruster, T.; Artioli, G.; Grice, J. D.; Liebau, F.; Nickel, E. H.; Peacor, D. R.; Rinaldi, R.; Ross, M.; Sheppard, R. A.; Vezzalini, G. Recommended Nomenclature for Zeolite Minerals: Report of the Subcommittee on Zeolites of the International Mineralogical Association, Commission on New Minerals and Mineral Names. Can. Mineral. 1997, 35, 1571–1606. 2. Liebau, F. Structural Chemistry of Silicates, Springer: Berlin Heidelberg, 1985. 3. Breck, D. W. Zeolite Molecular Sieves: Structure, Chemistry, and Use, Wiley, 1973. 4.  Cejka, J.; Corma, A.; Zones, S. Zeolites and Catalysis: Synthesis, Reactions and Applications, Wiley VCH, 2010. 5. Sherman, J. D. Synthetic Zeolites and Other Microporous Oxide Molecular Sieves. Proc. Natl. Acad. Sci. U. S. A. 1999, 96 (7), 3471–3478. https://doi.org/10.1073/ pnas.96.7.3471. 6. Glaser, R.; Weitkamp, J. The Application of Zeolites in Catalysis, Springer, 2004. 7. Wright, P. A. Microporous Framework Solids, Royal Society of Chemistry, 2008. 8. Baerlocher, C.; McCusker, L. B.; Olson, D. H. Atlas of Zeolite Framework Types, Elsevier, 2007. 9. Treacy, M. M. J.; Rivin, I.; Balkovsky, E.; Randall, K. H.; Foster, M. D. Enumeration of Periodic Tetrahedral Frameworks. II. Polynodal Graphs. Microporous Mesoporous Mater. 2004, 74 (1–3), 121–132. https://doi.org/10.1016/j.micromeso.2004.06.013. 10. Haag, W. O. Catalysis by Zeolites - Science and Technology. In Studies in Surface Science and Catalysis; Hölderich, W., Karge, H. G., Weitkamp, J., Pfeifer, H., Eds.; 1994; pp 1375–1394. 11. Weitkamp, J.; Karge, H. G. Molecular Sieves: Science and Technology, Springer-Verlag, 1998. 12. Plank, C. J. The Invention of Zeolite Cracking Catalysts. ACS Symp. Ser. 1983, 222 (4), 253–271. Available: http://cat.inist.fr/?aModele¼afficheN7B&7Dcpsidt¼9627522. 13. Jacobs, P. A.; et al. Introduction to Zeolite Science and Practice;, 2nd ed; . https://www.elsevier.com/books/introduction-to-zeolite-science-and-practice/jacobs/978-0-44482421-9 accessed 13 December 2020. 14. Martens, J. A.; Tielen, M.; Jacobs, P. A. Relation Between Paraffin Isomerisation Capability and Pore Architecture of Large-Pore Bifunctional Zeolites. Stud. Surf. Sci. Catal. 1989, 49–60. 15. Lok, B. M.; Messina, C. A.; Patton, R. L.; Gajek, R. T.; Cannan, T. R.; Flanigen, E. M. Silicoaluminophosphate Molecular Sieves: Another New Class of Microporous Crystalline Inorganic Solids. J. Am. Chem. Soc. 1984, 106 (20), 6092–6093. https://doi.org/10.1021/ja00332a063. 16. Lourenco, J. P.; Ribeiro, M. F.; Ribeiro, F. R.; Rocha, J.; Gabelica, Z.; Dumont, N.; Derouane, E. G. Study of catalytic properties of SAPO-40. Stud. Surf. Sci. Catal. 1994, 84 (C), 867–874. https://doi.org/10.1016/S0167-2991(08)64092-6. 17. Olsbye, U.; Svelle, S.; Lillerud, K. P.; Wei, Z. H.; Chen, Y. Y.; Li, J. F.; Wang, J. G.; Fan, W. B. The Formation and Degradation of Active Species During Methanol Conversion Over Protonated Zeotype Catalysts. Chem. Soc. Rev. 2015, 44 (20), 7155–7176. https://doi.org/10.1039/c5cs00304k. 18. Tian, P.; Wei, Y.; Ye, M.; Liu, Z. Methanol to Olefins (MTO): From Fundamentals to Commercialization. ACS Catal. 2015, 5 (3), 1922–1938. https://doi.org/10.1021/ acscatal.5b00007. 19. Uytterhoeven, J. B.; Christner, L. G.; Hall, W. K. Studies of the Hydrogen Held by Solids. VIII. The Decationated Zeolites. J. Phys. Chem. 1965, 69 (6), 2117–2126. https:// doi.org/10.1021/j100890a052. 20. Brown, W. G. Physical Organic Chemistry (Hammett, L. P.). J. Chem. Educ. 1940, 17 (11), 551. https://doi.org/10.1021/ed017p551.3. 21. Busca, G. Acid Catalysts in Industrial Hydrocarbon Chemistry. Chem. Rev. 2007, 107 (11), 5366–5410. https://doi.org/10.1021/cr068042e. 22. Li, Z.; Johnson, M. C.; Sun, M.; Ryan, E. T.; Earl, D. J.; Maichen, W.; Martin, J. I.; Li, S.; Lew, C. M.; Wang, J.; Deem, M. W.; Davis, M. E.; Yan, Y. Mechanical and Dielectric Properties of Pure-Silica-Zeolite Low-k Materials. Angew. Chemie 2006, 118 (38), 6477–6480. https://doi.org/10.1002/ange.200602036. 23. Solkan, V. N.; Kuz’min, I. V.; Kazanskii, V. B. Quantum-Chemical Study of Alkyl Carbenium Ions in 100% Sulfuric Acid. Kinet. Catal. 2001, 42 (3), 411–417. https://doi.org/ 10.1023/A:1010477718602.

Solid acid catalysis; Part I, the zeolite protonic site

291

24. Bordiga, S.; Lamberti, C.; Bonino, F.; Travert, A.; Thibault-Starzyk, F. Probing Zeolites by Vibrational Spectroscopies. Chem. Soc. Rev. 2015, 44 (20), 7262–7341. https:// doi.org/10.1039/c5cs00396b. 25. Bhan, A.; Gounder, R.; Macht, J.; Iglesia, E. Entropy Considerations in Monomolecular Cracking of Alkanes on Acidic Zeolites. J. Catal. 2008, 253 (1), 221–224. https:// doi.org/10.1016/j.jcat.2007.11.003. 26. Dijkstra, T. W.; Duchateau, R.; van Santen, R. A.; Meetsma, A.; Yap, G. P. A. Silsesquioxane Models for Geminal Silica Surface Silanol Sites. A Spectroscopic Investigation of Different Types of Silanols. J. Am. Chem. Soc. 2002, 124 (33), 9856–9864. https://doi.org/10.1021/ja0122243. 27. Paukshtis, E. A.; Yurchenko, E. N. Study of the Acid–Base Properties of Heterogeneous Catalysts by Infrared Spectroscopy. Russ. Chem. Rev. 1983, 52 (3), 242–258. https:// doi.org/10.1070/rc1983v052n03abeh002812. 28. Sauer, J.; Hill, J. R. The Acidity of Surface Silanol Groups. A Theoretical Estimate Based on Ab Initio Calculations on a Model Surface. Chem. Phys. Lett. 1994, 218 (4), 333– 337. https://doi.org/10.1016/0009-2614(94)00002-6. 29. Zecchina, A.; Buzzoni, R.; Bordiga, S.; Geobaldo, F.; Scarano, D.; Ricchiardi, G.; Spoto, G. Host-Guest Interactions in Zeolite Cavities. Stud. Surf. Sci. Catal. 1995, 97, 213– 222. https://doi.org/10.1016/S0167-2991(06)81892-6. 30. Lias, S. G.; Liebman, J. F.; Levin, R. D. Evaluated Gas Phase Basicities and Proton Affinities of Molecules; Heats of Formation of Protonated Molecules. J. Phys. Chem. Ref. Data 1984, 13 (3), 695–808. https://doi.org/10.1063/1.555719. 31. East, A. L. L.; Smith, B. J.; Radom, L. Entropies and Free Energies of Protonation and Proton-Transfer Reaction. J. Am. Chem. Soc. 1997, 119 (38), 9014–9020. https:// doi.org/10.1021/ja970891j. 32. Waller, S. E.; Belousov, A.; Kidd, R. D.; Nikolic, D.; Madzunkov, S. M.; Wiley, J. S.; Darrach, M. R. Chemical Ionization Mass Spectrometry: Applications for the In Situ Measurement of Nonvolatile Organics at Ocean Worlds. Astrobiology 2019, 19 (10), 1196–1210. https://doi.org/10.1089/ast.2018.1961. 33. Vener, M. V.; Rozanskay, X.; Sauer, J. Protonation of Water Clusters in the Cavities of Acidic Zeolites: (H 2 O) n ÁH-Chabazite, n ¼ 1-4w. Phys. Chem. Chem. Phys. 2009. https://doi.org/10.1039/b817905k. 34. Lowenstein, W. The Distribution of Aluminum in the Tetra-Hedra of Silicates and Aluminates. Am. Mineral. 1954, 39 (1-2), 92–96. 35. Engelhardt, G.; Zeigan, D.; Lippmaa, E.; Mogi, M. Zur Frage der Verteilung der Silicium- und Aluminiumatome im Kristallgitter von Zeolithen. Ein alternativer Strukturvorschlag fur Zeolithe vom Typ NaA. Zeitschrift fur Anorg. und Allg. Chemie 1980, 468 (1), 35–38. https://doi.org/10.1002/zaac.19804680105. 36. Bell, R. G.; Jackson, R. A.; Catlow, C. R. A. Löwenstein’s Rule in Zeolite A: A Computational Study. Zeolites 1992, 12 (7), 870–871. https://doi.org/10.1016/0144-2449(92) 90065-W. 37. Röhrig, C.; Gies, H. A New Zincosilicate Zeolite with Nine-Ring Channels. Angew. Chemie Int. Ed. English 1995, 34 (1), 63–65. https://doi.org/10.1002/anie.199500631. 38. Bu, X.; Feng, P.; Stucky, G. D. Novel Germanate Zeolite Structures with 3-Rings. Angew. Chem., Int. Ed. Engl. 1998, 120, 11204–11205. https://doi.org/10.1021/ ja982591n. 39. Kubelková, L.; Beran, S.; Lercher, J. A. Determination of Proton Affinity of Zeolites and Zeolite-Like Solids by Low-Temperature Adsorption of Carbon Monoxide. Zeolites 1989, 9 (6), 539–543. https://doi.org/10.1016/0144-2449(89)90052-3. 40. Gonzales, N. O.; Chakraborty, A. K.; Bell, A. T. A Density Functional Theory Study of the Effects of Metal Cations on the Brønsted Acidity of H-ZSM-5. Catal. Letters 1998, 50 (3–4), 135–139. https://doi.org/10.1023/A:1019095808885. 41. Vayssilov, G. N.; Rösch, N. Influence of Alkali and Alkaline Earth Cations on the Brønsted Acidity of Zeolites. J. Phys. Chem. B 2001, 105 (19), 4277–4284. https://doi.org/ 10.1021/jp0041048. 42. Weyda, H.; Lechert, H. The Crystallization of Silicoaluminophosphates with the Structure-type SAPO-5. Zeolites 1990, 10, 251–258. 43. Barthomeuf, D. Zeolite Acidity Dependence on Structure and Chemical Environment. Correlations with Catalysis. Mater. Chem. Phys. 1987, 17 (1–2), 49–71. https://doi.org/ 10.1016/0254-0584(87)90048-4. 44. Grajciar, L.; Areán, C. O.; Pulido, A.; Nachtigall, P. Periodic DFT Investigation of the Effect of Aluminium Content on the Properties of the Acid Zeolite H-FER. Phys. Chem. Chem. Phys. 2010, 12 (7), 1497–1506. https://doi.org/10.1039/b917969k. 45. van Santen, R. A.; Kramer, G. J. Reactivity Theory of Zeolitic Broensted Acidic Sites. Chem. Rev. 1995, 95 (3), 637–660. https://doi.org/10.1021/cr00035a008. 46. Schröder, K. P.; Sauer, J.; Leslie, M.; Richard, C.; Catlow, A.; Thomas, J. M. Bridging Hydrodyl Groups in Zeolitic Catalysts: A Computer Simulation of Their Structure, Vibrational Properties and Acidity in Protonated Faujasites (HY Zeolites). Chem. Phys. Lett. 1992, 188 (3–4), 320–325. https://doi.org/10.1016/0009-2614(92)90030-Q. 47. Tuma, C.; Sauer, J. Quantum Chemical Ab Initio Prediction of Proton Exchange Barriers between CH4 and Different H-Zeolites. J. Chem. Phys. 2015, 143 (10). https://doi.org/ 10.1063/1.4923086. 48. Jones, A. J.; Iglesia, E. The Strength of Brønsted Acid Sites in Microporous Aluminosilicates. ACS Catal. 2015, 5 (10), 5741–5755. https://doi.org/10.1021/ acscatal.5b01133. 49. Lercher, J. A.; Gründling, C.; Eder-Mirth, G. Infrared Studies of the Surface Acidity of Oxides and Zeolites Using Adsorbed Probe Molecules. Catal. Today 1996, 27 (3–4), 353–376. https://doi.org/10.1016/0920-5861(95)00248-0. 50. Hunger, M. Multinuclear Solid-State NMR Studies of Acidic and Non-Acidic Hydroxyl Protons in Zeolites. Solid State Nucl. Magn. Reson. 1996, 6 (1), 1–29. https://doi.org/ 10.1016/0926-2040(95)01201-X. 51. Heeribout, L.; Semmer, V.; Batamack, P.; Dorémieux-Morin, C.; Fraissard, J. Brønsted Acid Strength of Zeolites Studied by 1H NMR: Scaling, Influence of Defects. Microporous Mesoporous Mater. 1998, 21 (4–6), 565–570. https://doi.org/10.1016/S1387-1811(98)00025-0. 52. Koller, H.; Engelhardt, G.; van Santen, R. A. The Dynamics of Hydrogen Bonds and Proton Transfer in Zeolites – Joint Vistas from Solid-State NMR and Quantum Chemistry. Top. Catal. 1999, 9 (3/4), 163–180. https://doi.org/10.1023/A:1019131126634. 53. Nagy, J. B.; Gabelica, Z.; Debras, G.; Derouane, E. G.; Gilson, J. P.; Jacobs, P. A. 27Al-n.m.r. Characterization of Natural and Synthetic Zeolites. Zeolites 1984, 4 (2), 133– 139. https://doi.org/10.1016/0144-2449(84)90051-4. 54. Sierka, M.; Eichler, U.; Datka, J.; Sauer, J. Heterogeneity of Brønsted Acidic Sites in Faujasite Type Zeolites Due to Aluminum Content and Framework Structure. J. Phys. Chem. B 1998, 102 (33), 6397–6404. https://doi.org/10.1021/jp983843. 55. Kazansky, V. B.; Serykh, A. I.; Pidko, E. A. DRIFT Study of Molecular and Dissociative Adsorption of Light Paraffins by H-ZSM-5 Zeolite Modified with Zinc Ions: Methane Adsorption. J. Catal. 2004, 225 (2), 369–373. https://doi.org/10.1016/j.jcat.2004.04.029. 56. Yeh, Y.-H.; Gorte, R. J.; Rangarajan, S.; Mavrikakis, M. Adsorption of Small Alkanes on ZSM-5 Zeolites: Influence of Brønsted Sites. J. Phys. Chem. C 2016, 120 (22), 12132–12,138. https://doi.org/10.1021/acs.jpcc.6b03855. 57. Sievers, C.; Onda, A.; Olindo, R.; Lercher, J. A. Adsorption and Polarization of Branched Alkanes on H-LaX. J. Phys. Chem. C 2007, 111 (14), 5454–5464. https://doi.org/ 10.1021/jp067312u. 58. Chen, L.; Lin, L.; Xu, Z.; Zhang, T.; Xin, Q.; Ying, P.; Li, G.; Li, C. Fourier Transform-Infrared Investigation of Adsorption of Methane and Carbon Monoxide on HZSM-5 and Mo/ HZSM-5 Zeolites at Low Temperature. J. Catal. 1996, 161 (1), 107–114. https://doi.org/10.1006/jcat.1996.0167. 59. Khodakov, A. Y.; Leonid Kustov, S. M.; Kazansky, V. B.; Williams, C. N. Infrared Spectroscopic Study of the Interaction of Cations in Zeolites with Simple Molecular Probes Part 3.dAdsorption and Polarization of Methane and Ethane on Cationic Forms of Hig h-si lica Zeolites. J. Chem. Soc., Faraday Trans. 1993, 89, 1393–1395. 60. Zecchina, A.; Otero Areán, C. Diatomic Molecular Probes for mid-IR Studies of Zeolites. Chem. Soc. Rev. 1996, 25 (3), 187–197. https://doi.org/10.1039/CS9962500187. 61. Schoonheydt, R. A.; Geerlings, P.; Pidko, E. A.; van Santen, R. A. The Framework Basicity of Zeolites. J. Mater. Chem. 2012, 22 (36), 18705. https://doi.org/10.1039/ c2jm31366a. 62. Rabo, J. A.; Bezmam, R. D.; Poutsma, L. M. Unifying Principles of Zeolite Catalysis. Acta Phys Chem. Szeged 1978, 24, 39. 63. Manz, T. A. Introducing DDEC6 atomic population analysis: part 3. Comprehensive method to compute bond orders y. RSC Adv. 2017. https://doi.org/10.1039/c7ra07400j.

292

Solid acid catalysis; Part I, the zeolite protonic site

64. Liu, C.; Tranca, I. C.; Santen, V.; Hensen, R. A.; Pidko, C. L.; Tranca, I.; Van Santen, R. A.; Hensen, E. J. M.; Pidko, E. A. Scaling Relations for Acidity and Reactivity of Zeolites. J. Phys. Chem. C 2017, 121 (42), 23520–23530. https://doi.org/10.1021/acs.jpcc.7b08176. 65. van Santen, R. A. Modern Heterogeneous Catalysis: An Introduction, Wiley VCH, 2017. 66. Eichler, U.; Brändle, M.; Sauer, J. Predicting Absolute and Site Specific Acidities for Zeolite Catalysts by a Combined Quantum Mechanics/Interatomic Potential Function Approach. J. Phys. Chem. B 1997, 101 (48), 10035–10050. https://doi.org/10.1021/jp971779a. 67. Brändle, M.; Sauer, J. Acidity Differences Between Inorganic Solids Induced by Their Framework Structure. A Combined Quantum Mechanics/Molecular Mechanics Ab Initio Study on Zeolites. J. Am. Chem. Soc. 1998, 120 (7), 1556–1570. https://doi.org/10.1021/ja9729037. 68. O’Malley, P. J.; Dwyer, J. An Ab Initio Quantum Chemical Investigation on the Effect of the Magnitude of the T-O-T Angle on the Brønsted Acid Characteristics of Zeolites. J. Phys. Chem. 1988, 92 (10), 3005–3007. https://doi.org/10.1021/j100321a059. 69. Jacobs, W. P. J. H.; van Wolput, J. H. M. C.; van Santen, R. A.; Jobic, H. A Vibrational Study of the OH and OD Bending Modes of the Brönsted Acid Sites in Zeolites. Zeolites 1994, 14 (2), 117–125. https://doi.org/10.1016/0144-2449(94)90005-1. 70. Jobic, H.; Tuel, A.; Krossner, M.; Sauer, J. Water in Interaction With Acid Sites in H-ZSM-5 Zeolite Does Not Form Hydroxonium Ions. A Comparison Between Neutron Scattering Results and Ab Initio Calculations. J. Phys. Chem. 1996, 100 (50), 19545–19550. https://doi.org/10.1021/jp9619954. 71. van Santen, R. A.; Vogel, D. L. Lattice Dynamics of Zeolites. In Advances in Solid-State Chemistry; Catlow, C. R. A., Ed.; vol. 1; JAI Press Inc., 1989; pp 151–224. 72. Gedeon, A.; Fernandez, C. Solid-State NMR Spectroscopy in Zeolite Science; vol. 168; Elsevier B.V., 2007. Chapter 12. 73. Hunger, M.; Horvath, T.; Engelhardt, G.; Karge, H. G. Multi-Nuclear NMR Study of the Interaction of SiOHAl Groups with Cationic and Neutral Guest-Molecules in Dehydrated Zeolites Y and ZSM-5. Stud. Surf. Sci. Catal. 1995, 94 (C), 756–763. https://doi.org/10.1016/S0167-2991(06)81293-0. 74. Koller, H.; Meijer, E. L.; Van Santen, R. A. 27Al Quadrupole Interaction in Zeolites Loaded with Probe Molecules - A Quantum-Chemical Study of Trends in Electric Field Gradients and Chemical Bonds in Clusters. Solid State Nucl. Magn. Reson. 1997, 9 (2–4), 165–175. https://doi.org/10.1016/S0926-2040(97)00056-8. 75. Hunger, M.; Freude, D.; Pfeifer, H. Magic-Angle Spinning Nuclear Magnetic Resonance Studies of Water Molecules Adsorbed on Brønsted- and Lewis-Acid Sites in Zeolites and Amorphous Silica-Aluminas. J. Chem. Soc. Faraday Trans. 1991, 87 (4), 657–662. https://doi.org/10.1039/FT9918700657. 76. van Koningsveld, H.; Jansen, J. C.; van Bekkum, H. The Orthorhombic/Monoclinic Transition in Single Crystals of Zeolite ZSM-5. Zeolites 1987, 7 (6), 564–568. https:// doi.org/10.1016/0144-2449(87)90099-6. 77. Snurr, R. Q.; Bell, A. T.; Theodorou, D. N. Prediction of Adsorption of Aromatic Hydrocarbons in Silicalite from Grand Canonical Monte Carlo Simulations with Biased Insertions. J. Phys. Chem. 1993, 97 (51), 13742–13752. https://doi.org/10.1021/j100153a051. 78. Sartbaeva, A.; Wells, S. A.; Treacy, M. M. J.; Thorpe, M. F. The Flexibility Window in Zeolites. Nat. Mater. 2006, 5 (12), 962–965. https://doi.org/10.1038/nmat1784.  79. Gábová, V.; Dedecek, J.; Cejka, J. Control of Al Distribution in ZSM-5 by Conditions of Zeolite Synthesis. Chem. Commun. 2003, 3 (10), 1196–1197. https://doi.org/ 10.1039/b301634j. 80. Sastre, G.; Fornes, V.; Corma, A. On the Preferential Location of Al and Proton Siting in Zeolites: A Computational and Infrared Study. J. Phys. Chem. B 2002, 106 (3), 701– 708. https://doi.org/10.1021/jp013189p. 81. Wang, S.; Wei, Z.; Chen, Y.; Qin, Z.; Ma, H.; Dong, M.; Fan, W.; Wang, J. Methanol to Olefins Over H-MCM-22 Zeolite: Theoretical Study on the Catalytic Roles of Various Pores. ACS Catal. 2015, 5 (2), 1131–1144. https://doi.org/10.1021/cs501232r. 82. Haag, W. O.; Lago, R. M.; Weisz, P. B. The Active Site of Acidic Aluminosilicate Catalysts. Nature 1984, 309, 589–591. https://doi.org/10.1038/309589a0. 83. Sarazen, M. L.; Iglesia, E. Stability of Bound Species During Alkene Reactions on Solid Acids. Proc. Natl. Acad. Sci. U. S. A. 2017, 114 (20), E3900–E3908. https://doi.org/ 10.1073/pnas.1619557114. 84. Shustorovich, E. Chemisorption Phenomena: Analytic Modeling Based on Perturbation Theory and Bond-Order Conservation. Surf. Sci. Rep. 1986, 6 (1), 1–63. https://doi.org/ 10.1016/0167-5729(86)90003-8. 85. van Santen, R. A. Theoretical Heterogeneous Catalysis, 1st ed;, World Scientific, 1991. 86. Post, M. F. M.; Huizinga, T.; Emeis, C. A.; Nanne, J. M.; Stork, W. H. J. An Infrared and Catalytic Study of Isomorphous Substitution in Pentasil Zeolites. Stud. Surf. Sci. Catal. 1989, 46 (C), 365–375. https://doi.org/10.1016/S0167-2991(08)60993-3. 87. Jones, A. J.; Carr, R. T.; Zones, S. I.; Iglesia, E. Acid Strength and Solvation in Catalysis by MFI Zeolites and Effects of the Identity, Concentration and Location of Framework Heteroatoms. J. Catal. 2014, 312, 58–68. https://doi.org/10.1016/j.jcat.2014.01.007. 88. Beyer, H. K. Dealumination Techniques for Zeolites BT - Post-Synthesis Modification I. In Post-Synthesis Modification I, Springer: Berlin, Heidelberg, 2002; pp 203–255. 89. Gounder, R.; Jones, A. J.; Carr, R. T.; Iglesia, E. Solvation and Acid Strength Effects on Catalysis by Faujasite Zeolites. J. Catal. 2012. https://doi.org/10.1016/ j.jcat.2011.11.002. 90. Almutairi, S. M. T.; Mezari, B.; Filonenko, G. A.; Magusin, P. C. M. M.; Rigutto, M. S.; Pidko, E. A.; Hensen, E. J. M. Influence of Extraframework Aluminum on the Br{ø}nsted Acidity and Catalytic Reactivity of Faujasite Zeolite. ChemCatChem 2013, 5 (2), 452–466. https://doi.org/10.1002/cctc.201200612. 91. Liu, C.; Li, G.; Hensen, E. J. M.; Pidko, E. A. Nature and Catalytic Role of Extraframework Aluminum in Faujasite Zeolite: A Theoretical Perspective. ACS Catal. 2015, 5 (11), 7024–7033. https://doi.org/10.1021/acscatal.5b02268. 92. Deng, C.; Zhang, J.; Dong, L.; Huang, M.; Li, B.; Jin, G.; Gao, J.; Zhang, F.; Fan, M.; Zhang, L.; Gong, Y. The Effect of Positioning Cations on Acidity and Stability of the Framework Structure of y Zeolite. Sci. Rep. 2016, 6 (March), 1–13. https://doi.org/10.1038/srep23382. 93. Schüßler, F.; Pidko, E. A.; Kolvenbach, R.; Sievers, C.; Hensen, E. J. M. M.; Van Santen, R. A.; Lercher, J. A. Nature and Location of Cationic Lanthanum Species in High Alumina Containing Faujasite Type Zeolites. J. Phys. Chem. C 2011, 115 (44), 21763–21776. https://doi.org/10.1021/jp205771e. 94. Liu, C.; Van Santen, R. A.; Poursaeidesfahani, A.; Vlugt, T. J. H. H.; Pidko, E. A.; Hensen, E. J. M. M. Hydride Transfer versus Deprotonation Kinetics in the Isobutane–Propene Alkylation Reaction: A Computational Study. ACS Catal. 2017, 7 (12), 8613–8627. https://doi.org/10.1021/acscatal.7b02877. 95. Gauthier, C.; Chiche, B.; Finiels, A.; Geneste, P. Influence of Acidity in Friedel-Crafts Acylation Catalyzed by Zeolites. J. Mol. Catal. 1989. https://doi.org/10.1016/03045102(89)85065-5. 96. Feller, A.; Guzman, A.; Zuazo, I.; Lercher, J. A. On the Mechanism of Catalyzed Isobutane/Butene Alkylation by Zeolites. J. Catal. 2004, 224, 80–93. https://doi.org/10.1016/ j.jcat.2004.02.019. 97. Weitkamp, J. Isomerization of Long-Chain n-Alkanes on a Pt/CaY Zeolite Catalyst. Ind. Eng. Chem. Prod. Res. Dev. 1982, 21 (4), 550–558. https://doi.org/10.1021/ i300008a008. 98. Weitkamp, J.; Jacobs, P. A.; Martens, J. A. Isomerization and Hydrocracking of Cg through Cl6 n-Alkanes on Pt/HZSM-5 Zeolite. Appl. Catal. 1983. https://doi.org/10.1016/ 0166-9834(83)80058-X. 99. Teunissen, E. H.; Jansen, A. P. J.; van Santen, R. A.; Orlando, R.; Dovesi, R. Adsorption Energies of NH3 and NHþ 4 in zeolites corrected for the long-range electrostatic potential of the crystal. J. Chem. Phys. 1994, 101, 5865. https://doi.org/10.1063/1.467303. 100. Bordiga, S.; Platero, E. E.; Areán, C. O.; Lamberti, C.; Zecchina, A. Low temperature CO adsorption on Na-ZSM-5 zeolites: An FTIR investigation. J. Catal. 1992, 137 (1), 179– 185. https://doi.org/10.1016/0021-9517(92)90147-A. 101. Chester, A. W.; Derouane, E. G. Zeolite Chemistry and Catalysis, Springer Netherlands: Dordrecht, 2009. 102. Hattori, H.; Arudra, P.; Abdalla, A.; Aitani, A. M.; Al-Khattaf, S. S. Infrared Study of Silanol Groups on Dealuminated High Silica MFI Zeolite to Correlate Different Types of Silanol Groups with Activity for Conversion of 1-Butene to Propene. Catal. Lett. 2020, 150 (3), 771–780. https://doi.org/10.1007/s10562-019-02972-8. 103. Heitmann, G. P.; Dahlhoff, G.; Hölderich, W. F. Catalytically Active Sites for the Beckmann Rearrangement of Cyclohexanone Oxime to $3$-Caprolactam. J. Catal. 1999, 186 (1), 12–19. https://doi.org/10.1006/jcat.1999.2552.

6.13

Solid acid catalysis; Part II, catalytic chemistry of proton activation

Rutger A. van Santen, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, MB Eindhoven, The Netherlands © 2023 Elsevier Ltd. All rights reserved.

6.13.1 6.13.2 6.13.3 6.13.4 6.13.4.1 6.13.4.2 6.13.4.3 6.13.5 References

Introduction Elementary proton activated reactions Contribution of the adsorption free energy Confinement Transition state stabilization Stereoselectivity Micro pore equilibration Conclusion

293 294 297 300 300 303 303 306 308

Abstract A catalytic reaction is a complex cyclic reaction network of a sequence of elementary reactions. The initial part of this article considers the rate of elementary reactions where organic molecules are activated by zeolite protons. It will be seen that protonated intermediates as carbonium or carbenium cations are part of transition states are activated intermediates in solid acid catalysis. In part I an analysis of the proton affinity as a function of zeolite composition has been given. In this article we will emphasize the importance of confinement, that relates to match of molecular dimension and zeolite micro-cavity. Confinement causes large differences in adsorption free energies when zeolite structure is varied. A kinetic analysis is given that leads to the definition of a catalytic analogue of the classical Hammett function of acidity. Selected examples of catalytic reaction cycles are chosen that illustrate the interplay of confinement and proton affinity in the comparison of zeolites with different structures and composition.

6.13.1

Introduction

In this article we will discuss chemical reactivity of the zeolite proton. In the catalytic reaction protons activate chemical bonds of reactant molecules, that adsorb in the micro-cavities in the zeolite. Whereas the proton affinity of isolated protons only weakly depends on zeolite structure, this is very different when solid acids are applied as catalysts. Catalyst performance depends on match or mismatch of reactant, reaction intermediate or product molecules with size and shape of zeolite catalyst micro cavity. This dependence is called the confinement effect. It controls to a large extend the variation in activity of zeolites when reactivity is compared between zeolites of different structure with protons of comparable proton affinity (PA). Proton affinity depends strongly on composition of the zeolite, that we discussed in detail in the previous Chapter . (part I). Here we will see that chemical bond activation may very strongly depend on proton affinity and hence also zeolite composition. In part I we also defined the three classes of proton affinities with deprotonation energy differences of the order of 30 kJ/mol. Demands for a particular proton PA for it to activate chemical substrate bonds may differ for different molecular chemical bonds. Activation of saturated CeC bonds in alkanes requires protons of lower PA, whereas molecules with unsaturated C]C bonds can be activated also by protons of less reactivity and higher PA. From a chemical bonding perspective proton activation of a chemical bond and the confinement effect are quite different. Proton activation involves cleavage and formation of covalent chemical bonds. The confinement effect is dominated by the dispersive van der Waals interaction between adsorbed molecule and zeolite wall. This interaction is dominated by polarizability of mainly the zeolite oxygen atoms and the atoms of the adsorbed molecule. Since the anionic oxygen atoms, with a formal charge
2 are larger than the smaller Si4þ and Al3þ cations, the polarizability of the former is the larger. The dispersive van der Waals interaction between adsorbate and zeolite wall depends on the distance of the polarizable adsorbate bonds and the zeolite wall oxygen atoms. In a large pore a molecule interacts with a more limited number of oxygen atoms and has less of a steric constraint, than in a smaller pore with a larger curvature. This gives has a large consequence for micro-pore occupation due to differences of adsorption free energies when zeolites are compared.1 The larger the micro pore occupation with reactant the faster the reaction rate. This happens as long molecular adsorption is not too strong. Then micro pores become highly occupied and desorption rate of products can become rate limiting. Then equilibrium between molecules in the zeolite micro pore and molecules external of the zeolite will not be maintained and product distribution reflects internal molecular equilibria.

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00010-8

293

294

Solid acid catalysis; Part II, catalytic chemistry of proton activation

Steric constraints due to mismatch of size and shape will also inhibit particular reactions. Then stereo selective catalysis results. Catalyst stability will also be influenced because formation of large bulky deactivating molecules can become inhibited. The interplay of confinement and proton affinity will be an important topic of this article. In the first following section proton activated elementary reaction steps will be considered. Computational catalysis, that provides detailed information on activation free energies of elementary reaction events and on adsorption free energies of reaction intermediates has become important to relate structure and composition with experimentally measured catalyst performance.2–4 In the next section we will show how this energetics has to be integrated with adsorption of reactant and desorption of product molecules to give overall kinetics of the catalytic reaction. According to reaction kinetics, the rate controlling step defines the state of the surface. As we will see in Section 6.13.3 in solid acid catalysis this state is usually not the free proton, but a surface reaction intermediate. Often one reaction intermediate dominates. This Major Reaction Intermediate (MARI) may vary as a function of reaction conditions and depends on the mechanism of the reaction.5,6 We will define a Catalytic Acidity (CatA) function analogous to the Hammett function (Chapter ., part I, Eq. 1). This function explicitly depends on the definition of the MARI, the activation free energy of the rate controlling reaction step and adsorption free energy of reactant in the zeolite micro-pore. Confinement is the main subject of Section 6.13.4, where we focus on the stereo selectivity of zeolite catalyzed reactions. In Section 6.13.5 the reaction mechanisms of important catalytic reactions as hydro-isomerization, alkylation, catalytic cracking and methanol conversion are introduced and used to illustrate the complex relation between catalyst performance and inorganic catalyst material reactivity descriptors. This section will also serve as the summarizing and concluding section.

6.13.2

Elementary proton activated reactions

In this section we will consider the energetic changes of a molecule that transforms from its initial non reacted state, when adsorbed to the proton to a reaction intermediate molecule. We will see that notwithstanding the considerably lower reactivity of the zeolite proton compared to that of the protons in liquid acids, that reaction intermediates are analogous to those known from the physical organic chemistry of liquid acid catalysis.7,8 Protonation of unsaturated molecules gives a carbenium ion carbo-cations. A carbenium ion is a relatively stable cation. The positive charge is located on a carbon atom, that can be considered sp2 hybridized and is bonded to three neighbor atoms. The arrangement is planar. Attachment of a proton to the saturated molecular CeC bond gives a so-called non classical carbonium ion, since then the carbon atoms next to the protonated bond become five valent. Such a carbonium is unstable and will rapidly undergo subsequent bond transformation reactions. When formed on solid acids, carbenium and carbonium ions are part of transition states or are activated reaction intermediates. We will illustrate this here for solid acid catalyzed oligomerization of propylene, that contains a coordinatively unsaturated C] C bond and the bond cleavage reaction of coordinatively saturated propane. In Fig. 1 activation is compared with FAU zeolite models of low Al/(Al þ Si) framework composition, that are nonpromoted or promoted by three valent cationic oxy-hydroxy clusters (see Fig. 15, part I for the corresponding zeolite model structures). For the respective reactions DFT computed reaction energy diagrams are given. They show the differences in energies of initial adsorbed state, activated reaction intermediate or transition state and final reacted state. In Fig. 1A and B also the structures of the successive intermediate states are shown. As Fig. 1A illustrates protonation of adsorbed propylene leads to the propyl-alcoxy species as a stable reaction intermediate, that becomes covalently attached to a zeolite wall oxygen atom. It is slightly more stable than molecular propylene when hydrogen bonded to the zeolite proton. Formation of the alcoxy species is an activated process, that proceeds through an unstable transition state intermediate. The structure of this transition state intermediate is interesting, since it is similar to that of the propyl carbenium cation. In Fig. 1A not only the transformation of p adsorbed molecular propylene to the s bonded alcoxy intermediate is shown, but also the consecutive dimerization reaction step where a second propylene molecule reacts with the propyl alcoxy intermediate. CeC bond formation occurs in the positively charged transition state. As comparison of the respective FAU and La-FAU models illustrates, the lower DPE of the latter zeolite proton results in lower transition state energies. Bond cleavage of the covalent OH bond or OC bond depends similarly on the zeolite oxygen atom reactivity changes by the La oxy-hydroxy complex. The activation energy of the dimerization reactions decreases with decrease of the DPE. Fig. 1B illustrate that the activation of a s bond in propane proceeds through a substantially higher transition state energy than that of the unsaturated p bond of the alkene. The difference in activation energies is due the low proton affinity the CeC or CH s bonds. It leads to an unstable carbonium ion transition state intermediate that is the analogue of the non-classical carbonium ion. The relatively low activation energies of alkene oligomerization makes this reaction often relatively fast in comparison with other solid acid catalyzed reaction steps, as desirable reactions as CeC bond cleavage or hydrocarbon rearrangement reactions. Alkenes are common intermediates of solid acid catalyzed reactions. It causes solid acid catalyzed reaction to readily deactivate, since oligomerization reactions lead to carbonaceous products that deactivate the catalyst. As comparison of Fig. 1A and C illustrates the change in DPE by La or Al promotion affects the activation energies of the CeC bond cleavage or CeC bond formation reactions similarly. The respective transition states are late on their respective OH stretch

Solid acid catalysis; Part II, catalytic chemistry of proton activation

295

Fig. 1 Proton activation of propylene and propane. (A) Comparison of protonation and oligomerization of propylene by protons of low Al/(Al þ Si) framework FAU and La-FAU systems. Reaction intermediate, transition state energies and respective structures. (B1) The energy diagram of the propane CeC bond cleavage reaction activated by low Al/(Al þ Si) FAU. (B2) Proton activated propane dehydrogenation to H2 and Propylene on low Al/(Si þ Al) FAU. (C) Energy diagram of propane cracking in faujasite zeolites. EFAl-free (H-FAU Si/Al ¼ 7) and EFAl-containing (EFAl/H-FAU Si/ Al ¼ 7) faujasite models used for propane cracking reactions. Energy differences between adsorbed propane and carbonium ion transition intermediate for bond cleavage CeC bond cleavage reaction. (A and B) Reproduced with permission from Catlow, C. R. A.; Van Speybroeck, V.; Van Santen, R. A. In: Modelling and Simulation in the Science of Micro- and Meso-Porous Materials. Elsevier, 2017, Chapter 5; (C) Reproduced with permission from Liu, C.; Li, G.; Hensen, E. J. M.; Pidko, E. A. Nature and Catalytic Role of Extraframework Aluminum in Faujasite Zeolite: A Theoretical Perspective. ACS Catal. 2015, 5(11), 7024–7033, https://doi.org/10.1021/acscatal.5b02268.

reaction coordinates,2,3,9 so that covalent interaction between carbocation and zeolite lattice is essentially absent. Differences in activation energies are dominated by DPE differences and are of the same order of magnitude of 60 kJ/mol. Two different reactions happen when propane is activated by the proton. CeC bond cleavage may occur and final products are methane and ethylene (Fig. 1B1) or the CeH bond cleaves and H2 and propylene are products (Fig. 1B2). Whereas both reactions proceed through a carbonium ion intermediate, the respective transition states are different. This is a general phenomenon. Different reactions proceed through different transition states. As has also been found in earlier quantum-chemical cluster calculations10 the activation energy of CeC bond cleavage in propane is lower than that of the CeH bond cleavage reaction. Computational prediction and experiment are in quantitative agreement.11 The difference in activation energies relates to bond strength differences. The CH bond is stronger than CeC bond.12,13

296

Solid acid catalysis; Part II, catalytic chemistry of proton activation

The temperature dependence of elementary reaction rates derives from the Gibbs free activation energy. Differences in activation entropies may counteract differences in activation energies. In Ferrierite, that has one dimensional narrow 10 ring micro channels, the reaction rate of CeC bond cracking reaction is equal to the rate of dehydrogenation. This is in contrast to MFI that has a crossing channel system, that provides wider space, where the rate of CeC bond cracking is found to be faster14. This stereo selective structure dependence is due to frustrated rotation and limited mobility of larger molecular fragments in the constrained environment of narrow micro-pore channels. Consideration of rotational and translational entropy is also relevant to the question whether at reaction temperature carbenium ions are dominantly present as adsorbed alcoxy species or that mobile carbenium ion intermediates dominate. At a temperature of 600 K mobile carbenium ions are expected to dominate as long as the energy difference with respective alcoxy intermediates is less than 160 kJ/mol.15 The energy difference between alcoxy intermediate and carbenium ion not only varies largely with proton affinity. It is also structure dependent. The stability of adsorbed alcoxy will depend on curvature of zeolite cavity. Bulky carbenium ions as the tertiary carbenium ions experience repulsive interactions when zeolite cavity wall has strong curvature and contact of a positively charged carbon atom with negatively charged can become inhibited.16 This will favor the mobile carbenium ion state. One distinguishes primary, secondary and tertiary carbenium ions, that have different relative stabilities. The difference is whether the positive charge is at the primary carbon atom at the molecular end or is located on a secondary carbon atom with two carbon atom neighbors or a tertiary carbon atom with three carbon atom neighbors. The relative stabilities of primary, secondary or tertiary carbenium ions increase by increments of approximately 40 kJ/mol.17 Unless repulsive interactions of the CH2 or CH3 groups that are part of carbenium ions with the zeolite wall dominate, the strength of CeO alcoxy bonds does not vary when a primary, secondary or tertiary is attached. The large difference in relative stabilities of gasfase carbenium ions causes tertiary carbenium ions to be present as mobile intermediates in the zeolite micro pore at relatively low temperature. This is an important conclusion with respect to catalysis because tertiary carbenium ions tend to dominate as reaction intermediates in many solid acid catalyzed reactions. An interesting illustration of the difference in reactivity of primary, secondary and tertiary carbenium carbocations is provided by the calculations of Tuma and Sauer.18 Calculated Gibbs free energies of protonated intermediates of isobutene in the 10 ring zeolite FER structure are presented as a function of temperature in Fig. 2. The Gibbs energy of the tertiary isobutyl cation (4) in Fig. 2 is compared with that of the covalently adsorbed primary isobutyl alcoxy (5) and tertiary isobutyl alcoxy (3). When temperature increases the relative stability of the tertiary butyl cation increases steeply compared to the tertiary surface isobutyl alcoxy species and crosses the free energy of the primary alcoxy species at 120 K. The tertiary alcoxy species is destabilized with respect to the primary alcoxy species, that is end on connected of with a lattice oxygen atom, because its attachment to the zeolite wall though the carbon atom that has three methyl group neighbors leads to additional repulsive interaction due to large zeolite cavity curvature. In FER the isobutyl carbenium cation is not a transition state but an activated mobile reaction intermediate. It is more stable than the corresponding alcoxy species. But it is less stable than hydrogen bonded isobutene (2, in Fig. 19), that at finite temperature dominates as zeolite proton site.

Fig. 2 Schematic plot of the Gibbs energy difference DGT as a function of temperature T (at 1 atm) for the formation of protonation intermediates and gasfase isobutene and H-FER. (2) p adsorbed isobutene, (3) tert-butoxide alcoxy, (4) tert-butyl carbenium cation, (5) primary isobutoxide alcoxy. Due to their larger motional entropy the tertiary carbenium cation becomes increasingly more stable when temperature increases. Due to steric constraints the tertiary carbenium is already more stable than tertiary alcoxy at zero Kelvin in narrow pore Ferrierite. After Tuma, C.; Sauer, J. Protonated Isobutene in Zeolites:Tert-Butyl Cation or Alkoxide? Angew. Chem. Int. Ed. 2005, 44(30), 4769–4771, https://doi.org/10.1002/anie. 200501002.

Solid acid catalysis; Part II, catalytic chemistry of proton activation

297

In this section we discussed prototype examples of elementary reaction constants of proton activated reactions, where in the initial state of the reaction reactant is in contact with proton site. In the next section we discuss kinetics of a catalytic reaction where also adsorption and desorption of molecules between external part of zeolite and the internal part of zeolite have to be included.

6.13.3

Contribution of the adsorption free energy

To appreciate the role of reactant adsorption and product desorption in catalytic kinetics it is useful to consider the Langmuir reaction rate expression of a mono molecular reaction with a single reaction rate controlling step: R ¼ kr q ¼ kr Kads $

½Cr  1 þ Kads ½Cr 

(1a)

R ¼ kr ; Kads ½Cr i1

(1b)

R ¼ kr Kads ½Cr ; Kads ½Cr h1

(1c)

In Equations 1a–c kr is the elementary reaction rate constant of the rate controlling step. Kads is the adsorption equilibrium constant of reactant and q the surface coverage. [Cr] is reactant concentration. High versus low coverage defines the order of the reaction. When Kads[Cr] > 1, surface coverage is high and the surface is saturated with adsorbate. Reaction rate is zero order in reactant concentration and is equal the elementary reaction rate constant kr (Eq. 1b). The apparent activation energy of reaction is equal to Eact of kr. When Kads[Cr] < 1 reaction is first order in reactant concentration [Cr] and the activation energy of reaction becomes: Eapp act ¼ Eact þ Eads $ Eact (positive number). Eads (negative number) is the adsorption energy of reactant. The apparent reaction energy is now decreased. The reaction rate R is now proportional to kr and the adsorption equilibrium constant kads (Eq. 2). The free proton is the dominant surface state. Generally the apparent activation energy of the catalytic reaction will depend on surface coverage and vary with reaction condition: app

Eact ¼ Eact þ ð1
qÞ$Eads

(2)

An example of a monomolecular reaction with first order kinetics is the alkane cracking reaction, for which we the elementary proton activation reaction steps discussed in Fig. 1B. At low conversion the alkane molecules undergo direct CeC and CeH bond cleavage reactions according to the Haag-Dessau mechanism.19 This reaction has been experimentally studied for alkane molecules with varying reaction length catalyzed by low Al/Si ratio ZSM-5 by Nabeshuber et al.20 at conditions where q is low. Respective reaction rates as a function of temperature are shown in Fig. 3. The slopes of the rate curves give Eapp act . Using Eq. (2) the activation energy of the elementary reaction rate constants can be deduced by subtracting the adsorption energies of respective reactant molecules given in Table 1 (see also Fig. 4). Eapp act decreases with increasing hydrocarbon bond length, but the Eact of the elementary reaction rate constant is a constant. Its value is comparable to the calculated value of the CeC bond cleavage reaction of propane (Fig. 1B), that dominates the reaction. The adsorption energy of reaction is alkane chain length dependent because it depends on match of hydrocarbon size or shape with micro pore cavity dimension. It is a consequence of confinement. Activation energies of the bond activating elementary

Fig. 3 Arrhenius plot for protolytic cracking of n-alkanes of varying length catalyzed by H-ZSM-5. Reproduced with permission from Narbeshuber, T. F.; Vinek, H.; Lercher, J. A. Monomolecular Conversion of Light Alkanes over H-ZSM-5. J. Catal. 1995, 157(2), 388–395, https://doi.org/10.1006/ jcat.1995.1304.

298

Solid acid catalysis; Part II, catalytic chemistry of proton activation Table 1 Reactant Propane n-Butane n-Pentane n-Hexane

Heats of adsorption and respective activation energies for protolytic alkane cracking. Heat of adsorption (kJ/mol) 43 62 74 92

app Eact (kJ/mol) 155 135 120 105

Eact (kJ/mol) 198 197 194 197

After Fig. 20. Narbeshuber, T. F.; Vinek, H.; Lercher, J. A. Monomolecular Conversion of Light Alkanes over H-ZSM-5. J. Catal. 1995, 157(2), 388–395, https://doi.org/10.1006/jcat.1995. 1304.

reactions with the proton are similar. They are independent of hydrocarbon length since they react locally with individual chemical bonds of the molecules. The dependence of the activation energy of the catalytic reaction on adsorption energy is also the reason that zeolite catalyzed reactions are zeolite structure dependent. Differences of adsorption energies of zeolites with different micro pore size are illustrated in Fig. 4 for adsorption of n-alkanes in MFI and FAU. The adsorption energy in MFI with 10 ring channels for the longer alkanes is close to twice that of the wider micro pore FAU structure. The hydrocarbon becomes in closer contact with the zeolite wall oxygen atoms when the pore dimension is smaller. Fig. 4 also illustrates the incremental increase of alkane adsorption energy with carbon number that is also the result of increased contact with zeolite wall oxygen atoms. Compared to interaction with the siliceous zeolite wall, the hydrocarbon when adsorbed to the proton has a small additional interaction with the zeolite of the order of 5–10 kJ/mol. The structure dependence of a catalytic reaction we will illustrate for the hydro isomerization reaction of hexane catalyzed by zeolites with different micro-pore structure. In Table 2 the rate of hydro isomerization reaction of hexane is compared for four different zeolites. Two zeolites have one dimensional channels: zeolite Mordenite has a 12 ring microchannel, and zeolite HZSM22 (TON) has a 10 ring channel. The other two Zeolites b and ZSM-5 have a three dimensional microchannel structure that consist of respectively 12 and 10 rings, but their effective cavity sizes due to channel intersections are comparable. Table 2

Measured turn over frequencies (TOF, reaction rate normalized per proton) of the hydro isomerization reaction and adsorption equilibrium constants of hexane Kads (T ¼ 550 K; pH2/pnC6 ¼ 14).

Zeolite

TOF (s
1)

Kads (Pa
1)

kiso (s
1)

H-Beta H-MOR H-ZSM-5 H-ZSM-22

5.0  10
3 1.1  10
2 4.1  10
3 1.6  10
3

6.4  10
5 3.3  10
4 3.3  10
5 1.4  10
5

1.7  10
2 2.7  10
2 2.8  10
2 1.7  10
2

Elementary reaction rate kiso is deduced from expressions analogous to Eq. (1a). After Van Santen R. A.; De Gauw, F. J. M. M. Acidity in Zeolite Catalysis. Stud. Surf. Sci. Catal. 2000, 130A(2), 127–135, https://doi.org/10.1016/s0167-2991(00)80949-0.

Fig. 4 Heats of adsorption of alkanes in different zeolites. Siliceous MFI an FAU (closed squares and circles) are compared with proton containing materials (open squares or circles) as a function of n-alkane chain length. Reproduced with permission from Eder, F.; Lercher, J. A. Alkane Sorption in Molecular Sieves: The Contribution of Ordering, Intermolecular Interactions, and Sorption on Brønsted Acid Sites. Zeolites 1997, 18(1), 75–81, https://doi.org/10.1016/S0144-2449(96)00127-3.

Solid acid catalysis; Part II, catalytic chemistry of proton activation

299

In the hydro isomerization reaction a C5 or C6 n-alkane isomerizes to the corresponding isoalkane. In the hexane isomerization reaction the rate controlling step kiso is the rate of conversion of the adsorbed n-hexyl alcoxy intermediate to the isohexyl cation.21 By fitting measured reaction rate data to reaction rate expressions similar as Eq. (1a) and measurement of the Henry adsorption constant Kads at reaction temperature the elementary reaction rate constant, kiso can be determined.22 The data of the three dimensional network zeolites show little difference (their effective cavity dimensions do not differ significantly), but a large difference in reactivity is found for the one dimensional systems. An increased rate of reaction per proton (TOF) is measured for the one dimensional 12 ring Mordenite zeolite compared to the rate of reaction in the more narrow one dimensional 10 ring ZSM-22 (TON) system. Whereas the measured reaction rates (TOF) differ by a factor of 7, the values deduced for the elementary reaction rate constant kiso are essentially the same. Proton activation energies do not differ on the different zeolites as expected form their chosen low Al/Si ratio’s. In contrast the Henry coefficients of hexane adsorption differ by a factor 20. Different from expectation based on the adsorption energies, that are
19 kJ/mol for MOR and
35 kJ/mol for TON respectively the Henry coefficient is larger for the wide pore Mordenite than for the narrow pore H-ZSM-22. The adsorption equilibrium constants do not follow the trend in adsorption energies. The difference is dominated by differences in adsorption entropies. The adsorption entropy of adsorption in narrow pore TON is less than in the wide pore MOR because of constrained mobility in the narrow pore zeolite which causes the Henry coefficient of MOR to be the larger. The larger reactivity of Mordenite is due to the resulting a higher pore occupation by hexane in Mordenite compared to that in ZSM-22 (see Eq. 1b). Whereas at the particular reaction conditions used the site coverage q in Mordenite is 40% in H-ZSM-5 the coverage is only 6%. The compensation effect of adsorption entropy decrease when adsorption energy increases may cause reaction rates to vary much less with change in micropore dimension than one would expect from variation in apparent activation energies23 Generally at high temperature the entropy effect will dominate and large pore zeolites will be most active. At the lower temperatures the adsorption energy difference will dominate. In the case of the hydro isomerization reaction the entropy gain at the higher temperature for the wide pore zeolite dominates over adsorption energy loss. Since for the low Al/(Al þ Si) systems the DPE of the protons is independent of structure the invariance of kiso implies that entropy changes of initial reactant state and transition state in this case cancel. Isomerization does not change in molecular carbon atom number. For a detailed discussion of transition state entropies we refer to.24 Whereas from a catalytic point of view one would be tempted to assign the differences in catalytic rates of the zeolites to differences in acidity, the intrinsic reactivity of the protons is similar. Differences in reactivity relate to structural variation that results in varying adsorption equilibrium constants, that lead to different micro pore occupation. Reaction rate varies since it relates to micro pore occupation. One can define an expression analogous to the equilibrium Hammett acidity function (see part I, Chapter ., Eq. 1) for kinetic phenomenological catalytic acidity (CatA). According to Eyring transition reaction rate equation the transition state can be considered quasi equilibrated with the reactant state. We can use this quasi-equilibrium assumption to formulate an equilibrium equation for the transition state coverage of the reactive site equation as an equilibrium between gasfase reactant concentration [Cr] and transition state site coverage qTr: qTr =ð1
qTr Þ ¼ Kquasi ½Cr 

(3a)

In this expression Kquasi has be deduced from the reaction rate expression. For this we will use Eq. (1c), valid for a monomolecular reaction in the limit of low surface coverage: Kquasi ¼

h kr Kads kT

(3b)

Analogous to the definition of Hammett function we define CatA: Cat A ¼
ln½Cr  ¼ ln Kquasi
ln qTr =ð1
qTr Þ We can define the catalytic acidity function as the value of CatA, where qTr ¼ 1
qTr:     1 h Cat A qTr ¼ ¼ ln kr Kads ¼
ðGact:r þ Gads Þ=kT 2 kT

(4)

(5)

The Hammett function is more negative when proton reactivity increases. In contrast CatA becomes less negative when reactivity of catalyst increases. The state of the proton, can be the free proton free proton but also an intermediate alcoxy or mobile carbenium ion species (carbenium). This state is Major Reactant Intermediate (MARI) if the reaction and serves as reference to the calculation of kr. It may change with reaction conditions and depends on reaction mechanism. We see that the CatA explicitly depends on confinement25 through the adsorption equilibrium of reactant in the catalyst micro-pore. For the reactions considered in this section diffusion in the zeolite channels is assumed fast compared to reaction rate. Due to match of molecular size and shape with zeolite micro cavity size and shape diffusion is a floating motion and relatively fast. In the narrow micro pores it is not dominated by inter molecular interaction as in Knudsen diffusion but by interaction with the zeolite wall. Intra-zeolite diffusion constants can become are independent of hydrocarbon chain length.1,26,27

300

6.13.4

Solid acid catalysis; Part II, catalytic chemistry of proton activation

Confinement

The nanometer size of zeolite micro channels and cavities leads to the sensitivity of the interaction of reaction intermediates to the match of molecular shape and size and that of the zeolite micro cavity.26 The attractive dispersive van der Waals interactions with the polarizable zeolite wall oxygen atoms stabilizes molecules adsorbed in the micro channels. The charged carbenium ion or carbonium ion reaction intermediate molecules experience additional stabilization due to screening of charge that relates to the dielectric constant of the zeolite material. It depends on the distance of between the positive charges on the organo-cation atoms and the zeolite oxygen wall atoms. Mismatch of shape or size will lead to repulsive interactions. We discussed the repulsion of bulky alcoxy species adsorbed to zeolite cavities of large curvature. In Section 6.13.4.1 we will show how the balance of attractive and repulsive interactions affects methanol addition to aromatic molecules, an elementary reaction important to MTO (methanol to olefin) catalysis. Steric repulsion effects also give rise to stereo selective catalysis. In Section 6.13.4.2 we will discuss the stereo selectivity of the alkylation reaction of methanol with toluene as an example. In the previous section we discussed reaction kinetics where diffusion is not limiting. This is a valid assumption in zeolite catalysis for reactions with molecules without steric constraint as long as the micro pores are relatively empty. This changes when for reactions with larger molecules at intermediate temperatures and higher reactant pressures micro pores become highly occupied. This high micro pore occupation than will suppress molecular mobility. As we will discuss in the final Section 6.13.4.3 then equilibrium between the molecules adsorbed in the internal part of the zeolite and external of the zeolite is not maintained and reaction kinetics becomes dominated by equilibration of reaction intermediates internal of the zeolite.

6.13.4.1

Transition state stabilization

Confinement may stabilize as well as destabilize transition states. We will illustrate this here for a elementary reactions that are important to the Methanol to Gasoline28 and related Methanol to Olefin processes.29,30 Whereas the 10 ring channel system of H ZSM-5 (MFI) yields aromatics as main product production of the smaller olefins requires the use of zeolites that are large that are cavities connected by 8 rings (in Section 6.13.4.2 this reaction is discussed in more detail). Zeolite materials with this topology and low proton concentration are only available with the SieAlPO4 lattice where Si substitution for P in the AlPO4 material creates

SAPO-35

120

DNL-6

SAPO-34

120

120

100

100

100

80

80

80

60

60

60

40

40

40

20

20

20

0

0

0

TMB

PMB

HMB

TMB

PMB

HMB

TMB

PMB

HMB

Fig. 5 Three SAPO lattices are compared for methylation by methanol of tetramethyl benzene (TMB), pentamethyl benzene (PMB) and hexamethylbenzene (HMB). SAPO-35 with the smallest cavities (0.73  0.63 nm) exhibits the higher selectivity toward ethylene; SAPO-34 (1.09  0.67 nm) yields ethylene and propylene as the main products; DNL-6 with larger cavity space (1.27  1.16 nm) favors propylene and butylene formation. Reproduced with permission from Li, J.; Wei, Y.; Chen, J.; Xu, S.; Tian, P.; Yang, X.; Li, B.; Wang, J.; Liu, Z. Cavity Controls the Selectivity: Insights of Confinement Effects on MTO Reaction. ACS Catal. 2015, 5(2), 661–665, https://doi.org/10.1021/cs501669k.

Solid acid catalysis; Part II, catalytic chemistry of proton activation

301

Fig. 6 Confinement energetics without steric constraints (schematic). Comparison of stabilization in small (no prime) and large micro pore (prime) respectively. Eact0 > Eact. E0 is reactant energy in gas phase, G0 is free energy gas phase. (A) Energies and (B) Gibbs free energies.

the lattice surface charge that stabilizes a zeolite proton, that is attached to an oxygen atom that bridges a Si and P containing tetrahedron. In the methanol conversion reaction after a short initiation period reaction propagates though reaction of methanol with molecules that are part of the so-called hydrocarbon pool. They are substituted aromatic benzene molecules. The addition reaction with methanol depends strongly of match of micro pore and aromatic molecule.31 Fig. 532 illustrates the stabilization of cation intermediate with decrease in cavity size and onset destabilization when cationic reaction intermediate becomes too large. The DFT calculated free energies are done for the different SieAlPO4 lattices. Methylated benzene cations, that are key intermediates of the MTO reaction, are considered with varying degree of methylation.29,31 Selectivity derives from the relative stability of the respective methylated carbenium ions of tetramethyl benzene (TMB), pentamethylbenzene (PMB) and hexamethylbenzene (HMB). These respective carbenium ions have increasing molecular size. Methylation of TMB has the higher barrier in SAPO 34 compared to SAPO 35 or DNL-6, because this smaller methylated benzene cation is least electrostatically stabilized in the larger SAPO 34 cavity. Of the three methylbenzenes in small cavity SAPO 35 the largest has the highest activation energy. In this case steric constraint dominates. The larger cavity of DNL-6 does not show this steric constraint. Again the smaller cation has the least transition state stabilization. Schematic Fig. 6 summarizes confinement energetics. For a wide pore and small pore cavity apparent activation energy as well as app (the experimentally measured free energy differences are compared. When site coverage is high the apparent activation energy Eact activation energy) is equal to Eact the intrinsic activation energy of the elementary reaction rate constant of the rate controlling reaction step (Eq. 1b). Then in the absence of steric constraints the activation energy difference between reaction in wide pore and narrow pore cavity is small and equal to Eact0
Eact. It is due to the larger screening of cation charge in the smaller cavity. When cavity coverage is low adsorption affects the apparent activation energy. The apparent free energy of activation is equal to the difference of Eact and Eads (Eq. 1c). The activation barrier is now to be measured with respect to the reactant energy E0 in gas phase. 0 app The difference between large micro pore Eapp act and small micro pore Eact is equal to the difference in adsorption energies, but added is the difference in electrostatic stabilization of the cationic transition state (Eact0
Eact). The latter difference is similar for the activation free energies. The difference between G0 and G is less that E0 and E because of the higher entropy of the adsorbate in the larger cavity. As we have seen for the hydroisomerisation reaction at high temperature, when the entropy difference dominates the difference in the order of Gibbs free energies can invert.

302 Solid acid catalysis; Part II, catalytic chemistry of proton activation Fig. 7 Simulation of shape selective toluene alkylation. (A) Reaction intermediates of the alkylation reaction of toluene with methanol catalyzed by acidic zeolite to p-xylene. Re-p, coadsorbed toluene and methanol, Ts-p, transition state; within brackets protonated xylene, Pr-p adsorbed products xylene and water. (B) Dispersive energy corrected reaction energy diagram of the subsequent elementary reaction steps of the alkylation reaction of toluene with methanol catalyzed by acidic Mordenite. A comparison is made of reaction paths that lead to the formation of p-xylene, m-xylene, or o-xylene respectively (all values in kJ/mol). The values correspond to energies at 0 K. Reproduced with permission from Vos, A. M.; Rozanska, X.; Schoonheydt, R. A.; van Santen, R. A.; Hutschka, F.; Hafner, J. A Theoretical Study of the Alkylation Reaction of Toluene with Methanol Catalyzed by Acidic Mordenite. J. Am. Chem. Soc. 2001, 123(12), 2799–2809, https://doi.org/10.1021/ja001981i.

Solid acid catalysis; Part II, catalytic chemistry of proton activation

303

Fig. 8 Simulated hydrocracking equilibrium reaction product distributions. Schematic representation of simulated normalized DMB/n-C6 yield ratios (y) for various zeolite structures at T 577 K and 3000 kPa are shown as a function of zeolite structure. The DMB/n-C6 ratios were normalized by setting the value for the FAU-type zeolite at one. The experimental ratios (red) were determined from n-C16 hydro conversion experiments,39 the calculated ratios were taken from simulated adsorption isotherms of 2,2-DMB/n-C6 (yellow) or from Henry coefficients (green). The numbers in parentheses are the average pore sizes [Å]. DMB dimethylbutane, n-C6 n-hexane. After Schenk, M.; Calero, S.; Maesen, T. L. M.; Van Benthem, L. L.; Verbeek, M. G.; Smit, B. Understanding Zeolite Catalysis: Inverse Shape Selectivity Revised. Angew. Chem. Int. Ed. 2002, 41, 2499–2502, https://doi. org/10.1002/1521-3773(20020715)41:14 3.0.CO;2-T.

6.13.4.2

Stereoselectivity

We have seen in the previous section how when molecule size increases molecular interaction with the zeolite wall changes from attractive to repulsive. The onset of steric constraints leads to stereo selective reactivity. In a reaction with different competing reaction channels that reaction path is selected that has the optimum match of size and shape of reaction intermediate. Here we will discuss an additional example of stereo-selectivity of the alkylation reaction where we consider explicitly the relation between shape of zeolite cavity and transition state. Complementary to the simulations of Fig. 5 it shows that pre-transition state configuration of reactant molecules determines the difference in apparent activation energies. It illustrates as indicated in Fig. 6 that there is no difference in activation energies when micro-pore coverage is high. The reaction selected is the stereo selective alkylation of methanol with toluene.33 In the liquid phase alkylation to ortho, or para position of toluene occurs with little preference. P-xylene. Important to polymer industry, can be selectively produced when this reaction is catalyzed by Mordenite. This is the zeolite structure that has tubular one dimensional micro channels. Fig. 7A illustrates the successive reaction steps of the reaction. Toluene and methanol have to adsorb in close vicinity. Then of the different relative orientations with respect to each other the methyl group of toluene has to become oriented para with respect to the methyl group of methanol. This can be considered the pre-transition complex. Then activation of methanol leads to addition of CH3þ to the toluene molecule. Fig. 7B gives the computed energy reaction diagram of the different stages of the reaction near the proton in the Mordenite micro pore. The differences in energy of the three relevant transition complexes are given. They have the shape of related product molecules. The linear para-xylene molecule fits well along the axis in this linear channel but the bent ortho-xylene molecule as well metaxylene molecule has a less favorable interaction due to misfit of its shape with that of the micro pore tube. The pre-transition state complex of the para complex is more stable by 16 kJ/mol. The difference in activation energies of the para and ortho methyl addition reactions with respect to the pre-transition states is only 1 kJ/mol. This is very different from classical organic chemical34 substituent addition theory. In the homogenous phase nucleophilic addition does not discriminate between addition in ortho versus para position, but disfavors the meta position. The difference in stereo-selectivity of this reaction has at its root the steric difference of van der Waal stabilization of reactant molecules before reaction.

6.13.4.3

Micro pore equilibration

So far we have limited the analysis to catalytic systems where a particular elementary reaction step is rate controlling. This is the case interchange of molecules between the exterior and interior of the zeolite remains fast. However when adsorbate coverage of zeolite micro channels becomes high molecular mobility will become highly limited. This may inhibit equilibration of reactant, reaction intermediates and products occluded in the micro pore with the exterior of the zeolite.

304

Solid acid catalysis; Part II, catalytic chemistry of proton activation

Fig. 9 (A) Stereoselectivity of methanol conversion catalysis. The methanol to olefin reaction is catalyzed by SAPO-34 chabasite structure; Methanol to gasoline is catalyzed by ZSM-5. Reaction temperature 530 K.41,42 (B) Schematic representation of the complex reaction network of the acid catalyzed methanol to alkene reaction. After Yang, M.; Fan, D.; Wei, Y.; Tian, P.; Liu, Z. Recent Progress in Methanol-to-Olefins (MTO) Catalysts. Adv. Mater. 2019, 31(50), 1–15, https://doi.org/10.1002/adma.201902181.

Hydrocracking and related hydro isomerization reactions of long chain hydrocarbons provide an example. In the hydro isomerization reactions,35 that we shortly discussed in section 6.13.3 n-alkane is isomerized into isoalkane. In the hydrocracking reactions bond cleavage of long chain molecules gives shorter alkanes.36 Different from alkane cracking these reaction occur at milder temperatures and are catalyzed by protonic zeolites activated with a small amount of a noble metal as Pt or Pd. These transition metals activate CeH and H2. Alkane molecules are readily converted into the corresponding alkene molecules. Reaction is executed in excess hydrogen, so that alkanedalkene equilibrium is established. The protons activate the alkene molecules and form intermediate carbenium ions. The carbenium ions undergo isomerization or CeC bond cleavage reaction, that yield alkene product molecules. Hydrogenation of alkene molecules is again catalyzed by the noble metal component of the catalyst to give the alkane products. The catalyst is designed such, that when operated at high H2/ alkane ratio a low alkene concentration is maintained so that deactivating alkene oligomerization reactions are suppressed. When for short alkanes and low pressure concentration and molecular concentration external of the zeolite are in equilibrium, a rate expression as Eq. (1a) applies, where [Cr] ¼ Keq $ [alkane]/[H2], since q is the alkene concentration. Keq is the equilibrium constant of the alkane- alkene, H2 equilibrium and in Eq. (1a–c) Kads is the adsorption constant of alkene adsorption. The order of reaction rate R in H2 concentration will be
1 when alkane pressure is low and ultimately decreases to zero when alkane concentration is high. The orders of reaction in alkane and H2 will be the same, but of opposite sign.21 In contrast when there is no equilibrium between concentrations internal or external of the zeolite in Eq. (1a) q becomes alkane concentration and Kads is the adsorption constant of alkane. In Eq. (1b) kr is to be multiplied with Kads/H2. In this case at high hydrocarbon pressure the reaction maintains the order of
1 in H2 pressure and the order of reaction in alkane changes from þ 1 to 0. An expression related to the latter expression is used in kinetic studies of high pressure hydrocracking of heavy molecules.37 In high pressure hydrocracking experiments of n-C16 catalyzed by different zeolites product distributions simulations show that they reflect local product equilibria in the micro-pore. Statistical mechanical grand–canonical equilibrium simulations of adsorption isotherms have been done using force fields deduced by fitting with experimental adsorption isotherms in siliceous zeolite

Solid acid catalysis; Part II, catalytic chemistry of proton activation

305

(A)

Reactant

(B)

Alkene’’

Alkene’

H+

- H+

Carbenium ion’’

Alkane’’’

C-C cleavage

Carbonium ion Cracking

- H+

Carbenium ion’

Carbenium ion’’’ Alkane’’’ Hydride transfer i-Alkane Alkene

- H2 Alkene

- H+

Additional alkenes

Alkene’’’’ Hydride transfer Additional propagation cycles: - Dialkenes - Ringclosure - Aromatics - Small alkanes

Oligomerisation Deactivation - Paring reactions

Fig. 10 (A) Comparison of stereo selectivity of the alkane cracking reaction to aromatics versus short alkenes and alkanes. (B) The two major catalytic reaction cycles of the catalytic cracking reaction. The initiation cycle proceeds through carbonium ion intermediates. Once they decompose to give a carbenium ion intermediate, the propagation cycle takes over. This cycle is maintained by hydride transfer reactions illustrated in Fig. 11. Reproduced with permission from Catlow, C. R. A.; Van Speybroeck, V.; Van Santen, R. A. In: Modelling and Simulation in the Science of Micro- and Meso-Porous Materials. Elsevier, 2017, Chapter 5.

models replace.38 This is illustrated in Fig. 8. It shows very good agreement with experiment. Computed dimethyl butane/hexane ratios are compared with measured product ratios from the C16 cracking experiment. This agreement is only found when interaction between adsorbed intermediates is explicitly taken into account. In a micro pore of low occupation, where molecules do not interact Henry coefficient differences between n- and i-alkanes do not essentially differ. The interaction energies of both molecules with the zeolite wall are comparable. Then cavities do not induce shape selectivity. Kinetics then dominates selectivity. The origin of the difference in adsorption equilibria originates from non-ideal mixing behavior at high concentration of hexane. It is an entropic effect. According to the authors of40: “When pores become occupied with high concentration of adsorbate more small molecules can occupy the same space than larger molecules. This is important when we compare the linear n-hexane molecules with the shorter branched C6 isomers. In a small pore linear molecules become stretched and hence have a larger effective size than the more bulky branched molecules that have a smaller diameter. The smaller isomer molecules will push the linear hexane molecules from the pores. This will enhance their relative concentration and favor selectivity of reaction towards formation of the branched products. The effect will decrease with increasing micro pore size. An optimal selectivity is found for micro pores of intermediate size.”

306

Solid acid catalysis; Part II, catalytic chemistry of proton activation

A1 H2 C H3C

H2 C C H2

H2 C

H+

CH3

H3C

C H2

H2 C

CH3 C H2

+ CH2 H

H C

Carbonium ion

H3C

H2 C

+

CH3

H3C

-H+

CH3

Cracking

CH3

H2C

CH3 C H2

H2C

+H-

C H2

Primary carbenium ion

H3C

+alkane

H2 C

CH3

CH3

H3C

C H2

+ carbenium ion

Hydride transfer

C-C bond cleavage A2

H2 C

H3C

H

H2 C C H2

CH3

H

+

H+

C H2

H2 C

C H

H3C

CH3

C H2

C H2

Carbonium ion

H2

H2 C H3C

+

-H+

H2 C C H2

H2 C

+alkane

H3C

CH3

H3C

CH

H3C

C C H2

C H2

C H2

H2 C

CH3

H3C

C H2

HH2 C

C H2

C H2

Carbenium ion

H2 C

H3C CH3

Hydride transfer

C H2

C H2

CH3

+ carbenium ion

C H2

H2 C C H2

Alkane

H C C

H C CH3

H3C

H C CH3

CH3

CH3 H2 C

CH3 C H2

H C C

H2C

-H+

CH3

H-

H C

H3C

H2 C

B2

CH3 CH3

+ H2

Hydride transfer

C-H bond cleavage B1

CH3 C H2

Dehydrogenation +H-

Secondary carbenium ion

C H2

H2C

CH3

C H2

H2 C

H C

H3C

CH3 C H2

Heterolytic dehydrogenation

Fig. 11 Elementary reaction steps of catalytic cracking reaction. (A) CeC (a1) and CH (a2) bond cleavage reaction paths. Initially an unstable carbonium ion is formed. This decomposed rapidly into a carbenium ion. Upon deprotonation the carbenium ion leads to an alkene product molecule. This is counteracted by reaction of the carbenium ion with a molecules that transfer a hydride ion to the carbenium ion that hence reacts to give alkane. (B) The hydride transfer reaction. b1 transfer between alkane and carbenium ion, the reacting alkane molecule becomes another carbenium alkyl cation; b2 transfer between alkene and carbenium ion. Hydride transfer by an alkene combined with deprotonation leads to dehydrogenation of the alkene molecule. In both cases the initial carbenium ion converts to an alkane molecule.

6.13.5

Conclusion

Solid acid catalysts are needed to catalyze reactions that for thermodynamic reasons require temperatures where liquid acids cannot be used. Hydrocarbon conversion reactions that involve CeC or CeH bond cleavage as well as alcohol dehydration reactions belong to this category. The unique additional benefit of zeolite catalysts is their stereo selectivity. A major issue intrinsic to solid acid catalysis is catalyst stability. This is because usually in the course of reaction deactivating alkene molecules are formed. They can be undesirable coproducts, but are often also essential intermediates or desirable products. Alkene molecules cause catalyst deactivation by consecutive oligomerization and ring closure reactions that lead to non-reactive carbonaceous residue that blocks reactive sites. Choice of proton affinity and zeolite micro pore shape and dimensions optimizes catalyst selectivity and maximizes catalyst life. Rational design requires insight in de the mechanism of the reaction. Theoretical catalysis and simulations have become a useful tool to relate inorganic catalyst material properties with chemical reactivity.

Solid acid catalysis; Part II, catalytic chemistry of proton activation

307

Hydride transfer versus oligomerisation (alkene addition) CH3

CH3 H3C

H3C

CH C H2

C C H2

CH3

CH3

HH C

H3C C H2

H C

CH3

CH3

H2C

Alkene addition

CH3

-H+ CH H2C H2 C

H3C C H2

H C

H3C CH3

Alkane Hydride transfer product

C H

H3C CH3

CH C H2

CH3

Alkene

Fig. 12 Hydride transfer versus oligomerization. CeC bond formation between carbenium ion and alkene competes with hydride transfer reaction that reacts carbenium ion to alkane.

This we will illustrate with a short discussion of the methanol conversion reaction to hydrocarbons and the alkane catalytic cracking reaction. In Fig. 9A shape selectivity of the methanol dehydration reactions is illustrated. As we shortly discussed earlier in Section 6.13.4.1 when catalyzed by ZSM-5, that has a 10 ring three dimensional channel system, it will give as major product aromatics. This is the methanol to gasoline (MTG) process.41 Ethylene or propylene will be major products when catalyzed by the SAPO34. This AlPO4 zeolite structure related material is a polymorph of the chabasite structure. It has relatively large cavities, that can accommodate the bulky reactions intermediates of the reaction (see Fig. 5). Products can only escape through narrow 8 rings that connect the micro cavities in this material. As illustrated in Fig. 9B the reaction starts with a slow induction period in which by an activated process that converts methanol the first molecular CeC bond is formed and short alkene is produced. Subsequent reaction with methanol leads to a complex series of methylation and alkene oligomerization reactions. At quasi-steady state a so-called Hydro Carbon Pool forms (HCP) is present that consists of bulky cationic aromatic intermediates, that act as organic catalysts.29 In a competition with deactivation reactions the HCP intermediates will release light weight olefins and can be regenerated by ongoing methylation reactions. Whereas in a narrow pore system only light weight olefins are product, in a large pore system the aromatic molecules that also derive from the HCP intermediates can also be released as products. Fig. 9B show schematically the succession of reaction intermediates of this reaction. Differences in micro pore size control the difference in stereo selectivity. Proper tuning of Proton Affinity is essential. The intermediate strength OH Proton affinity of the SAPO material is optimum. When proton affinity is too large, the cyclopentadienyl cation intermediates that are part of the Hydrocarbon pool (HCP, see middle of Fig. 9B) become too stable. This also causes catalyst deactivation since stable cationic deposits remove reactive protons from the system.43 As is illustrated in Fig. 10A another important reaction that is shape selective is the catalytic cracking reaction of alkanes. CeC bond activation and CeH bond cleavage reactions, that are part of its reaction mechanism we have met before in Sections 6.13.2 and 6.13.3. Fig. 10A illustrates schematically difference in product selectivity when catalysis by a wide pore zeolite structure (FAU), that has 12 ring cavity connections, is compared with catalysis by the more narrow pore (MFI) zeolite. The large micro pore faujasite structure, that is the basis of the commercially used material, will give mainly a mixture of aromatics and short alkanes. On the other hand ZSM-5 with smaller pores will give dominantly short alkanes and alkenes. At the high temperature of reaction of 850 K oligomerization of unsaturated intermediate hydrocarbons will lead rapidly to deactivating polyaromatic carbonaceous material. The size restriction of the smaller ZSM-5 micro-pores will give these materials an extended lifetime compared to the Faujasite structure. The reaction mechanism evolves differently for the two catalysts.44 As illustrated in Fig. 10B when catalysis, apart from catalyst deactivating reactions, is essentially without steric constraints the reaction consists of two reaction cycles. This is the case when reaction is catalyzed by faujasite structure zeolites. One distinguishes an initiation reaction cycle and propagation reaction cycle. Proton activation through carbonium ion intermediates that lead to carbenium ions define the initiation reaction cycle. In the propagation

308

Solid acid catalysis; Part II, catalytic chemistry of proton activation

cycle carbenium ions become regenerated through hydride transfer processes. Carbonium ion intermediates are not formed in the propagation reaction cycle. The elementary reaction steps that dominate in the two catalytic reaction cycles are given in Fig. 11. In Fig. 11A elementary reaction steps of the initiation reaction cycle are summarized. Carbonium ion intermediate formation is a monomolecular reaction. Decomposition of this activated cation leads to CeC bond cleavage as well as CeH bond cleavage with formation of a carbenium ion. Upon deprotonation alkenes will be formed and reaction continues by regeneration of the carbonium ions with incoming reactant molecules and back donated proton. This initiation reaction cycle is converted into the propagation reaction when instead of deprotonation the carbenium ion is converted into product alkane by a hydride transfer reaction with reactant alkane. This regenerates a carbenium ion and the catalytic reaction cycle propagates. The carbenium ions can undergo secondary reactions that leads to additional CeC bond cleavage, isomerization as well as dehydrogenation reactions. In contrast to monomolecular carbonium ion formation the hydride transfer reaction (see Fig. 11B) is a bimolecular reaction. This bimolecular reaction has to be fast compared to carbenium ion deprotonation in order for the propagation reaction cycle to dominate. As is also illustrated in Fig. 11B hydride transfer in combination with proton back donation is an important reaction to give higher olefins. This is an essential reaction step to give aromatics as product. The more narrow ZSM-5 pore compared to that of Faujasite will inhibit hydride transfer reactions. Hence only CeC and CH bond cleavage reactions of alkanes will happen through intermediate carbonium ion formation and in contrast to Faujasite no aromatics as products are formed. In ZSM-5 catalyst the products only derive from the initiation reaction cycle. This is the HaagDessau reaction catalytic cracking mechanism.19,45 Because now carbonium ion intermediate formation is essential the apparent activation energy is the higher than for the Faujasite catalyzed reaction. Catalyst deactivation occurs through alkene oligomerization reactions. This reaction is suppressed by hydride transfer reactions. As illustrated in Fig. 12, when hydride transfer is fast compared the carbenium ion deprotonation, the carbenium ion will be converted to non-reactive alkane. Otherwise it will deprotonate to give alkene or lead to oligomerization by reaction with another alkene. Deactivation will be suppressed as long as alkene formation does not occur and carbenium ions are the dominant reaction intermediates. In order for the hydride transfer reaction to be fast compared to the carbenium ion deprotonation, reaction has to occur on reaction sites of low proton affinity that favor proton donation to the alkene versus proton acceptance from the carbenium ion.46 The other reason why the cracking reaction is a low proton affinity demanding reaction is that initiation proceeds through carbonium ions that have low proton affinity transfer reaction. Reactions that are catalyzed at high temperature prefer wide pore zeolites as catalysts. At high temperature differences in adsorption equilibrium constants are dominated by adsorption entropy. The loss of entropy upon adsorption is less in the wide micro pores, that are able to maintain a finite micro-pore adsorbate concentration. Whereas stereo selectivity is usually caused by confinement effects and contributions due to differences in diffusion rates are minor, molecular sieving properties of zeolites due to small differences in diffusion rates can be exploited by use of zeolite crystallites of large size at conditions where micropores are highly occupied. An example is the disproportionation reaction of toluene to benzene and xylene catalyzed by ZSM-5.47 The above discussion of methanol and alkane activation illustrates the two consequences of confinement: steric inhibition and variation of cavity concentration. In the case of methanol activation reaction stereo-selectivity arises primarily because cavity connection radius controls transport. Larger radii connection allows for aromatics transport, the smaller only that of alkenes. Cavities of large volume are needed in order to allow form formation of aromatics that are the organic catalysts of this reaction. The difference in stereo selectivity of the catalytic cracking reaction is due to the larger cavity size requirement for the bimolecular hydride transfer reaction. Selectivity difference in this case is from a change in reaction mechanism between wide micro-pore zeolite and smaller micropore zeolite. The proton affinity requirements if the methanol conversion and alkane cracking reaction respectively are different. The methanol conversion reaction deactivates due formation of substituted cyclopentadienyl cations43 by hydride transfer reaction of the olefin oligomers. Their stability is such, that proton back donation will not happen when zeolite PA is too small. This reaction requires protons of intermediate PA. In contrast the catalytic cracking reaction requires high temperature because it is initiated by formation of intermediate carbonium ions. In this article we have highlighted the complex relationship of catalyst structure and compositions and catalyst functionality. Progress in zeolite catalysis science has made large strides over the past half century. Here and in the previous Chapter ., part I we have discussed the deepened understand that is currently available due to experimentation with well-defined materials, refined characterization methods and detailed mechanistic and kinetics studies. This provides a basis to the design of detailed catalyst and catalytic reaction models that have been widely explored with advanced computational simulation techniques.48 Kinetic simulation techniques are becoming available that enable to incorporate the full non linearity of the zeolite catalytic reaction cycle. Reactivity descriptors of solid acid catalysis can be considered well understood. This provides a useful base to future exploration of zeolite catalysis.

References 1. Bates, S. P.; Van Santen, R. A. The Molecular Basis of Zeolite Catalysis: A Review of Theoretical Simulations. Adv. Catal. 1998, 42, 1–114. https://doi.org/10.1016/S03600564(08)60627-6. 2. Catlow, C. R. A.; van Santen, R. A.; Smit, B. Computer Modelling of Microporous Materials, Elsevier, 2004.

Solid acid catalysis; Part II, catalytic chemistry of proton activation

309

3. Catlow, C. R. A.; Van Speybroeck, V.; van Santen, R. A. Modelling and Simulation in the Science of Micro- and Meso-Porous Materials, Elsevier, 2017. 4. Sarazen, M. L.; Iglesia, E. Stability of Bound Species During Alkene Reactions on Solid Acids. Proc. Natl. Acad. Sci. U. S. A. 2017, 114 (20), E3900–E3908. https://doi.org/ 10.1073/pnas.1619557114. 5. Boudart, M.; Djéga-Mariadassou, G. Kinetics of Heterogeneous Catalytic Reactions, Princeton University Press, 1984. 6. Sengar, A.; van Santen, R. A.; Kuipers, J. A. M. Deactivation Kinetics of the Catalytic Alkylation Reaction. ACS Catal. Jun. 2020, 6988–7006. https://doi.org/10.1021/ acscatal.0c00932. 7. Olah, G. A.; Surya Prakash, G. K.; Molnr, R.; Sommer, J. Superacid Chemistry, John Wiley & Sons, Inc.: Hoboken, NJ, 2009. 8. Brouwer, D. M.; Hogeveen, H. Electrophilic Substitutions at Alkanes and in Alkylcarbonium Ions. Prog. Phys. Org. Chem. 1972, 9, 179. 9. van Santen, R. A.; Neurock, M. Molecular Heterogenous Catalysis, Wiley-VCH, 2006. 10. Zheng, X.; Blowers, P. Reactivity of Alkanes on Zeolites: A Computational Study of Propane Conversion Reactions. J. Phys. Chem. A Dec. 2005, 109 (47), 10734–10741. https://doi.org/10.1021/jp054605z. 11. Gounder, R.; Iglesia, E. Catalytic Consequences of Spatial Constraints and Acid Site Location for Monomolecular Alkane Activation on Zeolites. J. Am. Chem. Soc. Feb. 2009, 131 (5), 1958–1971. https://doi.org/10.1021/ja808292c. 12. Collins, S. J.; O’Malley, P. J. The Mechanism of Alkane Activation over Zeolite Brønsted Acid Sites. A Density-Functional Study. Chem. Phys. Lett. Dec. 1995, 246 (6), 555– 561. https://doi.org/10.1016/0009-2614(95)01149-3. 13. Esteves, P. M.; Mota, C. J. A.; Ramírez-Solís, A.; Hernández-Lamoneda, R. Potential Energy Surface of the C3H9 þ Cations. Protonated Propane. J. Am. Chem. Soc. Apr. 1998, 120 (13), 3213–3219. https://doi.org/10.1021/ja973784y. 14. Teunissen, E. H.; Jansen, A. P. J.; van Santen, R. A.; Orlando, R.; Dovesi, R. Adsorption Energies of NH3 and NHþ4 in Zeolites Corrected for the Long-Range Electrostatic Potential of the Crystal. J. Chem. Phys. 1994, 101, 5865. https://doi.org/10.1063/1.467303. 15. van Santen, R. A.; Niemantsverdriet, J. W. Chemical Kinetics and Catalysis, Plenum Press, 1995. 16. Rozanska, X.; van Santen, R. A.; Demuth, T.; Hutschka, F.; Hafner, J. A Periodic DFT Study of Isobutene Chemisorption in Proton-Exchanged Zeolites: Dependence of Reactivity on the Zeolite Framework Structure. J. Phys. Chem. B Feb. 2003, 107 (6), 1309–1315. https://doi.org/10.1021/jp021646b. 17. Gates, B. C.; Katzer, J. R.; Schuit, G. C. A. Chemistry of Catalytic Processes, Wiley, 1979. 18. Tuma, C.; Sauer, J. Protonated Isobutene in Zeolites:Tert-Butyl Cation or Alkoxide? Angew. Chem. Int. Ed. Jul. 2005, 44 (30), 4769–4771 https://doi.org/10.1002/ anie.200501002. 19. Kotrel, S.; Knözinger, H.; Gates, B. C. C. The Haag–Dessau Mechanism of Protolytic Cracking of Alkanes. Microporous Mesoporous Mater. Apr. 2000, 35–36, 11–20. https:// doi.org/10.1016/S1387-1811(99)00204-8. 20. Narbeshuber, T. F. T. F.; Vinek, H.; Lercher, J. A. J. A. Monomolecular Conversion of Light Alkanes over H-ZSM-5. J. Catal. Dec. 1995, 157 (2), 388–395. https://doi.org/ 10.1006/jcat.1995.1304. 21. Van De Runstraat, A.; Van Grondelle, J.; Van Santen, R. A. Microkinetics Modeling of the Hydroisomerization of n-Hexane. Ind. Eng. Chem. Res. 1997, 36, 3116–3125. https:// doi.org/10.1021/ie960661y. 22. De Gauw, F. J. M. M. J. M. M.; Van Grondelle, J.; Van Santen, R. A.; Van Grondelle, J.; Van Santen, R. A. V. A.; Van Grondelle, J.; Van Santen, R. A. The Intrinsic Kinetics of nHexane Hydroisomerization Catalyzed by Platinum-Loaded Solid-Acid Catalysts. J. Catal. Mar. 2002, 206 (2), 295–304. https://doi.org/10.1006/jcat.2001.3479. 23. Hölderich, W.; Karge, H. G.; Weitkamp, J.; Pfeifer, H. Zeolites and Related Microporous Materials: State of the Art 1994; 1st edn.; vol. 84; Elsevier Sciencee, 1994. 24. De Wispelaere, K.; Vanduyfhuys, L.; Van Speybroeck, V. Entropy Contributions to Transition State Modeling. In Modelling and Simulation in the Science of Micro- and MesoPorous Materials, Elsevier, 2018; pp 189–223. 25. Sastre, G.; Corma, A. The Confinement Effect in Zeolites. J. Mol. Catal. A Chem. Jun. 2009, 305 (1–2), 3–7. https://doi.org/10.1016/j.molcata.2008.10.042. 26. Derouane, E. G.; Andre, J. M.; Lucas, A. A. Surface Curvature Effects in Physisorption and Catalysis by Microporous Solids and Molecular Sieves. J. Catal. Mar. 1988, 110 (1), 58–73. https://doi.org/10.1016/0021-9517(88)90297-7. 27. Kärcher, J. Sitzungsberichte der S¨achsischen Akademie der Wissenschaften zu Leipzig, Mathematisch-Naturwissenchaftliche Klasse, vol. Band 128 Heft 6, 2003. 28. Chang, C. D.; Silvestri, A. J. The Conversion of Methanol and Other O-Compounds to Hydrocarbons over Zeolite Catalysts. J. Catal. 1977, 47, 249–259. 29. Olsbye, U.; Svelle, S.; Lillerud, K. P.; Wei, Z. H.; Chen, Y. Y.; Li, J. F.; Wang, J. G.; Fan, W. B. The Formation and Degradation of Active Species during Methanol Conversion over Protonated Zeotype Catalysts. Chem. Soc. Rev. 2015, 44 (20), 7155–7176. https://doi.org/10.1039/c5cs00304k. 30. Yang, M.; Fan, D.; Wei, Y.; Tian, P.; Liu, Z. Recent Progress in Methanol-to-Olefins (MTO) Catalysts. Adv. Mater. 2019, 31 (50), 1–15. https://doi.org/10.1002/ adma.201902181. 31. Lesthaeghe, D.; De Sterck, B.; Van Speybroeck, V.; Marin, G. B.; Waroquier, M. Zeolite Shape-Selectivity in the Gem-Methylation of Aromatic Hydrocarbons. Angew. Chem. Int. Ed. 2007, 46 (8), 1311–1314. https://doi.org/10.1002/anie.200604309. 32. Li, J.; Wei, Y.; Chen, J.; Xu, S.; Tian, P.; Yang, X.; Li, B.; Wang, J.; Liu, Z. Cavity Controls the Selectivity: Insights of Confinement Effects on MTO Reaction. ACS Catal. 2015, 5 (2), 661–665. https://doi.org/10.1021/cs501669k. 33. Vos, A. M.; Rozanska, X.; Schoonheydt, R. A.; van Santen, R. A.; Hutschka, F.; Hafner, J. A Theoretical Study of the Alkylation Reaction of Toluene with Methanol Catalyzed by Acidic Mordenite. J. Am. Chem. Soc. 2001, 123 (12), 2799–2809. https://doi.org/10.1021/ja001981i. 34. Hammett, L. Physical Organic Chemistry. Reaction Rates, Equilibria and Mechanisms, McGraw-Hill, 1940. 35. Weitkamp, J. Isomerization of Long-Chain N-Alkanes on a Pt/CaY Zeolite Catalyst. Ind. Eng. Chem. Prod. Res. Dev. 1982 Dec, 21 (4), 550–558. https://doi.org/10.1021/ i300008a008. 36. Weitkamp, J. Catalytic Hydrocracking-Mechanisms and Versatility of the Process. ChemCatChem 2012, 4 (3), 292–306. https://doi.org/10.1002/cctc.201100315. 37. Steijns, M.; Froment, G. F. Hydroisomerization and Hydrocracking. 3. Kinetic Analysis of Rate Data for n-Decane and n-Dodecane. Ind. Eng. Chem. Prod. Res. Dev. Dec. 1981, 20 (4), 660–668. https://doi.org/10.1021/i300004a014. 38. Smit, B.; Maesen, T. L. M. Molecular Simulations of Zeolites: Adsorption, Diffusion, and Shape Selectivity. Chem. Rev. Oct. 2008, 108 (10), 4125–4184. https://doi.org/ 10.1021/cr8002642. 39. Santilli, D. S.; Harris, T. V.; Zones, S. I. Inverse Shape Selectivity in Molecular Sieves: Observations, Modelling, and Predictions. Microporous Mater. Sep. 1993, 1 (5), 329– 341. https://doi.org/10.1016/0927-6513(93)80049-Z. 40. Schenk, M.; Calero, S.; Maesen, T. L. M.; Van Benthem, L. L.; Verbeek, M. G.; Smit, B. Understanding Zeolite Catalysis: Inverse Shape Selectivity Revised. Angew. Chem. Int. Ed. 2002, 41, 2499–2502. https://doi.org/10.1002/1521-3773(20020715)41:143.0.CO;2-T. 41. Chang, C. D. MTG Revisited. Stud. Surf. Sci. Catal. Jan. 1991, 61 (C), 393–404. https://doi.org/10.1016/S0167-2991(08)60105-6. 42. Xu, S.; Zhi, Y.; Han, J.; Zhang, W.; Wu, X.; Sun, T.; Wei, Y.; Liu, Z. Advances in Catalysis for Methanol-to-Olefins Conversion. Adv. Catal. 2020 Mar, 61, 2017. 43. Marcus, D. M.; McLachlan, K. A.; Wildman, M. A.; Ehresmann, J. O.; Kletnieks, P. W.; Haw, J. F. Experimental Evidence from H/D Exchange Studies for the Failure of Direct C– C Coupling Mechanisms in the Methanol-to-Olefin Process Catalyzed by HSAPO-34. Angew. Chem. May 2006, 45 (19), 3205–3208. https://doi.org/10.1002/ ange.200504372. 44. Krannila, H.; Kranilla, H. Monomolecular and Bimolecular Mechanisms of Paraffin Cracking: N-Butane Cracking Catalyzed by HZSM-5. J. Catal. May 1992, 135 (1), 115–124. https://doi.org/10.1016/0021-9517(92)90273-K. 45. Haag, W. O. Catalysis by Zeolites - Science and Technology. In Studies in Surface Science and Catalysis; Hölderich, W., Karge, H. G., Weitkamp, J., Pfeifer, H., Eds., Elsevier, 1994; pp 1375–1394.

310

Solid acid catalysis; Part II, catalytic chemistry of proton activation

46. Liu, C.; Van Santen, R. A.; Poursaeidesfahani, A.; Vlugt, T. J. H. H.; Pidko, E. A.; Hensen, E. J. M. M. Hydride Transfer Versus Deprotonation Kinetics in the Isobutane–Propene Alkylation Reaction: A Computational Study. ACS Catal. Dec. 2017, 7 (12), 8613–8627. https://doi.org/10.1021/acscatal.7b02877. 47. Olson, D. H.; Haag, W. O. Structure-Selectivity Relationship in Xylene Isomerization and Selective Toluene Disproportionation. In Catalytic Materials: Relationship Between Structure and Reactivity; ACS Symposium Series, ACS Publication, 1984; pp 276–307. https://doi.org/10.1021/bk-1984-0248.ch014. 48. van Santen, R. A. Modern Heterogeneous Catalysis: An Introduction. Wiley VCH, 2017.

6.14 Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins Hao Zhang, Emiel J.M. Hensen, and Nikolay Kosinov, Laboratory of Inorganic Materials and Catalysis, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, Eindhoven, The Netherlands © 2023 Elsevier Ltd. All rights reserved.

6.14.1 6.14.2 6.14.2.1 6.14.2.1.1 6.14.2.1.2 6.14.2.1.3 6.14.2.2 6.14.2.2.1 6.14.2.2.2 6.14.2.3 6.14.2.3.1 6.14.2.3.2 6.14.3 6.14.3.1 6.14.3.2 6.14.3.3 6.14.3.4 6.14.3.5 6.14.4 References

Introduction Non-oxidative dehydroaromatization of methane Mo/ZSM-5 Preparation of Mo/ZSM-5 Induction period and catalyst deactivation Reaction mechanism Alternative catalysts Zeolite-based catalysts Non-zeolite based catalysts Zeolite modification Tuning zeolite acidity Constructing hierarchical and core-shell zeolite structures Non-oxidative dehydrodimerization of methane Fe©SiO2 Other metal oxide-based catalysts Metal phosphides and metal nitrides Zeolite-based catalysts Alternative catalysts Summary and outlook

311 313 313 313 315 315 316 316 317 317 317 318 318 318 321 322 323 323 323 324

Abstract Direct conversion of methane, the main component in natural gas, into value-added chemical building blocks remains an important target of the chemical industry due to the ever-increasing demand for aromatics and light olefins as well as the needs for feedstock diversification. Methane activation is challenging because of the strong CeH bonds in combination with its negligible electron affinity and low polarizability. In the last 30 years, a significant amount of study has been devoted to methane non-oxidative dehydroaromatization (MDA), involving the conversion of methane to mainly benzene and hydrogen. Recently, the possibility of methane dehydrodimerization (MDD) to olefines has also been proposed. This chapter focuses on inorganic heterogeneous catalysts for non-oxidative conversion of methane to aromatics and olefins. The main types of catalysts of MDA and MDD are reviewed together with the discussion of the proposed reaction mechanisms. We also sketch an outlook for this burgeoning research field after summarizing the insights from literature.

6.14.1

Introduction

The quest for an efficient process to directly convert methane, the main component of natural gas, into value-added chemicals is driven by the increasing demand for aromatics and olefins and low pricing of natural gas feedstock.1–4 Activating methane is challenging because of the strong CeH bonds in combination with its negligible electron affinity and low polarizability.1,5 Fig. 1 summarizes the main chemical routes for methane upgrading to other chemicals and fuels. In indirect routes, synthesis gas (syngas, a mixture of CO and H2) is first produced by reforming (e.g., steam reforming or similar processes) or gasification, followed by, for example, Fischer-Tropsch synthesis (FTS) or methanol synthesis. While in current industrial practice FTS is mainly used to produce liquid transportation fuels and some valuable chemicals, this technology can also be tuned to produce olefins, aromatics, and oxygenates.6 Methanol is mainly used as an intermediate chemical in the production of formaldehyde, but can also be converted to olefins and aromatics (i.e., methanol-to-olefins, methanol-to-gasoline, and methanol-to-aromatics).7 The multistep nature of these indirect methane conversion processes implies that they can only be carried out in a cost-effective manner at a very large scale.1 It would be desirable to develop medium- or even smallscale processes in which methane can be converted in a single step to valuable products, especially for monetization of natural gas in remote parts of the world where it cannot be economically transported to end use markets. Typically, a distinction is made between oxidative and non-oxidative routes in direct methane conversion. Oxidative coupling of methane (OCM) to C2 hydrocarbons (ethylene, ethane) has been actively investigated during the past four decades. A major

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00026-1

311

312

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

Fig. 1 Main processes of methane conversion to higher hydrocarbons and oxygenates. Red box indicates chemicals; orange and blue boxes indicate the commercialized and uncommercialized processes, respectively.

challenge is the over-oxidation of the desired olefinic products to CO and CO2, resulting in the well-known activity-selectivity tradeoff.8 Thermal (non-catalytic) pyrolysis of methane has also been used in industry to produce acetylene or carbon black products. However, the selectivity to ethylene is limited in this approach.9 The electric arc process to produce acetylene at very high temperature in a plasma gas from motor fuel can also be run with methane, yet it suffers from poor economics and competing soot formation.10 In the last three decades, a significant amount of research has been devoted to methane dehydroaromatization (MDA), involving the conversion of methane to mainly benzene and hydrogen. More recently, the possibility of methane dehydrodimerization (MDD) has also been studied. MDD is also referred to non-oxidative coupling of methane (NOCM). MDA and MDD represent routes to convert low-value methane to chemical building blocks such as benzene and ethylene in a single step. However, commercialization of these technologies is hampered among others by the availability of suitable catalysts that display high selectivity and stability for these non-oxidative methane conversion routes. Unlike oxidative routes, non-oxidative conversion of methane to value-added chemicals is limited by thermodynamics in a wide range of temperatures (Fig. 2). Therefore, to achieve reasonable conversion levels, these processes require temperatures above

Fig. 2 Gibbs free energies of methane conversion processes. Reproduced with permission from Ref. Schwach, P.; Pan, X.; Bao, X., Chem. Rev. 2017, 117 (13), 8497–8520. Copyright 2017 American Chemistry Society.

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

313

600
C.11 A major challenge in non-oxidative conversion of hydrocarbons is that graphitic carbon is thermodynamically preferred over all other products. The harsh reaction conditions and the formation of coke that can easily cover active sites make the development of stable catalysts cumbersome. Successful catalysts for MDA and MDD should display high rates of CeH bond activation and provide a high selectivity to light hydrocarbon products, thereby avoiding as much as possible the formation of heavy hydrocarbons including coke. This chapter will focus on inorganic heterogeneous catalysts for non-oxidative conversion of methane to aromatics and olefins. We will review the main types of catalysts under development of MDA and MDD. The proposed reaction mechanisms will also be discussed. After summarizing the main insights from the available literature, we will sketch an outlook for this burgeoning research field.

6.14.2

Non-oxidative dehydroaromatization of methane

The report of Wang et al. in 1993 spawned widespread interest in methane dehydroaromatization.12 The authors identified Mo/ ZSM-5 as the preferred MDA catalyst. Earlier, Bragin et al. had already reported about the use of zeolite catalysts for the coaromatization of ethane and methane in 1982 and methane aromatization in 1989.13,14 In the following section, we will review the main aspects of the preparation, structure and catalytic performance of Mo/ZSM-5 catalysts for methane conversion.

6.14.2.1

Mo/ZSM-5

Wang et al. demonstrated the promising performance of Mo-modified ZSM-5 zeolite, which remains the most studied MDA catalyst.1,4,15 At typical reaction temperatures (600–800
C), Mo/ZSM-5 can convert about 5–15% of methane feed with an aromatics selectivity of 60–90% in a broad range of space velocities.11 While other metals such as Fe, V, Re, W, and Cr also display some activity in combination with ZSM-5 zeolite, their performance is much lower in comparison with Mo/ZSM-5.12,16–20 As the pore size of the 10-membered ring (10MR) pores of ZSM-5 zeolite of  0.55 nm is close to the kinetic diameter of benzene ( 0.59 nm), ZSM-5 (MFI topology) exhibits suitable shape-selective properties during methane conversion, explaining the high yield of mono-aromatics. Consequently, other 10MR zeolites such as MCM-22 (MWW) and ZSM-11 (MEL) have been shown to exhibit good MDA performance in combination with Mo.21–23 However, rapid coking deactivation is a common problem for these catalysts, which hinders practical implementation of the process.24 The coke that is believed to be a mixture of polyaromatics blocks both the zeolite pores and the active Mo sites. Many studies have been devoted to understanding the nature of the active Mo centers, the occurrence of an induction period, and, in general, the mechanism of MDA reaction over Mo/ZSM-5 including the mechanism of deactivation.21,22,25

6.14.2.1.1

Preparation of Mo/ZSM-5

Mo/ZSM-5 catalysts are typically prepared by impregnation or solid-state ion exchange methods using ammonium heptamolybdate or molybdenum trioxide as the Mo precursors. A heat treatment is generally required to induce the migration of Mo into the zeolite micropores. The structure of the active Mo-sites in Mo/ZSM-5 has been a topic of interest for several decades.26–30 The structure of the Mo(VI)-oxo species in as-prepared catalysts has also attracted widespread attention, because they are the precursors that are reduced and carburized and form Mo-sites, which can activate CeH bonds in methane. The structure of these precursor species and their role in the MDA reaction are not fully understood yet. Mo2O52þ dimers have been proposed to be the precursors of the active sites for methane activation (Fig. 3).31–33 Li et al. suggested the presence of intrazeolitic [Mo5O12]6þ species upon solid state ion exchange of MoO3 with HZSM-5.34 Gao et al. claimed that MoO22þ centers can also be stabilized through two adjacent cationic exchange site and, therefore, might also be a precursor of the catalytically active centers. Zhou and co-workers proposed that MoO22þ can be localized on one cation exchange site.35–37 Ma et al. used EPR to evidence the presence of two types of Mo species in as-prepared Mo/ZSM-5 catalysts, namely (i) polynuclear Mo-oxo species located on the external surface of zeolite crystals, either in the form of octahedral MoO3 clusters or MoOx with a square pyramidal structure, and (ii) Mo-oxo species dispersed inside the zeolite pores and associated with framework Al atoms.38 These authors observed a linear correlation between the rate of aromatics formation and the amount of exchanged Mo species inside zeolite pores. This finding underpins the importance of intrazeolitic Mooxo species as the precursor to the active sites for the MDA reaction. The Si/Al ratio of the zeolite framework will also influence the structure of the Mo-oxo precursors dispersed inside the micropores. Previous studies indicate that MoO22þ species are dominant in

Fig. 3

Candidate structures of cationic mononuclear and dinuclear Mo-oxo complexes in zeolites.

314

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

samples with a low Si/Al ratio (a high number of Brønsted acid sites), whereas Mo2O52þ dimers can be found in larger numbers in catalysts with higher Si/Al ratios.1,39,40 Recently, Liu et al. directly observed isolated Mo atoms in ZSM-5 channels by using integrated differential phase-contrast scanning transmission electron microscopy (iDPC-STEM).41 Enlarged images of three representative areas are shown in Fig. 4A–D. The line scanning profiles in Fig. 4E indicate the presence of single-atom Mo centers anchored to different T-sites of the zeolite framework. Framework Al plays a key role in driving the formation of atomically dispersed Mo in zeolite micropores. Statistical analysis indicates that Al most preferentially occupies T1 sites ( 40%), followed by T2 and T5 sites, as shown in Fig. 4F. Although this microscopic study represents a direct evidence of the presence of isolated Mo-oxo species inside the zeolite pores, microscopy typically provides local information, which may not always be representative for the whole sample. Other techniques such as electron energy loss spectroscopy combined with STEM can further consolidate these microscopic findings and provide more information related to the geometric and electronic structures of Mo in Mo/ZSM-5 catalysts. As discussed above, the structure of the Mo-oxo species will be influenced by factors such as the Mo loading, the Si/Al ratio, and the catalyst preparation method. Another aspect worth mentioning is the nature of the supporting zeolite. The typically used commercial HZSM-5 samples are not homogeneous with respect to particle size and morphology. To fully understand the actual structure of Mo precursors, model systems with a better-defined zeolite morphology, Si/Al ratio, and Mo speciation should be used in future studies.

Fig. 4 (A) iDPC-STEM image of Mo/ZSM-5 catalyst. (B)–(D) Enlarged areas of 1 (empty channel), 2 (channels containing a MoO3H cluster bonded at T8 site), and 3 (MoO3H sites bonded at T1 sites). Each panel includes the STEM image (top), calculated structural model (middle), and the simulated projected electrostatic potential (bottom); Si blue, Al green, H white, Mo pink, and O red. (E) Line scanning profiles of the intensities in image (B)–(D). (F) Statistical analysis of Al occupancy at different T sites. Reproduced with permission from Ref. Liu, L.; Wang, N.; Zhu, C.; Liu, X.; Zhu, Y.; Guo, P.; Alfilfil, L.; Dong, X.; Zhang, D.; Han, Y., Angew. Chem. Int. Ed. 2020, 59(2), 819–825. Copyright 2020 Wiley-VCH.

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins 6.14.2.1.2

315

Induction period and catalyst deactivation

In 1996, Wang et al. reported that Mo-oxo precursors are reduced to Mo-carbides, leading to the activation of the Mo components for methane activation.42 The initial gas-phase products evolving from the catalyst bed are H2O, CO, CO2, and H2.43 This period when the catalyst is activated towards benzene formation is usually called the induction period.44–49 The reduction of Mo has been characterized by various methods such as X-ray absorption spectroscopy (XAS), nuclear magnetic resonance (NMR), and X-ray photoelectron spectroscopy (XPS).18,37,50,51 In addition to formation of reduced Mo phases, the generation of inactive Al2(MoO4)3 through dealumination of zeolites during reaction has been revealed by EPR and NMR.52 Many studies have been conducted to probe the actual geometric and electronic structures of reduced MoOxCy species, but more efforts are needed to fully understand them. Early work by Iglesia and co-workers suggested the formation of small MoCx clusters (0.6–1 nm) during MDA reaction on the basis of extended X-ray absorption fine structure (EXAFS) analysis.53 Combining in situ XAS experiments with density functional theory (DFT) calculations, Zhang et al. demonstrated that the C/Mo ratio of the carbidic Mo phase is higher than unity. This ratio is higher than in typical Mo-carbides such as a-MoC, b-Mo2C, and d-MoC.54 Furthermore, Lezcano-Gonzalez et al. studied the evolution of Mo using a combination of high energy resolution fluorescence detection X-ray absorption near edge structure (HERFD-XANES) and X-ray emission spectroscopy (XES).55 Isolated Mo-oxo centers initially present after calcination can be converted to a metastable MoCxOy phase, followed by further carburization to MoC3 clusters and eventually Mo2C nanoparticles (Fig. 5). These transformations are thought to cause catalyst deactivation. MoCx species generated from Mo-oxo precursors inside zeolite pores and associated with Brønsted acid sites are more stable and active than those formed from MoOx particles, which are dominant on the external surface.56,57 Kosinov et al. studied the active Mo centers during the whole process of the MDA reaction. Isolated MoCx centers were present during reaction over Mo/ZSM-5 with a low Mo loading amount (1 wt%), whereas Mo2C particles generated on the external surface of Mo/ZSM-5 at higher loadings (2 wt% and 5 wt%). The authors also proposed that the partially reduced isolated Mo sites inside zeolite pores are the active sites for the MDA reaction.57 The large amount of coke generated during the MDA reaction makes the investigation of active sites difficult. Vollmer et al. used CO pretreatment to generate reduced Mo centers without coke formation, and reported that mono- and dimeric Mo (oxy)carbide centers are present in addition to Mo2C nanoparticles on the external surface.58 However, it is still unclear whether the Mo active species generated through CO treatment are the same as the Mo-centers formed during MDA reaction in methane. The relatively high reaction temperature (> 650
C) and the fast deactivation and accumulation of coke make it challenging to identify the structure of active Mo-centers during reaction. There is a significant body of work demonstrating that isolated MoCx centers or MoCx clusters inside zeolite pores, which are generated during the induction period, are the active sites for benzene formation, whereas larger MoCx particles on the external zeolite surface produce mainly coke. The exact nature of the intrazeolitic Mo species will depend on the Mo loading and the Si/Al ratio of ZSM-5, while other factors such as the zeolite morphology and Mo loading method and further pretreatment will also influence the Mo speciation.

6.14.2.1.3

Reaction mechanism

Currently, a bifunctional mechanism, involving active Mo centers and Brønsted acid sites, is most widely used to explain the reactivity of Mo/ZSM-5 catalysts.1,21,59,60 Mo centers activate the CeH bonds of methane to form CHx species, followed by the formation of ethylene or acetylene together with hydrogen. Ethylene and acetylene can serve as reaction intermediates that are converted to aromatics over the second catalytic function, i.e., Brønsted acid sites.61–66 Benzene is the main product, while other products are toluene, naphthalene, xylene, ethane, ethylene, and coke (Fig. 6). This mechanism is supported by the observations that methane activation over HZSM-5 is negligible and that the conversion of methane is enhanced by loading Mo species. Furthermore, there is a strong correlation between the benzene formation rate and the Brønsted acidity of the Mo/HZSM-5 catalyst.60,61,67 However, there are also a number of experimental observations that cannot be explained by the bifunctional mechanism. It has, for instance, been reported that Mo loaded on a non-acidic Silicalite-1 (same MFI topology as HZSM-5) exhibits a substantial activity in the MDA reaction, which suggests that MDA is an intrinsic ability of Mo species in zeolites with a proper pore topology.30 The bifunctional mechanism hinges on ethylene as a reaction intermediate in the conversion from methane to aromatics. Nevertheless, the catalytic performance of Mo/ZSM-5 towards aromatization of ethylene is different from that of methane under similar reaction conditions.68

Fig. 5 Evolution of Mo centers during MDA reaction determined by Mo K-edge HERFD-XANES/XES. Reproduced with permission from Ref. Lezcano-González, I.; Oord, R.; Rovezzi, M.; Glatzel, P.; Botchway, S. W.; Weckhuysen, B. M.; Beale, A. M. Angew. Chem. Int. Ed. 2016, 55(17), 5215–5219. Copyright 2016 Wiley-VCH.

316

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

Fig. 6 Mechanisms of MDA reaction over Mo/ZSM-5. Reproduced with permission from Ref. Kosinov, N.; Hensen, E. J. M. Adv. Mater. 2020, 32(44), 2002565. Copyright 2020 Wiley-VCH.

Based on isotopic labelling techniques, Kosinov et al. recently presented an alternative mechanism involving a specific role of hydrocarbon species retained in the micropores of Mo/ZSM-5.57 The Mo centers anchored in zeolite framework activate methane, producing reactive C2Hx or radical species, which react with polyaromatic hydrocarbon species occluded in the zeolite pores. This process shares some similarities with the well-understood methanol-to-hydrocarbons (MTH) reaction in which a pool of hydrocarbon intermediates occluded in the micropores act as reaction intermediates.7 It was moreover postulated that, in contrast to the carbocation chemistry for MTH reactions, MDA likely involves radicals. As such, this MDA mechanism places less emphasis on the role of the Brønsted acid sites for converting initially formed intermediates to aromatics (Fig. 6). Further studies are needed to elucidate the species and their evolution processes in the proposed hydrocarbon pool. The influence of the zeolite topology, the Si/Al ratio, and the active metal on the hydrocarbon pool also needs to be understood.

6.14.2.2

Alternative catalysts

Besides Mo/ZSM-5, other types of zeolites and non-zeolite based catalysts were also prepared with the purpose of gauging their performance in the MDA reaction. Although some are active in MDA process, these catalysts do not exhibit the promising performance reported for Mo/ZSM-5.

6.14.2.2.1

Zeolite-based catalysts

Shape selectivity is a key requirement for efficient MDA catalysts in order to suppress the undesired formation of polyaromatics. It has been firmly established that 10MR zeolites with pore sizes close to the dynamic diameter of benzene ( 0.59 nm) are preferred over zeolites with smaller or larger pores. Previous studies showed that 10MR Mo-zeolite combinations like Mo/ZSM-5, Mo/ZSM11, Mo/MCM-22, and Mo/MCM-49 exhibit better catalytic performance than 12MR and 8MR zeolites.30,60,69,70 When using 12MR zeolites such as X, Y, and MOR as (acidic) catalyst supports, only a small amount of ethylene was obtained without aromatics.69 12MR zeolites cannot provide shape selective environment for benzene production, but may generate polycyclic aromatics and lead to very rapid coking deactivation. As expected, mesoporous Mo/MCM-41 and Mo/SBA-15 also exhibited limited performance in MDA reaction.70 Moreover, the 8MR Mo-zeolite combinations Mo/SSZ-13, Mo/SAPO-34, and Mo/ERS-7 showed low MDA activity.30,69,70 It has been postulated that 8MR zeolites, which typically contain small apertures between larger cavities, does not allow benzene to leave the pore system, explaining the rapid coking deactivation.30,60 Because of their favorable pore sizes for the MDA reaction, we will further discuss the utilization of 10MR zeolites in the following section. Metal-loaded ZSM-5. Many alternative metals have been explored to replace Mo. Similar to the preparation of Mo/ZSM-5, incipient-wetness impregnation and solid-state ion-exchange methods followed by high-temperature activation treatment were used to induce metal migration into the zeolite pores.71 Among the first-row transition metals, Fe is the most studied alternative to Mo.72,73 Isolated Fe species, Fe2O3 and Fe3O4 particles were observed in fresh Fe/ZSM-5 catalysts.74,75 It has been established that Fe-oxo species are slowly reduced to FeOxCy-type species during the MDA reaction, which is similar to Mo in Mo/ZSM5.16,75 Further studies indicated that coking can be lowered by increasing the Fe dispersion.74 Denardin et al. investigated the effect of promoters for Fe/ZSM-5. The use of Cu resulted in a shorter induction period and higher benzene formation rate, which may be related to the promoting role of Cu in Fe reduction. Ca-modified Fe/ZSM-5 also exhibited a significantly lower coke selectivity and a higher aromatics yield.76 Lim et al. showed that Mn/ZSM-5 is active in the MDA reaction.77 They mentioned that isolated Mn3þ species, isolated Mn2þ ions, agglomerated MnO, and agglomerated Mn2O3 species are present in the fresh Mn/ZSM-5 catalysts, while MDA operation induces the formation of new reduced Mn species suggested to be the active sites. Co/ZSM-5 has also been studied.78 Isolated Co2þ ions at cation-exchange positions can activate CeH bonds in methane and lead to CHx intermediates.78 The isolated nature and strong binding to the zeolite framework may explain why Co2þ was not reduced during the MDA reaction. As Zn cations in zeolites can effectively activate methane,79,80 it is reasonable to expect the Zn/ZSM-5 is also active in the MDA reaction. In the work of Abdelsayed et al.,81 two types of Zn species were proposed, namely anchored stable [Zn(OH)]þ and

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

317

loosely bounded ZnO nanoparticles. The activity of Zn/ZSM-5 in the MDA reaction decreased rapidly due to the large amounts of coke deposited. In addition, V/ZSM-5 and Cr/ZSM-5 are also active MDA catalysts, but exhibit lower activity than Mo/ZSM-5.18 The second-row transition metals such as Zr, Nb, Ru, Rh, and Pd have been mainly investigated as promoters for Mo/ZSM-5 and Fe/ ZSM-5 catalysts rather than as active components themselves.82–85 Among the third-row transition metals, W/ZSM-5 has been well studied, albeit that it exhibits a lower activity than Mo/ZSM-5.18,86 XAS analysis by Iglesia and co-workers indicated the presence of isolated W species in fresh catalysts, while WCx clusters were generated during the MDA reaction.87 However, other studies using XPS and 13C NMR indicated that no carbides were formed in W/ZSM-5 during the MDA process. Instead, the role of reduced WOx phases as the active phase was proposed.44,88 Re/ZSM-5 has also been identified as a promising alternative to Mo, as it can achieve a similar methane conversion and comparable benzene formation rates. EXAFS and thermogravimetric studies suggest that isolated Re-oxo species in the fresh catalysts can be reduced to metallic Re during the MDA reaction.19,82 Due to a significantly lower activity, there are not many systematic studies of other metal-loaded ZSM-5 catalysts in MDA reaction. The reducibility and location of active metal sites can greatly influence the catalytic performance, which should be further investigated. Moreover, it would be also useful to compare Mo/ZSM-5 and other metal-loaded ZSM-5 catalysts to resolve the main reason for these differences and identify descriptors for the catalytic performance. Other 10MR zeolites. Zeolite MCM-22 (MWW topology) also contains 10MR channels and has therefore been investigated. Shu et al. explored the use of Mo/MCM-22 in the MDA reaction and found a comparable performance to Mo/ZSM-5 in terms of activity and selectivity.89 ITQ-2 is a related type of zeolite prepared by delamination of MCM-22. The catalytic performance of Mo/ITQ-22 was studied by Martinez et al.90 Mo/ITQ-2 with a Si/Al ratio of 15 showed the highest Brønsted acidity and also exhibited the highest aromatic yield in their study. The authors also reduced the surface acidity of ITQ-2 by using oxalic acid, which largely inhibited the formation of naphthalene and resulted in an enhanced benzene selectivity of 75%. MCM-49 is another example of a zeolite material with the MWW topology. A high benzene selectivity of 90% was obtained with a yield of 10% at 973 K using a 6 wt% Mo/MCM-49 catalyst. The catalytic activity was maintained for 150 h reaction through several air-regeneration processes.91 Although other than ZSM-5 10MR zeolites show promising catalytic performance in MDA reaction, replacing the relatively cheap and stable ZSM-5 zeolite does not seem particularly beneficial.

6.14.2.2.2

Non-zeolite based catalysts

Although zeolite-based catalysts have been extensively studied, other types of catalysts for MDA reaction have also been explored. For example, catalysts based on sulfated metal oxides and two-dimensional layered metal carbides (MXene) have been investigated. However, the selectivity, stability and activity of these catalysts are far from those presented by zeolite-based systems. Mo/sulfated hafnia and Mo/sulfated zirconia. Abedin et al. investigated the catalytic performance of Mo-loaded sulfated hafnia (SH) catalysts.92 A 5 wt% Mo-loaded SH exhibited a higher methane conversion of 9.5% in comparison with a methane conversion of 7.5% achieved by Mo/ZSM-5 under similar reaction conditions. Kanitkar et al. prepared and tested Mo-loaded sulfated zirconia (SZ) in the MDA reaction.93 It was observed that MoO3 converted to MoOxCy species, followed by further reduction to Mo2C. The Mo/SZ catalyst showed a methane conversion in the range of 5–20% at temperatures between 600 and 700
C, comparable to conversion levels achieved by Mo/ZSM-5. However, also in this case catalyst deactivation and the formation of relatively large amounts of naphthalene remain problems. V2CTx MXene. MXene is a family of 2D inorganic materials, which consists of few-atoms-thick layers of transition metal carbides, carbonitrides, or nitrides. The interlamellar space between well-defined layers of MXene can be tuned to be close to the kinetic diameter of benzene ( 0.59 nm), which may be favorable for the shape-selective MDA reaction. Thakur et al. studied the catalytic performance of layered two-dimensional vanadium carbide with an interlamellar space of  0.7 nm. The as-prepared catalysts showed a methane conversion of 11.8% with a 4.84% benzene yield at a temperature of 700
C.94 It is thought that the confinement of 2D MXene catalyst facilitates the oligomerization of C2 intermediates inside the layered structure. Despite the promising performance, the catalyst lacks stability as evident from the very rapid deactivation.

6.14.2.3

Zeolite modification

Besides developing new catalyst compositions, modification of existing catalysts can yield improved catalysts in MDA process. Among the reported strategies, tuning the acidity and morphology of zeolites appear more promising methods to improve MDA catalysts.

6.14.2.3.1

Tuning zeolite acidity

A common strategy to enhance lifetime of zeolite-containing catalysts in hydrocarbon conversion reaction is to decrease zeolite acidity. A pre-treatment method by heating in N2 flow was employed to induce a partial removal of tetrahedral Al from the zeolite framework in HZSM-5 support, leading to a decrease in the number of excess strong Brønsted acid sites that are responsible for coke formation.95 The Mo/HZSM-5 catalysts prepared using the dealuminated support exhibited an enhanced benzene yield and a longer durability in MDA reaction compared with the conventional Mo/HZSM-5. Similarly, Ma et al. used steam treatment of ZSM-5 to remove a part of framework Al, reducing the number and strength of the Brønsted acid sites.96 The resulting catalyst displayed a substantially higher benzene yield and a longer lifetime as compared to the parent catalyst. Song et al. used a hydrothermal post-synthesis method to modify the acidity of ZSM-5.97 By re-crystallization and extraction of framework Al, the zeolite structure was stabilized, which led to enhanced stability and selectivity towards aromatics. Kikuchi et al. synthesized a series of SiO2-modified

318

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

Mo/ZSM-5 by silanation.98 The addition of basic group-substituted bulky silyl compounds resulted in a decrease of the number of Brønsted acid sites on the external surface of ZSM-5, which caused an increase of the benzene selectivity related to a significant decrease of naphthalene and coke formation.

6.14.2.3.2

Constructing hierarchical and core-shell zeolite structures

To further improve the catalytic activity and stability, zeolites with hierarchical and core-shell structures have been tested in MDA reaction.25,99–101 Hu et al. used a hierarchically porous Mo/TNU-9 zeolite prepared by phenyltriethoxysilane as mesoporogen to improve the mass transfer of methane and products.102 By using a hierarchical structure, methane can easily diffuse inside zeolites instead of reacting on the outer regions and forming polycyclic aromatics, which can block zeolite pores. The unsaturated hydrocarbon products can also diffuse outside of zeolites faster, which helps to avoid further dehydrogenation to coke. Both benzene yield and catalyst stability were increased compared with the non-mesoporous TNU-9 counterpart. Dimethyloctadecyl [3-(trimethoxysilyl)propyl]ammonium chloride was used to synthesize hierarchical Mo/MCM-22 zeolite by Tempelman et al., which led to a higher benzene selectivity compared to the microporous Mo/MCM-22 reference.103 Wu et al. investigated the effects of hierarchy factor of zeolites by evaluating a series of Mo/ZSM-5 catalysts with varying micro /mesoporous MFI textures. The results indicate a volcano-type dependence between hierarchy factor and the aromatics production rate, which can facilitate the rational design and synthesis of more active zeolite catalysts.104 Covering HZSM-5 with a non-acidic shell can eliminate the Brønsted acid sites on the external surface.105,106 Zhu et al. synthesized a catalyst with hollow HZSM-5@Silicalite-1 zeolite capsule structure, which showed enhanced benzene yield and decreased the amount of coke formed in the MDA reaction.107 Such structuring suppressed external coke generation and thus enhanced the lifespan of catalysts. Although the modification of zeolite-based catalysts can result in enhanced catalytic performance, it cannot completely prevent deactivation through coking. As such, these methods do not allow yet to explore the potential of MDA catalysts for scaling up to industrial scale. In addition, some of the reported modification strategies are complicated by their multi-step approach or the use of expensive chemicals.

6.14.3

Non-oxidative dehydrodimerization of methane

As mentioned in the introduction, commercialization of OCM has been actively pursued, because ethylene is the world’s largestvolume and most energy-intensive commodity chemical.108 Although thermal cracking is versatile with respect to the feedstock (mainly, naphtha from crude oil and ethane from natural gas), there remains a need to further diversify the feedstock and debottleneck ethylene manufacturing opportunities. For completeness, we mention here that currently methanol is used as an intermediate platform to manufacture light olefins.109,110 Zeolites are used to catalyze the required CeC bond formation reactions. Such methanol conversion chemistry has been known since the 1980s. The commercial viability of the methanol-to-olefins processes hinges on cheap carbon sources such as coal or shale gas and, accordingly, most industrial plants are located in China. While methanol is obtained from synthesis gas, the direct conversion of synthesis gas to light olefins is also being investigated.111–113 The underlying chemistry is similar to Fischer-Tropsch synthesis, yet limiting chain growth to light olefins is challenging mainly because of the associated high methane selectivity. Insights into the catalytic chemistry underlying CH4 activation and CO hydrogenation led Koerts and Van Santen to propose a two-step process in which methane is first converted at elevated temperatures to CHx surface intermediates followed by CeC coupling at lower temperature in a hydrogen stream.114,115 The interest in high-temperature methane coupling is strongly linked to investigations about the promoting role of solid surfaces in methane pyrolysis.4 Carbon fibers and ThO2/SiO2 surface can improve the selectivity toC2 hydrocarbons during non-oxidative pyrolysis of methane at temperatures higher than 1000 K.116,117 Platinum-loaded sulfated zirconia catalysts were found to exhibit an ethylene selectivity of 90% at 773 K, although the methane conversion was limited to 0.23%.118 Research in this direction was spawned by a breakthrough reported in 2014 by the Bao group, who presented a FeÓSiO2 catalyst with a high ethylene yield from methane under non-oxidative conditions.119,120 This work led to the exploration of other catalysts for such a high-temperature MDD process. However, the trade-off relation between C2 hydrocarbons selectivity and methane conversion remains a challenging issue. So far, only a limited number of catalysts have been investigated (Fig. 7). To realize the MDD process in practice, it is necessary to develop highly active and stable catalysts. Designing novel efficient catalysts will require better understanding of the reaction mechanism. Herein, we will discuss recent literature on MDD with a focus on the exploration of catalysts, the optimization of the reaction conditions, and attempts to understand the underlying reaction mechanism.

6.14.3.1

Fe©SiO2

In the work of Bao et al., a FeÓSiO2 catalyst was obtained by fusing SiO2 and Fe2SiO4 (fayalite) at 1700
C under atmospheric pressure.119,120 The Fe content of the resulting material was low (0.5 wt%) and the surface area was lower than 1 m2 g 1. At 1700
C, fayalite is liquid, while quartz is converted to cristobalite and reaches a “soft” intermediate between the liquid and solid state, which is thought to promote the formation of isolated species in the silica network. Although acid leaching was conducted after hightemperature treatment at 1700
C, the fresh catalyst still contains nanometer-sized Fe2O3 and Fe3O4 particles ( 3–4 nm) on SiO2 support. These Fe2O3 and Fe3O4 particles were dispersed into single-atom Fe sites in the form of FeSiC2 after the activation

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

319

Fig. 7 Catalytic performance of catalysts for the high-temperature MDD reaction. Methane conversion and C2-hydrocarbon selectivity are shown in the figure. Data from Refs. 118, 119, 121–140.

in methane at 900
C during reaction (Fig. 8A). The presence of isolated Fe-sites was evidenced by HAADF-STEM and EXAFS analysis. The resulting catalyst can convert  8% of a methane feed into ethylene ( 45%) and benzene ( 30%) at 950
C (Fig. 8B). The main products reported by the Bao group are ethylene, benzene, and naphthalene. The absence or at least a very small amount of produced coke allows stable catalytic performance for at least 60 h at 1020
C (Fig. 8C). A longer stability test of 960 h did not lead to any obvious deactivation neither.1 It was speculated that the nature of Fe sites in the active phase hinders surface CeC coupling reactions and coke generation. In contrast, an Fe-loaded SiO2 catalyst containing 2–5 nm Fe oxide nanoparticles mainly produce coke ( 95%), which highlights the importance of single-atom Fe centers. Although the FeÓSiO2 catalyst exhibits a high activity for MDD, the process suffers from a relatively high selectivity to naphthalene, which has a low value in practice. The seminal work of the Bao group also provided important insights into the relevance of gas-phase reactions in addition to reactions on the surface. Methyl radicals were identified as plausible intermediates by using online vacuum ultraviolet soft photoionization molecular-beam mass spectrometry (VUV-SPI-MBMS), as shown in Fig. 8D. The influence of the flow rate was also investigated by this technique: an improved ethylene selectivity was obtained at higher flow rates together with a decrease of the benzene and naphthalene selectivity (Fig. 8E). Supported by DFT calculations, the authors proposed that the MDD involves the formation of methyl radicals on the catalyst surface, followed by a series of radical reactions in the gas phase (Fig. 8F). Two •CH3 radicals can

combine to C2H6, followed by C2H6 dehydrogenation into C2H4 and •H radicals. A •C2H3 radical generated from C2H4 can react with another C2H4 molecule, opening up a reaction channel involving dehydrogenation and cyclization reactions to benzene. Dehydrogenation by •H and further radical reactions can also yield naphthalene. Later, Li and co-workers performed a systematic DFT study to understand the formation mechanism of the active FeSiC2 phase.141 The authors proposed that the formation of FeSiC2@SiO2 proceeds via two steps: the removal of oxygen around Fe followed by the insertion of carbon atoms. Recently, they discussed a mechanism in which the activation of methane occurs at the single-atom Fe sites. The methyl moiety is transferred to an adjacent carbon atom of the catalyst where CeC coupling and hydrogen transfer take place generating ethylene. This mechanism is different from the gas-phase reaction mechanism for CeC bond formation proposed by Bao and co-workers.142 Another theoretical work by Kim et al. of the single-atom Fe catalyst showed that methyl radicals and hydrogen were the main reaction products at temperatures above 1027
C. At lower temperatures, CeC coupling reaction occurs on catalyst surface and acetylene production is dominant over ethylene production.143 Sensitivity analysis indicated that the enhancement of selectivity to C2-products by optimizing reaction conditions was limited. Although such theoretical studies provide valuable suggestions about the formation of the active phase and the MDD reaction mechanism, it remains very difficult to ascertain any of these findings by experiments, which is related to the harsh conditions of the MDD reaction. Employing a synthetic strategy of melt-fusing similar to Bao’s work, Han et al. prepared a Fe@cristobalite material.125 This catalyst exhibited

320 Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

Fig. 8 (A) High-angle annular dark field scanning transmission electron microscopy (HAADF-STEM) analysis of the used Fe©SiO2 catalyst for the MDD reaction with the inset showing the proposed structure of the active Fe sites in the FeSiC2 active phase. (B) Catalytic activity and (C) stability tests. VUV-SPI-MBMS at (D) 1193 K and (E) 1223 K at different flow rates. (F) Gas phase reaction pathways at 1225 K. Reproduced with permission from Ref. Guo, X.; Fang, G.; Li, G.; Ma, H.; Fan, H.; Yu, L.; Ma, C.; Wu, X.; Deng, D.; Wei, M.; Tan, D.; Si, R.; Zhang, S.; Li, J.; Sun, L.; Tang, Z.; Pan, X.; Bao, X. Science 2014, 344(6184), 616–619. Copyright 2014 American Association for the Advancement of Science.

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

321

a methane conversion of 6.5% with 51.3% C2 selectivity, 14.3% C3-C5 selectivity, 32.4% aromatics and 2.0% coke at 1020
C. The reported methane conversion is substantially lower than the one reported by Bao and co-workers under similar reaction conditions (methane conversion of 32%). These authors also observed the presence of FeCx nanoparticles after 10 h of reaction. There are still many questions to be answered regarding the FeÓSiO2 catalyst, for instance, related to the synthesis. It is important that the results are verified in order to further optimize catalyst synthesis towards optimum performance and to expand the characterization of the active phase and reaction mechanism. With respect to the mechanism, it is important to understand in detail the role of gas-phase radical chemistry. As coke is absent or at least not formed in such amounts that deactivation sets in immediately, it is interesting to understand the formation of benzene and naphthalene and the reasons why the larger polyaromatics are hardly observed. It is also clear that the reactor configuration can play an important role in this kind of high-temperature chemistry including aspects such as the catalyst bed void fraction.

6.14.3.2

Other metal oxide-based catalysts

In addition to the single-atom FeÓSiO2 catalyst presented by Bao and co-workers, other supported metal oxide catalysts have been evaluated. Catalysts based on Fe(II)/SiO2 and Pt/CeO2 were reported to display some activity in the MDD reaction.134 DFT calculations predicted that single-atom Ru/TiO2 might also be a favorable system.144 Other systems that have been explored are CuSO4/ Al2O3 and liquid active phase in In/SiO2.130 Sot et al. synthesized a series of Fe(II)/silica catalysts by a surface organometallic chemistry approach.140 The as-prepared catalyst contains isolated Fe(II) centers on SiO2 with a high surface area of 173 m2 g 1, exceeding by far the surface area of the FeÓSiO2 catalyst. The authors reported a 15–22% hydrocarbons selectivity at a methane conversion of 3–4% at 1000
C. Fe species reduced to Fe-carbide clusters during the MDD reaction. Nevertheless, deactivation of this catalyst was very fast ( 1 min). EPR analysis also indicated the presence of C-containing radicals. While typically a high surface area is beneficial for performance of heterogeneous catalysts, an important corollary of the MDD findings reported so far is that promising catalytic performance can be associated with a low surface area. This might have to do with the involvement of radical reactions in the gas phase. Single-atom catalysts that can limit CeC coupling reaction and, therefore, hinder coke formation are suggested to be promising materials for MDD. Xie et al. developed a series of single-atom Pt/CeO2 catalysts which were able to convert about 13% methane to C2 products with a selectivity of  80% at 975
C (Fig. 9A).127 No coke was observed under these conditions, enabling stable operation for 40 h (Fig. 9B). XAS analysis demonstrated that the Pt species were present as single atoms in the as-prepared catalyst (Fig. 9C). Further characterization confirmed that the ceria structure was maintained in the reductive methane atmosphere (Fig. 9D) and that the Pt single atoms did not sinter extensively (Fig. 9E and F). It should be noted that the relatively high methane conversion ( 13%) was at least in part due to the dilution (1 vol% methane), resulting in an overall low reaction rate. Unlike the

Fig. 9 (A) Catalytic performance of 0.5 wt% Pt1/CeO2 catalyst evaluated at various temperatures. (B) 40 h stability test of 0.5 wt% Pt/CeO2 catalyst at 975
C. (C) XAS analysis of fresh catalyst and the corresponding references. (D) X-ray diffraction (XRD) patterns of 0.5 wt% Pt/CeO2 catalyst before and after the stability test. HAADF-STEM images of (E) fresh and (F) used 0.5 wt% Pt/CeO2 catalyst. Reproduced with permission from Ref. Xie, P.; Pu, T.; Nie, A.; Hwang, S.; Purdy, S. C.; Yu, W.; Su, D.; Miller, J. T.; Wang, C. ACS Catal. 2018, 8(5), 4044–4048. Copyright 2018 American Chemistry Society.

322

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

gas-phase reaction pathway reported by Bao et al.,119 the authors claimed that the single-atom Pt sites can not only activate the CeH bonds of methane by dehydrogenation, but also catalyze CeC coupling towards C2 hydrocarbons. This may explain the different product composition. While Pt/CeO2 leads to ethane, ethylene, and acetylene, the FeÓSiO2 mainly produces ethylene.119 Ex-situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) suggested that Pt/CeO2 after methane activation at 900
C can stabilize adsorbed C2 species, acting as reaction intermediates on the catalyst surface. Bajec and co-workers also synthesized single-atom Pt/CeO2 catalysts using a similar method. Although a small amount of coke was generated, the catalyst was stable during a 245 h stability test.134 Recently, Wen et al. performed theoretical calculations to elucidate the reaction mechanism of methane conversion of Pt clusters supported on the CeO2(111) surface.145 Two reaction pathways were proposed, i.e., CH4* / C* þ 4H* and CHx* þ CHx* / C2Hx*. The highest activation energy of the CH4* / C* þ 4H* was predicted for a single-atom Pt1 catalyst, which also showed the highest C2 selectivity (93% ethylene and 6.8% acetylene) compared with Pt3 and Pt10 clusters. Although the catalytic performance of such single-atom Pt/CeO2 catalysts is much lower than that of the original FeÓSiO2 catalyst, these materials exhibit promising activity and stability in the MDD reaction. Ogihara et al. proposed an alternative strategy for methane valorization by combining thermal decomposition of methane and non-oxidative methane coupling.135 A Pt/SiO2 catalyst achieved  10% methane conversion at 700
C with a  0.12% C2 products yield. The main products under these conditions were hydrogen and carbon nanofibers. Unlike conventional Ni-, Fe-, and Co-based catalysts for complete methane dehydrogenation, the resulting Pd-Si bimetallic phase was suggested to promote methane coupling. Liquid indium metal catalysts have been discussed by Nishikawa et al.130 In/SiO2 was effective for methane conversion above 750
C with the production of a variety of higher hydrocarbons including ethane, acetylene, ethylene, propylene, benzene, toluene, and naphthalene. A total hydrocarbons selectivity of 96% was achieved at 825
C with a methane conversion of less than 1%. A following study of this catalyst investigated the structural evolution of In during the MDD reaction using XAS, uncovering a set of complex structural transformations.146 Some of these structural changes were verified by a combination of DFT calculations and first-principles molecular dynamics calculations.147 An important insight from these works was that In-CH3 species generated at the liquid In surface can be involved in CH3-CH3 coupling reaction via a conventional Langmuir-Hinshelwood mechanism. Moreover, it was suggested that the solubility of H species in bulk In helps to improve these coupling reactions by supressing CH4 formation.

6.14.3.3

Metal phosphides and metal nitrides

Dipu et al. developed a Ni-P/SiO2 catalyst that was evaluated in the MDD reaction.122 Among several Ni-containing catalysts screened for this purpose, Ni-P/SiO2 showed the best performance (Fig. 10A). The catalyst exhibits a methane conversion of  0.5% at 900
C with a  75% C2 selectivity. The products include ethylene, ethane, acetylene, propylene, benzene, toluene, and naphthalene. Bulk Ni2P also exhibited a noticeable activity in methane coupling, while the performance of pure SiO2 as a reference was found to be lower than Ni2P and Ni-P/SiO2 (Fig. 10B and C). The total hydrocarbon selectivity increased with decreasing reaction temperature and the C2 selectivity was as high as 99.9% at 1123 K (Fig. 10D). However, the methane conversion only reached 0.08% at this temperature. The same group further studied the active phase structure in Ni-P/SiO2 with different Ni/P ratios using EXAFS: the stable phase under the reaction conditions (12 h MDD, 1173 K) was found to be Ni2P.148 These studies demonstrate the potential applications of metal phosphides in this field. However, more insightful understandings related to mechanism and reaction pathways over such metal phosphide catalyst are needed. Dutta et al. screened a series of GaN/SBA-15 catalysts.133 Ethylene was the main product with a molar selectivity of 71% at 700
C. However, methane conversion was limited to 0.35%. Later, the same group developed several methods for the regeneration of such GaN-based catalysts.132 When the catalysts were regenerated in air at 500–550
C, ethylene yield would decrease by 16% after 5 cycles, which is likely caused by the oxidation of GaN to Ga2O3. However, the authors found that the ethylene yield could be

Fig. 10 (A) Hydrocarbons yield of Ni-M/SiO2 catalysts at 1173 K. (B) Hydrocarbons yield and (C) methane conversion of N2P, Ni-P/SiO2, and SiO2 at 1173 K. (D) Product distribution and methane conversion of Ni-P/SiO2 catalyst. Reproduced with permission from Ref. Dipu, A. L.; Ohbuchi, S.; Nishikawa, Y.; Iguchi, S.; Ogihara, H.; Yamanaka, I. ACS Catal. 2020, 10(1), 375–379. Copyright 2020 American Chemistry Society.

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

323

maintained over multiple cycles after an intermediate re-nitridation step following air regeneration. DFT calculations suggested a mechanism in which CH3* and H* fragments are obtained by CH4 dissociation on Ga and N sites, respectively.149 The ratedetermining step is the cleavage of the CeH bond in CH3* species on Ga. C2H4 would be formed from two CH2* species adsorbed on adjacent Ga-N sites. In these studies, a surface CeC coupling pathway on GaN has been proposed through theoretical calculations. Such a surface reaction mechanism is reasonable given the relatively low reaction temperature. Nevertheless, radical reactions in the gas phase should not be excluded.

6.14.3.4

Zeolite-based catalysts

Zeolite-based catalysts, usually operated in MDA at  700
C, have also been evaluated for the MDD reaction. Active species can be confined in the pores of zeolites to achieve a highly dispersed state, which may prevent sintering under the harsh MDD reaction conditions. A Mo/[B]ZSM-5 catalyst reported by Shen et al. reached  90% C2H4 selectivity at 650
C.129 Boron was employed instead of aluminium in zeolite synthesis in order to decrease the strength of the zeolite acidity. Mo-oxo precursors were reduced during reaction and formed a MoCx phase for methane activation. No obvious deactivation could be observed in an 18 h stability test. However, the methane conversion was very low (0.8%), even when diluted (5 vol%) methane was used. Gerceker et al. investigated a series of PtSn/ZSM-5 catalysts and their MDD performance.131 Methane conversion with the best performing catalyst was only 0.06% at 700
C. Xiao et al. studied the Pt/Bi ratio in PtBi/ZSM-5 catalysts.128 The authors reported that the surface Pt species serve as the active sites for methane activation, while Bi promotes C2 product formation. The C2 selectivity was higher than 80%, and no obvious decrease of methane conversion and C2 selectivity could be observed during an 8 h stability test at 650
C. The methane conversion reached 2% although diluted methane (10 vol%) was used. Later, the same group studied the active site structure by using in situ XRD and in situ XAS, showing the formation of a cubic intermetallic Pt3Bi shell on a metallic Pt core.150 The authors demonstrated that a low Bi loading was effective for MDD, whereas high Bi loading deactivated the catalytic surface and resulted in activity loss. Moreover, Miao et al. investigated the influence of Agþ ions in ZSM-5 on methane coupling.151 DFT calculations suggested that isolated Ag species in ZSM-5 play a key role in methane activation and subsequent non-oxidative CeC coupling.152 To summarize, the majority of MDD studies with zeolite-based catalysts focused on ZSM-5. Low reaction temperature ( 700
C) was usually used with zeolite catalysts aiming at a low rate of coking deactivation. However, due to the low reaction temperature methane conversion levels remain low.

6.14.3.5

Alternative catalysts

Pb- and Fe-doped albite catalysts showed promising activity in the MDD reaction at 873–1073 K.137 Albite is a natural mineral with the chemical formula of A[T4O8], where T represents Si and Al tetrahedrally coordinated by oxygen and connected in such way that 4- and 8-membered ring channels are formed. Pb was introduced to replace alkali and alkaline-earth metal ions in albite. The starting albite material showed a methane conversion of 3.32%, whereas a 3.4 wt% Pb/albite exhibited an enhanced methane conversion of 8.19% with a C2-hydrocarbon selectivity of  99% at 1073 K. Employing a high-temperature shockwave method, Yao et al. synthesized a single-atom Pt/C3N4 catalyst with periodic on-off heating (55 ms on, 550 ms off).153 The high synthesis temperature (1500–2000 K) resulted in the dispersion of Pt species into single sites on the support. In the MDD reaction, the selectivity to ethylene, ethane, and benzene of the resulting Pt/C3N4 catalyst was higher than 90%, while no coke was generated in a 50 h stability test at 973 K. These catalysts display promising activity in the MDD reaction, suggesting that novel materials and synthesis strategies can provide opportunities in enhancing the catalytic performance of MDD catalysts.

6.14.4

Summary and outlook

Non-oxidative conversion of methane remains high on the list of reactions to be implemented in the chemical industry aiming at a direct route to chemicals for natural gas monetization. However, the limited activity and lifetime of most catalysts represent major hurdles in the further development of catalysts and the related chemical processes. Two main routes of non-oxidative conversion of methane are MDA and MDD. The MDA reaction has already been intensively studied in the last three decades. Various catalysts, reaction mechanisms, and process intensification strategies have been proposed. Especially, advanced characterization methods have increased our understanding of the active sites in metal (Mo)-modified zeolite catalysts and the associated mechanisms for MDA. Compared with MDA, research on the MDD reaction is still in its infancy. Although some catalysts and reaction strategies have been explored, the catalytic performance is poor. Complications arise from a stronger contribution of gas-phase radical reactions and conditions that give rise to rapid coking deactivation. Nevertheless, some promising leads have been identified. The following main aspects deserve attention in order to move this exciting field forward: Catalyst development. Challenges ahead are mainly in the materials science field, i.e., identifying materials with preferably isolated reaction centers for methane activation followed by radical CeC bond formation in a porous matrix conducive to rapid transport of the products away from the reaction zone. In this respect, the synthesis of tailored single-atom catalysts may be a worthwhile approach. For example, single-atom Pt catalysts prepared by the shockwave high-temperature synthesis technique exhibit promising activity and stability in MDD. A supercritical solvothermal synthesis method was shown to enhance the dispersion of active Mo sites in HZSM-5.153,154 Newly developed 2D MXene materials and sulfated hafnia also exhibit promising activity in MDA.92,94 Besides

324

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

thermal catalysis, photocatalytic and plasma-assisted processes display promising catalytic performance in methane non-oxidative conversion.121,155 Rapid screening methods might be useful for identifying novel catalysts compositions for MDA and especially MDD, although the challenging reaction conditions imply high development cost. Theoretical calculations can also guide the design of improved catalysts, especially with respect to the activation of methane and the role of gas-phase radical reactions. Finally, the combination of theoretical calculations with detailed spectroscopic characterization will also speed up the development of novel catalytic materials. Reaction mechanism. By understanding reaction mechanism, educated guesses can be made about catalyst design. Hightemperature in situ/operando characterization techniques such as DRIFTS, XPS, XRD, and XAS are required to fully understand the mechanism under the actual reaction conditions. Studying the gas-phase species present is crucial to understand these MDA/ MDD processes as radical species are likely playing an important role. The first studies have been performed providing insight in the actual reaction pathways in MDD and the hydrocarbon pool species in MDA. VUV-PIMS, planar laser induced fluorescence, and EPR may help to identify the reaction intermediates and shed light on the complicated reaction mechanisms. X-ray free electron laser-based techniques might also provide new opportunities to detect “live” intermediates during the reaction. Besides advanced characterization, well-defined model catalysts with homogeneous active sites should be used in mechanistic studies, single-atom catalysts forming a good starting point. Transient reaction techniques such as temporal analysis of products combined with isotope labelling experiments (e.g., steady-state isotopic transient kinetic analysis, SSITKA) can provide new insights in MDA and MDD mechanisms as well. Economic analysis. Detailed economic analysis of promising catalyst systems and processes should be carried out to assess whether a potential process is practical and economically attractive. While catalyst development and production cost may play a role, process concepts that can deal with the high reaction temperature (including bringing in the heat to drive the endothermic reactions) and the likely need to regenerate the catalyst frequently are very much required. To summarize, major efforts are still needed to make a next step in the commercialization of MDA and MDD processes. Based on the state of the art, there is an ample room for fundamental studies into the mechanisms of the unusual high-temperature reactions, guiding the design and fabrication of novel catalytic materials. Ideally, these studies should be performed in a concerted manner with process development efforts.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.

Schwach, P.; Pan, X.; Bao, X. Chem. Rev. 2017, 117 (13), 8497–8520. Olivos-Suarez, A. I.; Szécsényi, À.; Hensen, E. J. M.; Ruiz-Martinez, J.; Pidko, E. A.; Gascon, J. ACS Catal. 2016, 6 (5), 2965–2981. Taifan, W.; Baltrusaitis, J. Appl. Catal. B Environ. 2016, 198, 525–547. Choudhary, T. V.; Aksoylu, E.; Wayne Goodman, D. Catal. Rev. 2003, 45 (1), 151–203. Meng, X.; Cui, X.; Rajan, N. P.; Yu, L.; Deng, D.; Bao, X. Chem 2019, 5 (9), 2296–2325. Luk, H. T.; Mondelli, C.; Ferré, D. C.; Stewart, J. A.; Pérez-Ramírez, J. Chem. Soc. Rev. 2017, 46 (5), 1358–1426. Olsbye, U.; Svelle, S.; Bjørgen, M.; Beato, P.; Janssens, T. V. W.; Joensen, F.; Bordiga, S.; Lillerud, K. P. Angew. Chem. Int. Ed. Engl. 2012, 51 (24), 5810–5831. Horn, R.; Schlögl, R. Catal. Lett. 2015, 145 (1), 23–39. Huang, K.; Miller, J. B.; Huber, G. W.; Dumesic, J. A.; Maravelias, C. T. Joule 2018, 2 (2), 349–365. Fincke, J. R.; Anderson, R. P.; Hyde, T.; Detering, B. A.; Wright, R.; Bewley, R. L.; Haggard, D. C.; Swank, W. D. Plasma Chem. Plasma Process. 2002, 22 (1), 105–136. Kosinov, N.; Hensen, E. J. M. Adv. Mater. 2020, 32 (44), 2002565. Wang, L.; Tao, L.; Xie, M.; Xu, G.; Huang, J.; Xu, Y. Catal. Lett. 1993, 21 (1), 35–41. Bragin, O. V.; Vasina, T. V.; Preobrazhenskii, A. V.; Minachev, K. M. Biol. Bull. Acad. Sci. USSR 1989, 38 (3), 680. Bragin, O. V.; Vasina, T. V.; Isakov, Y. I.; Nefedov, B. K.; Preobrazhenskii, A. V.; Palishkina, N. V.; Minachev, K. M. Biol. Bull. Acad. Sci. USSR 1982, 31 (4), 847. Tang, P.; Zhu, Q.; Wu, Z.; Ma, D. Energy Environ. Sci. 2014, 7 (8), 2580–2591. Weckhuysen, B. M.; Wang, D.; Rosynek, M. P.; Lunsford, J. H. Angew. Chem. Int. Ed. Engl. 1997, 36 (21), 2374–2376. Gao, J.; Zheng, Y.; Tang, Y.; Jehng, J.-M.; Grybos, R.; Handzlik, J.; Wachs, I. E.; Podkolzin, S. G. ACS Catal. 2015, 5 (5), 3078–3092. Weckhuysen, B. M.; Wang, D.; Rosynek, M. P.; Lunsford, J. H. J. Catal. 1998, 175 (2), 338–346. Wang, L.; Ohnishi, R.; Ichikawa, M. J. Catal. 2000, 190 (2), 276–283. Wang, L.; Xu, Y.; Wong, S.-T.; Cui, W.; Guo, X. Appl. Catal. A Gen. 1997, 152 (2), 173–182. Vollmer, I.; Yarulina, I.; Kapteijn, F.; Gascon, J. ChemCatChem 2019, 11 (1), 39–52. Sun, K.; Ginosar, D. M.; He, T.; Zhang, Y.; Fan, M.; Chen, R. Ind. Eng. Chem. Res. 2018, 57 (6), 1768–1789. Zhang, Z.-G. Carbon Resour. Convers. 2019, 2 (3), 157–174. Xu, Y.; Lin, L. Appl. Catal. A Gen. 1999, 188 (1), 53–67. Kosinov, N.; Hensen, E. J. M. Nonoxidative Dehydroaromatization of Methane. In Nanotechnology in Catalysis, Wiley-VCH, 2017; pp 469–482. Solymosi, F.; Cserényi, J.; Szöke, A.; Bánsági, T.; Oszkó, A. J. Catal. 1997, 165 (2), 150–161. Mosqueira, L.; Fuentes, G. A. Mol. Phys. 2002, 100 (19), 3055–3057. Xu, Y.; Liu, W.; Wong, S.-T.; Wang, L.; Guo, X. Catal. Lett. 1996, 40 (3), 207–214. Chen, L. Y.; Lin, L. W.; Xu, Z. S.; Li, X. S.; Zhang, T. J. Catal. 1995, 157 (1), 190–200. Kosinov, N.; Coumans, F. J. A. G.; Uslamin, E. A.; Wijpkema, A. S. G.; Mezari, B.; Hensen, E. J. M. ACS Catal. 2017, 7 (1), 520–529. Borry, R. W.; Kim, Y. H.; Huffsmith, A.; Reimer, J. A.; Iglesia, E. J. Phys. Chem. B 1999, 103 (28), 5787–5796. Kim, Y.-H.; Borry, R. W.; Iglesia, E. Microporous Mesoporous Mater. 2000, 35-36, 495–509. Tessonnier, J.-P.; Louis, B.; Walspurger, S.; Sommer, J.; Ledoux, M.-J.; Pham-Huu, C. J. Phys. Chem. B 2006, 110 (21), 10390–10395. Li, B.; Li, S.; Li, N.; Chen, H.; Zhang, W.; Bao, X.; Lin, B. Microporous Mesoporous Mater. 2006, 88 (1), 244–253. Gao, J.; Zheng, Y.; Jehng, J.-M.; Tang, Y.; Wachs, I. E.; Podkolzin, S. G. Science 2015, 348 (6235), 686. Zhou, D.; Ma, D.; Liu, X.; Bao, X. J. Chem. Phys. 2001, 114 (20), 9125–9129. Li, W.; Meitzner, G. D.; Borry, R. W.; Iglesia, E. J. Catal. 2000, 191 (2), 373–383.

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106.

325

Ma, D.; Shu, Y.; Bao, X.; Xu, Y. J. Catal. 2000, 189 (2), 314–325. Zhao, K.; Jia, L.; Wang, J.; Hou, B.; Li, D. New J. Chem. 2019, 43 (10), 4130–4136. Tessonnier, J.-P.; Louis, B.; Rigolet, S.; Ledoux, M. J.; Pham-Huu, C. Appl. Catal. A Gen. 2008, 336 (1), 79–88. Liu, L.; Wang, N.; Zhu, C.; Liu, X.; Zhu, Y.; Guo, P.; Alfilfil, L.; Dong, X.; Zhang, D.; Han, Y. Angew. Chem. Int. Ed. 2020, 59 (2), 819–825. Wang, D.; Lunsford, J. H.; Rosynek, M. P. Top. Catal. 1996, 3 (3), 289–297. Solymosi, F.; Erdöhelyi, A.; Szöke, A. Catal. Lett. 1995, 32 (1), 43–53. Weckhuysen, B. M.; Wang, D.; Rosynek, M. P.; Lunsford, J. H. J. Catal. 1998, 175 (2), 347–351. Ma, D.; Shu, Y.; Cheng, M.; Xu, Y.; Bao, X. J. Catal. 2000, 194 (1), 105–114. Jiang, H.; Wang, L.; Cui, W.; Xu, Y. Catal. Lett. 1999, 57 (3), 95–102. Vollmer, I.; Li, G.; Yarulina, I.; Kosinov, N.; Hensen, E. J.; Houben, K.; Mance, D.; Baldus, M.; Gascon, J.; Kapteijn, F. Catal. Sci. Technol. 2018, 8 (3), 916–922. Khan, T. S.; Balyan, S.; Mishra, S.; Pant, K. K.; Haider, M. A. J. Phys. Chem. C 2018, 122 (22), 11754–11764. López-Martín, A.; Caballero, A.; Colón, G. Mol. Catal. 2020, 486, 110787. Zheng, H.; Ma, D.; Bao, X.; Hu, J. Z.; Kwak, J. H.; Wang, Y.; Peden, C. H. F. J. Am. Chem. Soc. 2008, 130 (12), 3722–3723. Rahman, M.; Sridhar, A.; Khatib, S. J. Appl. Catal. A Gen. 2018, 558, 67–80. Ha, V. T. T.; Sarıoglan, A.; Erdem-S¸enatalar, A.; Taârit, Y. B. J. Mol. Catal. A Chem. 2013, 378, 279–284. Ding, W.; Li, S.; D Meitzner, G.; Iglesia, E. J. Phys. Chem. B 2001, 105 (2), 506–513. Zheng, Y.; Tang, Y.; Gallagher, J. R.; Gao, J.; Miller, J. T.; Wachs, I. E.; Podkolzin, S. G. J. Phys. Chem. C 2019, 123 (36), 22281–22292. Lezcano-González, I.; Oord, R.; Rovezzi, M.; Glatzel, P.; Botchway, S. W.; Weckhuysen, B. M.; Beale, A. M. Angew. Chem. Int. Ed. 2016, 55 (17), 5215–5219. Liu, H.; Shen, W.; Bao, X.; Xu, Y. Appl. Catal. A Gen. 2005, 295 (1), 79–88. Kosinov, N.; Wijpkema, A. S. G.; Uslamin, E.; Rohling, R.; Coumans, F. J. A. G.; Mezari, B.; Parastaev, A.; Poryvaev, A. S.; Fedin, M. V.; Pidko, E. A.; Hensen, E. J. M. Angew. Chem. Int. Ed. 2018, 57 (4), 1016–1020. Vollmer, I.; Kosinov, N.; Szécsényi, Á.; Li, G.; Yarulina, I.; Abou-Hamad, E.; Gurinov, A.; Ould-Chikh, S.; Aguilar-Tapia, A.; Hazemann, J.-L.; Pidko, E.; Hensen, E.; Kapteijn, F.; Gascon, J. J. Catal. 2019, 370, 321–331. Spivey, J. J.; Hutchings, G. Chem. Soc. Rev. 2014, 43 (3), 792–803. Liu, S.; Wang, L.; Ohnishi, R.; Ichikawa, M. J. Catal. 1999, 181 (2), 175–188. Ha, V. T. T.; Tiep, L. V.; Meriaudeau, P.; Naccache, C. J. Mol. Catal. A Chem. 2002, 181 (1), 283–290. Huang, X.; Jiao, X.; Lin, M.; Jia, L.; Hou, B.; Li, D. J. Fuel Chem. Technol. 2018, 46 (9), 1087–1100. Baba, T.; Miyaji, A. Conversion of Methane to Aromatic Hydrocarbons. In Catalysis and the Mechanism of Methane Conversion to Chemicals: C-C and C-O Bonds Formation Using Heterogeneous, Homogenous, and Biological Catalysts; Baba, T., Miyaji, A., Eds., Springer Singapore: Singapore, 2020; pp 127–163. Xu, Y.; Liu, S.; Guo, X.; Wang, L.; Xie, M. Catal. Lett. 1994, 30 (1), 135–149. Yin, F.; Li, M.-R.; Wang, G.-C. Phys. Chem. Chem. Phys. 2017, 19 (33), 22243–22255. Razdan, N. K.; Kumar, A.; Foley, B. L.; Bhan, A. J. Catal. 2020, 381, 261–270. Xu, Y.; Bao, X.; Lin, L. J. Catal. 2003, 216 (1), 386–395. Vollmer, I.; Abou-Hamad, E.; Gascon, J.; Kapteijn, F. ChemCatChem 2020, 12 (2), 544–549. Zhang, C.-L.; Li, S.; Yuan, Y.; Zhang, W.-X.; Wu, T.-H.; Lin, L.-W. Catal. Lett. 1998, 56 (4), 207–213. Shu, Y.; Ma, D.; Xu, L.; Bao, X.; Xu, Y. Chin. J. Catal. 2002, 23 (1), 18–24. Kiani, D.; Sourav, S.; Tang, Y.; Baltrusaitis, J.; Wachs, I. E. Chem. Soc. Rev. 2021, 50 (2), 1251–1268. Tan, P. J. Catal. 2016, 338, 21–29. Bajec, D.; Kostyniuk, A.; Pohar, A.; Likozar, B. Int. J. Energy Res. 2019, 43 (13), 6852–6868. Lai, Y.; Veser, G. Catal. Sci. Technol. 2016, 6 (14), 5440–5452. Vollmer, I.; Ould-Chikh, S.; Aguilar-Tapia, A.; Li, G.; Pidko, E.; Hazemann, J.-L.; Kapteijn, F.; Gascon, J. J. Am. Chem. Soc. 2019, 141 (47), 18814–18824. Denardin, F.; Perez-Lopez, O. W. Fuel 2019, 236, 1293–1300. Lim, T. H.; Kim, D. H. Appl. Catal. A Gen. 2019, 577, 10–19. Xu, Y.; Chen, M.; Wang, T.; Liu, B.; Jiang, F.; Liu, X. J. Catal. 2020, 387, 102–118. Kazansky, V. B.; Serykh, A. I.; Pidko, E. A. J. Catal. 2004, 225 (2), 369–373. Kazansky, V. B. Kinet. Catal. 2014, 55 (6), 737–747. Abdelsayed, V.; Smith, M. W.; Shekhawat, D. Appl. Catal. A Gen. 2015, 505, 365–374. Shu, Y.; Ohnishi, R.; Ichikawa, M. Appl. Catal. A Gen. 2003, 252 (2), 315–329. Abedin, M. A.; Kanitkar, S.; Bhattar, S.; Spivey, J. J. Catal. Today 2020, 343, 8–17. Iliuta, M. C.; Iliuta, I.; Grandjean, B. P. A.; Larachi, F. Ind. Eng. Chem. Res. 2003, 42 (14), 3203–3209. Denardin, F. G.; Perez-Lopez, O. W. Microporous Mesoporous Mater. 2020, 295, 109961. Xiong, Z.-T.; Chen, L.-L.; Zhang, H.-B.; Zeng, J.-L.; Lin, G.-D. Catal. Lett. 2001, 74 (3), 227–232. Ding, W.; Meitzner, G. D.; Marler, D. O.; Iglesia, E. J. Phys. Chem. B 2001, 105 (18), 3928–3936. Yang, J.; Ma, D.; Deng, F.; Luo, Q.; Zhang, M.; Bao, X.; Ye, C. Chem. Commun. 2002, 24, 3046–3047. Shu, Y.; Ma, D.; Xu, L.; Xu, Y.; Bao, X. Catal. Lett. 2000, 70 (1), 67–73. Martínez, A.; Peris, E.; Sastre, G. Catal. Today 2005, 107-108, 676–684. Wang, D. Y.; Kan, Q. B.; Xu, N.; Wu, P.; Wu, T. H. Catal. Today 2004, 93-95, 75–80. Abedin, M. A.; Kanitkar, S.; Bhattar, S.; Spivey, J. J. Catal. Today 2020, 343, 8–17. Kanitkar, S.; Abedin, M. A.; Bhattar, S.; Spivey, J. J. Appl. Catal. A Gen. 2019, 575, 25–37. Thakur, R.; Hoffman, M.; Vahid Mohammadi, A.; Smith, J.; Chi, M.; Tatarchuk, B.; Beidaghi, M.; Carrero, C. A. ChemCatChem 2020, 12 (14), 3639–3643. Dong, X.; Song, Y.; Lin, W. Catal. Commun. 2007, 8 (3), 539–542. Ma, D.; Lu, Y.; Su, L.; Xu, Z.; Tian, Z.; Xu, Y.; Lin, L.; Bao, X. J. Phys. Chem. B 2002, 106 (34), 8524–8530. Song, Y.; Sun, C.; Shen, W.; Lin, L. Catal. Lett. 2006, 109 (1), 21–24. Kikuchi, S.; Kojima, R.; Ma, H.; Bai, J.; Ichikawa, M. J. Catal. 2006, 242 (2), 349–356. Chu, N.; Wang, J.; Zhang, Y.; Yang, J.; Lu, J.; Yin, D. Chem. Mater. 2010, 22 (9), 2757–2763. Sarıoglan, A.; Savas¸çı, Ö. T.; Erdem-S¸enatalar, A.; Tuel, A.; Sapaly, G.; Ben Taârit, Y. J. Catal. 2007, 246 (1), 35–39. Huang, X.; Jiao, X.; Lin, M.; Wang, K.; Jia, L.; Hou, B.; Li, D. Catal. Sci. Technol. 2018, 8 (22), 5740–5749. Hu, J.; Wu, S.; Liu, H.; Ding, H.; Li, Z.; Guan, J.; Kan, Q. RSC Adv. 2014, 4 (51), 26577–26584. Tempelman, C. H. L.; Portilla, M. T.; Martínez-Armero, M. E.; Mezari, B.; de Caluwé, N. G. R.; Martínez, C.; Hensen, E. J. M. Microporous Mesoporous Mater. 2016, 220, 28–38. Wu, Y.; Emdadi, L.; Oh, S. C.; Sakbodin, M.; Liu, D. J. Catal. 2015, 323, 100–111. Ghorbanpour, A.; Gumidyala, A.; Grabow, L. C.; Crossley, S. P.; Rimer, J. D. ACS Nano 2015, 9 (4), 4006–4016. Jin, Z.; Liu, S.; Qin, L.; Liu, Z.; Wang, Y.; Xie, Z.; Wang, X. Appl. Catal. A Gen. 2013, 453, 295–301.

326

Heterogeneous catalysts for the non-oxidative conversion of methane to aromatics and olefins

107. Zhu, P.; Yang, G.; Sun, J.; Fan, R.; Zhang, P.; Yoneyama, Y.; Tsubaki, N. J. Mater. Chem. A 2017, 5 (18), 8599–8607. 108. Gao, Y.; Neal, L.; Ding, D.; Wu, W.; Baroi, C.; Gaffney, A. M.; Li, F. ACS Catal. 2019, 9 (9), 8592–8621. 109. Ye, M.; Li, H.; Zhao, Y.; Zhang, T.; Liu, Z. MTO Processes Development: The Key of Mesoscale Studies. In Advances in Chemical Engineering; Marin, G. B., Li, J., Eds.; vol. 47; Academic Press, 2015; pp 279–335. Chapter Five. 110. Tian, P.; Wei, Y.; Ye, M.; Liu, Z. ACS Catal. 2015, 5 (3), 1922–1938. 111. Ali, K. A.; Abdullah, A. Z.; Mohamed, A. R. Renew. Sust. Energ. Rev. 2015, 44, 508–518. 112. Masudi, A.; Jusoh, N. W. C.; Muraza, O. Catal. Sci. Technol. 2020, 10 (6), 1582–1596. 113. Snel, R. Catal. Rev. 1987, 29 (4), 361–445. 114. Koerts, T.; van Santen, R. A. J. Chem. Soc. Chem. Commun. 1991, 18, 1281–1283. 115. Koerts, T.; Deelen, M. J. A. G.; van Santen, R. A. J. Catal. 1992, 138 (1), 101–114. 116. Mochida, I.; Aoyagi, Y.; Fujitsu, H. Chem. Lett. 1990, 19 (9), 1525–1526. 117. Fang, T.; Yeh, C.-T. J. Catal. 1981, 69 (1), 227–229. 118. Kurosaka, T.; Matsuhashi, H.; Arata, K. J. Catal. 1998, 179 (1), 28–35. 119. Guo, X.; Fang, G.; Li, G.; Ma, H.; Fan, H.; Yu, L.; Ma, C.; Wu, X.; Deng, D.; Wei, M.; Tan, D.; Si, R.; Zhang, S.; Li, J.; Sun, L.; Tang, Z.; Pan, X.; Bao, X. Science 2014, 344 (6184), 616–619. 120. Ruitenbeek, M.; Weckhuysen, B. M. Angew. Chem. Int. Ed. 2014, 53 (42), 11137–11139. 121. García-Moncada, N.; van Rooij, G.; Cents, T.; Lefferts, L. Catal. Today 2021, 369, 210–220. 122. Dipu, A. L.; Ohbuchi, S.; Nishikawa, Y.; Iguchi, S.; Ogihara, H.; Yamanaka, I. ACS Catal. 2020, 10 (1), 375–379. 123. Oh, S. C.; Schulman, E.; Zhang, J.; Fan, J.; Pan, Y.; Meng, J.; Liu, D. Angew. Chem. Int. Ed. 2019, 58 (21), 7083–7086. 124. Hao, J.; Schwach, P.; Fang, G.; Guo, X.; Zhang, H.; Shen, H.; Huang, X.; Eggart, D.; Pan, X.; Bao, X. ACS Catal. 2019, 9 (10), 9045–9050. 125. Han, S. J.; Lee, S. W.; Kim, H. W.; Kim, S. K.; Kim, Y. T. ACS Catal. 2019, 9 (9), 7984–7997. 126. Sakbodin, M.; Wu, Y.; Oh, S. C.; Wachsman, E. D.; Liu, D. Angew. Chem. Int. Ed. 2016, 55 (52), 16149–16152. 127. Xie, P.; Pu, T.; Nie, A.; Hwang, S.; Purdy, S. C.; Yu, W.; Su, D.; Miller, J. T.; Wang, C. ACS Catal. 2018, 8 (5), 4044–4048. 128. Xiao, Y.; Varma, A. ACS Catal. 2018, 8 (4), 2735–2740. 129. Sheng, H.; Schreiner, E. P.; Zheng, W.; Lobo, R. F. ChemPhysChem 2018, 19 (4), 504–511. 130. Nishikawa, Y.; Ogihara, H.; Yamanaka, I. ChemistrySelect 2017, 2 (16), 4572–4576. 131. Gerceker, D.; Motagamwala, A. H.; Rivera-Dones, K. R.; Miller, J. B.; Huber, G. W.; Mavrikakis, M.; Dumesic, J. A. ACS Catal. 2017, 7 (3), 2088–2100. 132. Dutta, K.; Shahryari, M.; Kopyscinski, J. Ind. Eng. Chem. Res. 2020, 59 (10), 4245–4256. 133. Dutta, K.; Chaudhari, V.; Li, C.-J.; Kopyscinski, J. Appl. Catal. A Gen. 2020, 595, 117430. 134. Bajec, D.; Kostyniuk, A.; Pohar, A.; Likozar, B. Chem. Eng. J. 2020, 396, 125182. 135. Ogihara, H.; Imai, N.; Yoshida-Hirahara, M.; Kurokawa, H. Chem. Lett. 2020, 49 (3), 236–239. 136. Zeng, Z.-Y.; Chen, J.; Lin, J.-H. Front. Res. Today 2018, 1, 1007. 137. Chen, Y.; Wang, X.; Luo, X.; Lin, X.; Zhang, Y. Chin. J. Chem. 2018, 36 (6), 531–537. 138. Tajima, H.; Ogihara, H.; Yoshida-Hirahara, M.; Kurokawa, H. New J. Chem. 2020, 44 (40), 17198–17202. 139. Abedin, M. A.; Bhattar, S.; Spivey, J. J. SN Appl. Sci. 2020, 2 (12), 2012. 140. Sot, P.; Newton, M. A.; Baabe, D.; Walter, M. D.; van Bavel, A. P.; Horton, A. D.; Copéret, C.; van Bokhoven, J. A. Chem. Eur. J. 2020, 26 (36), 8012–8016. 141. Li, T.-H.; Yan, M.; Liu, Y.; Huang, Z.-Q.; Chang, C.-R.; Li, J. J. Phys. Chem. C 2020, 124 (25), 13656–13663. 142. Liu, Y.; Liu, J.-C.; Li, T.-H.; Duan, Z.-H.; Zhang, T.-Y.; Yan, M.; Li, W.-L.; Xiao, H.; Wang, Y.-G.; Chang, C.-R.; Li, J. Angew. Chem. Int. Ed. 2020, 59 (42), 18586–18590. 143. Kim, S. K.; Kim, H. W.; Han, S. J.; Lee, S. W.; Shin, J.; Kim, Y. T. Commun. Chem. 2020, 3 (1), 58. 144. Ma, X.; Sun, K.; Liu, J.-X.; Li, W.-X.; Cai, X.; Su, H.-Y. J. Phys. Chem. C 2019, 123 (23), 14391–14397. 145. Wen, J.-H.; Wang, G.-C. J. Phys. Chem. C 2020, 124 (24), 13249–13262. 146. Kashaboina, U.; Nishikawa, Y.; Wakisaka, Y.; Sirisit, N.; Nagamatsu, S.-i.; Bao, D.; Ariga-Miwa, H.; Takakusagi, S.; Inami, Y.; Kuriyama, F.; Dipu, A. L.; Ogihara, H.; Iguchi, S.; Yamanaka, I.; Wada, T.; Asakura, K. Chem. Lett. 2019, 48 (9), 1145–1147. 147. Nishikawa, Y.; Ohtsuka, Y.; Ogihara, H.; Rattanawan, R.; Gao, M.; Nakayama, A.; Hasegawa, J.-y.; Yamanaka, I. ACS Omega 2020, 5 (43), 28158–28167. 148. Al Rashid, M. H.; Dipu, A.; Nishikawa, Y.; Ogihara, H.; Inami, Y.; Obuchi, S.; Yamanaka, I.; Nagamatsu, S.; Kido, D.; Asakura, K. e-J. Surf. Sci. Nanotechnol. 2020, 18, 24–27. 149. Chaudhari, V.; Dutta, K.; Li, C.-J.; Kopyscinski, J. Mol. Catal. 2020, 482, 110606. 150. Zhu Chen, J.; Wu, Z.; Zhang, X.; Choi, S.; Xiao, Y.; Varma, A.; Liu, W.; Zhang, G.; Miller, J. T. Catal. Sci. Technol. 2019, 9 (6), 1349–1356. 151. Miao, S.; Wang, Y.; Ma, D.; Zhu, Q.; Zhou, S.; Su, L.; Tan, D.; Bao, X. J. Phys. Chem. B 2004, 108 (46), 17866–17871. 152. Ding, B.; Huang, S.; Wang, W. Appl. Surf. Sci. 2008, 254 (16), 4944–4948. 153. Yao, Y.; Huang, Z.; Xie, P.; Wu, L.; Ma, L.; Li, T.; Pang, Z.; Jiao, M.; Liang, Z.; Gao, J.; He, Y.; Kline, D. J.; Zachariah, M. R.; Wang, C.; Lu, J.; Wu, T.; Li, T.; Wang, C.; Shahbazian-Yassar, R.; Hu, L. Nat. Nanotechnol. 2019, 14 (9), 851–857. 154. Julian, I.; Roedern, M. B.; Hueso, J. L.; Irusta, S.; Baden, A. K.; Mallada, R.; Davis, Z.; Santamaria, J. Appl. Catal. B Environ. 2020, 263, 118360. 155. Yu, X.; Zholobenko, V. L.; Moldovan, S.; Hu, D.; Wu, D.; Ordomsky, V. V.; Khodakov, A. Y. Nat. Energy 2020, 5 (7), 511–519.

6.15

Inorganic catalysis for methane conversion to chemicals

Guangzong Fanga, Dunfeng Gaoa, Xiulian Pana, Guoxiong Wanga, and Xinhe Baoa,b, a The State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, The Chinese Academy of Science, Beijing, China; and b University of Science and Technology of China, Hebei, China © 2023 Elsevier Ltd. All rights reserved.

6.15.1 6.15.2 6.15.3 6.15.3.1 6.15.3.1.1 6.15.3.1.2 6.15.3.2 6.15.3.2.1 6.15.3.2.2 6.15.3.2.3 6.15.3.2.4 6.15.3.3 References

Background Physicochemical properties of methane Development of methane activation and conversion technologies Thermo-catalytic conversion of methane Indirect methane conversion Direct methane conversion Electrocatalytic conversion of methane Introduction Fundamentals of electrocatalytic conversion of methane Electrocatalytic conversion of methane to fuels and chemicals Photoelectrocatalytic conversion of methane Summary

327 329 329 330 330 331 343 343 345 346 350 351 351

Abstract As an important fossil energy, natural gas has played an important role as a fuel for power and heat generation. With the development of renewable energies under the pressure of the climate change and environmental conservation, there is an urge to turn the fossil fuel era to fossil chemicals era to meet the ever increasing demanding for carbon-containing materials. As the main component of the abundant natural gas resources, methane conversion to chemicals is one of the keys for this purpose. This chapter overviews the well-developed indirect conversion technologies via syngas intermediate stage and direct conversion technologies which are still under development, including the thermocatalytic conversion and emerging electrocatalytic and photoelectrocatalytic conversion technologies. While significant progress has been made in both mechanistic understanding and catalyst optimization, there are still great challenges for industrial applications and innovations for more efficient catalysts are still highly needed.

6.15.1

Background

With the reducing reserves and the price fluctuations of crude oil, natural gas receives increasing attention as an important source of clean fossil energy. By the end of 2019, the reserve of natural gas was proved to be 198.8 trillion cubic meters globally. In the meanwhile, large reserves of shale gas, coalbed methane, and methane hydrate have been discovered recently. The BP Statistical Review of World Energy 2020 shows the distribution of natural gas distribution globally including shale gas (Fig. 1).1,2 The most abundant natural gas is located in North America, Middle East, Asia Pacific. Natural gas is mainly composed of methane with a small amount of ethane, propane, nitrogen, and butane. Methane is also the main component of shale gas, coalbed methane, and methane hydrate. It can react with oxygen, giving CO2 and H2O as the products, and at the same time, producing heat (Eq. 1). CH4 þ 2O2 ¼ CO2 þ 2H2 O DH ¼ 191:8 kJ=mol

(1)

Therefore, it is widely used as a fuel in automobiles, households, and so on. It is also utilized for power generation because it has the advantages of high-power efficiency, less emission than coal, and low construction cost. Eq. (1) shows that combustion of one mole methane would release 191.8 kJ/mol, and at the same time giving one mole CO2. Moreover, methane molecule contains carbon and thus can be used as a clean feedstock for the production of carbon-containing chemicals, which traditionally relies on oil. According to the Natural Gas Annual 2019 report from U.S. Energy Information Administration,3 natural gas is mainly used in electrical power generation, industrial, residential, and commercial, as shown in Fig. 2. Other sectors such as transportation could be included in electrical power generation and are negligible while comparing to the above-mentioned applications. With the development of the society and increasing energy demand, CO2 emission also increases in different sectors, as shown in Fig. 3.4 This has imposed great pressure on the sustainable development of society to reduce the carbon footprint. In comparison, oxidation of hydrogen would release 115.6 kJ/mol but giving environmentally benign water as a product (Eq. 2). 2H2 þ O2 ¼ 2H2 O DH ¼  115:6 kJ=mol

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00048-0

(2)

327

328

Inorganic catalysis for methane conversion to chemicals

Natural gas: Production by region

Natural gas: Consumption by region

Billion cubic meters

Billion cubic meters

Asia Pacific Africa Middle East CIS Europe S. & Cent. America North America

94

99

04

09

14

19

4000

4000

3500

3500

3000

3000

2500

2500

2000

2000

1500

1500

1000

1000

500

500

0

94

99

04

09

14

19

0

Fig. 1 Global natural gas production and consumption by regions in 2019. Data source: the BP Statistical Review of World Energy 2020. BP plc, BP Statistical Review of World Energy, 2020.

Therefore, there is a great urge to develop renewable energies, such as wind, solar, tide, and nuclear for power and heat generation, but utilizing fossils as a carbon resource for the production of chemicals, such as olefins, aromatics, liquid fuels, and hydrogen. However, most natural gas and shale gas resources are located in remote areas. Long-distance transportation is not economically viable. Therefore, it is desirable to develop highly efficient technologies to transform natural gas on site into high energy density and high value-added basic chemicals, which are easy to transport at a lower cost.

trillion cubic feet 12

10

8

6

4 2

0 2015

2016

2017

2018

2019

Fig. 2 Natural gas consumption by sector, reported by U.S. Energy Information Administration, Natural Gas Annual 2019. Energy Information Administration of US. Natural Gas Annual 2019. 2019 https://www.eia.gov/naturalgas/annual/, p. 49.

Inorganic catalysis for methane conversion to chemicals

329

Fig. 3 Global CO2 emissions by sector in 2017, with emissions due to electricity and heat generation further reallocated by end-use sector (purple, pink, orange and green represent industry, buildings, transportation and other sectors, respectively). International Energy Agency, CO2 Emissions From Fuel Combustion 2019, 2019.

6.15.2

Physicochemical properties of methane

Methane is a colorless and odorless gas with a boiling point of 111.6 K. It is flammable with its concentration ranging 5.4–17% in the air at ambient pressure. It is also an important greenhouse gas because it absorbs infrared radiation in a region of the terrestrial IR spectrum. Methane molecule has four equivalent CeH bonds organized in a tetrahedral structure due to the sp3 hybridization of the central carbon atom. It is the most stable and smallest organic molecule. The dissociation energy of the CeH bond is as high as 439 kJ/mol. It is higher than the CeH bond energy in higher hydrocarbons but lower than that in benzene. This highly symmetrical molecule exhibits extremely small polarizability (2.84  10 40 C2 m2 J 1).5 It implies that an extremely strong electric field will be needed to polarize the molecule or to make it feasible for a nucleophilic or electrophilic attack. This intrinsically chemical inertness frequently requires harsh reaction conditions such as high temperature and strong oxidants to activate the methane molecule. However, the greater challenge lies in controlled activation, steering the selective conversion rather than deep oxidation giving CO2 or deep dehydrogenation (pyrolysis) giving coke. Therefore, its direct conversion to value-added chemicals remains a grand challenge for decades. It has been long considered as a “Holy Grail” in the field of catalysis and even chemistry.

6.15.3

Development of methane activation and conversion technologies

As the main component of natural gas, shale gas, coal-bed methane, and methane hydrate, methane conversion is the key for optimal utilization of these carbon resources. There are extensive studies exploring methane conversion to chemicals and fuels, as shown in Fig. 4. The most mature technology nowadays is the indirect conversion via synthesis gas (so-called syngas). Syngas is then upgraded to valuable chemicals and fuels such as olefins, gasoline, diesel, aromatics via Fischer-Tropsch synthesis, methanol-based technology such as methanol-to-olefins (MTO) commercialized by Dalian Institute of Chemical Physics and UOP,6 MTG, MDA (methane dehydroaromatization),7–13 and alternatively the emerging OXZEO platform technology.14 In comparison, direct conversion of methane would be much more energy- and cost-effective and therefore is under extensive study for decades. Various catalysts and catalytic processes have been studied, for instance, methane sulfonation, halogenation, borylation, selective oxidation of methane (SOM) to oxygenates, oxidative coupling of methane (OCM) to ethylene, nonoxidative direct conversion dehydroaromatization to aromatics (MDA), and to olefins, aromatics and hydrogen (MTOAH).1,15–17 In addition to traditional thermo-catalytic conversion, there are also increasing efforts exploring the electrochemical activation of methane and photoelectrocatalytic conversion of methane.

330

Inorganic catalysis for methane conversion to chemicals

Halogenation

Fig. 4

Thermo-catalytic methane activation

Sulfonation Borylation Seletive oxidation Oxidative coupling Dehydroarom atization

Hydrocarbons

MTOAH

Indirect route

Other technologies

Catalytic Methane Conversion

Direct route

Methanol

DRM, SRM, POM

Electrocatalytic conversion of methane

Syngas

Hydrocarbons

C2+ oxygenates

Photoelectrocatalytic conversion of methane

Methanol

Different routes of catalytic methane conversion.

6.15.3.1 6.15.3.1.1

Thermo-catalytic conversion of methane Indirect methane conversion

Methane conversion to syngas is well-practiced in industry, which is also an important intermediate stage of indirect methane conversion to chemicals. There are three technologies available, i.e., steam reforming, partial oxidation, and drying reforming. They are all operated at elevated temperatures, i.e., 873–1373 K. CO2 ð g Þ þ CH4 ð g Þ ¼ 2COð g Þ þ 2H2 ð g ÞDH0r;298K ¼ 247 kJ=mol

(3)

H2 Oð g Þ þ CH4 ð g Þ ¼ COð g Þ þ 3H2 ð g ÞDH0r;298K ¼ 206 kJ=mol

(4)

1 þ CH4 ð g Þ ¼ COð g Þ þ 2H2 ð g ÞDH0r;298K ¼  35kJ=mol 2O2 ð g Þ

(5)

The main features of these syngas production technologies are summarized in Table 1. Steam reforming of methane is one of the most important and economical hydrogen production methods nowadays. This process has been developed since 1926 and it is a rather mature technology for hydrogen production.18–20 Almost 50% of the worldwide hydrogen ( 55  106 tons/year) is derived from natural gas.21 In comparison, dry reforming attracts increasing attention in recent years, because it can reduce carbon footprint by converting both greenhouse gasses in one process, i.e., CO2 and CH4 into valuable syngas. The resulting syngas has a H2/CO ratio of 1/1. Despite extensive studies on mechanistic understanding, catalyst optimization, reaction conditions, kinetics, and so on,22–28 there are still many challenges, including (1) deactivation of catalysts caused by catalyst sintering and coke deposition; (2) cold spots in the catalyst bed due to the strong endothermic endothermal reaction. Partial oxidation of methane to syngas is exothermic. Its reaction rate is generally one or two orders of magnitude faster than those of steam reforming and dry reforming and consequently has also attracted much attention.29–32 The resulting syngas has a H2/CO ratio of about 2, which is ideal as the feedstock for Fischer-Tropsch synthesis. This has been the key for Gas-to-Liquid, so-called GTL technology. Additionally, POM has the advantages of a smaller size reactor, a higher efficiency and a lower energy consumption compared with the endothermic SRM and DRM processes. However, it also faces great challenges, for instance, (1)

Inorganic catalysis for methane conversion to chemicals

331

Table 1

Main features of SRM/DRM/POM.

Routes a

Advantages

Disadvantages

SRM

Syngas: H2/CO3. Mature technology for hydrogen production

DRM

High yield; Syngas: H2/CO1 Syngas: H2/CO2; Lower energy cost; High conversion, good selectivity, and short residence time

Endothermal reaction; Steam requirement; Deactivation of catalysts Endothermal reaction Deactivation of catalysts Exothermal reaction Hot spots on catalysts; Desulphurization requirement; Safety issue

POM

a

SRM, steam reforming of methane; DRM, dry reforming of methane; POM, partial oxidation of methane.

the explosive risks because of oxygen as a reactant; (2) the separation cost for unreacted O2 and CO2 from syngas; (3) local hot spots in the catalyst bed due to the exothermicity, which may lead to sintering and hence deactivation of catalysts; (4) carbon deposition.

6.15.3.1.2

Direct methane conversion

6.15.3.1.2.1 Low and medium temperature methane conversion Low and medium temperature (typically below 773 K) methane conversion was dated back to the 1970s.33 There are several typical processes, including sulfonation, halogenation, and borylation. 6.15.3.1.2.1.1 Methane sulfonation Protic acids such as concentrated sulfuric acid or trifluoroacetic acid are usually used as solvents for sulfonation. Under such harsh conditions, methane could be converted to esters and thus avoid over-oxidation to give a promising selectivity. Meanwhile, the boiling points of ester compounds are usually relatively low and could be easily separated from the solvent and followed by hydrolysis to generate methanol. Snyder and Grosse34 first reported methane sulfonation with sulfur trioxide using mercury salt catalyst forming CH3eO, CH3eS, and SeCH2eS bonds in the 1950s. Periana et al.35 reported Hg(O3SCF3)2 in triflic acid could also catalyze methane to CH3OSO2CF, with the selectivity of 85% and yield of 50%. H2PtCl6/H2PtCl4 catalyst was first reported by Shilov et al.36 But it was not stable during reaction due to the reduction of the catalyst forming Pt black. Thereafter, efforts were made to modify the coordination environment to stabilize the catalyst. For instance, Periana et al. developed several catalysts for methane oxidation to methanol.37 For example, using Pt(NH3)2Cl2 as the catalyst, methane was converted to methyl sulfate in highly concentrated sulfuric acid under 493 K, 3.5 MPa, which gave a yield as high as 70%.37 However, due to the irreversible protonation of Pt(NH3)2Cl2 to PtCl2 in this system, the catalyst rapidly deactivated. Then 2-20 -bipyridine was introduced as a ligand, which prevented protonation and hence exhibited a good stability in concentrated acid. As shown in Fig. 5, methane (3.4 MPa, 115 mmol) was successfully converted to methyl bisulfate (81% selectivity), equivalent to  1 M methyl bisulfate within 2.5 h in 80 mL of 102% H2SO4 at 493 K.38 In comparison to the mercury catalyst system, the Pt-ligated system can operate in the presence of excess SO3 without forming the undesirable CeS bonded by-product, CH3SO3H.38 Uremia et al.39 reported a yield of methanesulfonic acid as high as 99% by reacting methane with SO3 using K2S2O8 as a catalyst. The electrophilic initiator based on a sulfonyl peroxide derivative was protonated under super acidic conditions, which was considered to be the key. Thus, a highly electrophilic oxygen species was obtained, which was capable of activating methane forming CH3þ as a key intermediate. The authors envisioned that this method was scalable to a pilot plant, as displayed in Fig. 6, by connecting reactors in series and expected to have a production capacity of up to 20 metrics.39 6.15.3.1.2.1.2 Methane halogenation/oxy-halogenation Studies demonstrated that various halogens can activate methane. However, fluorine is too reactive while reacting with methane, and methyl iodide is likely to decompose. Only chlorine (Eq. 6) and bromine (Eq. 7) are of practical interest. For instance, Degirmenci et al.40 report that methane reacted with Br2 catalyzed by 25 wt% SO42/ZrO2 modified SBA-15 giving methane conversion 69% and methane bromide selectivity 99% at 613 K. CH4 ð g Þ þ Cl2 ð gÞ ¼ CH3 Clð g Þ þ HClð g ÞDH0r;298K ¼  99:6 kJ=mol

(6)

CH4 ð g Þ þ Br 2 ð g Þ ¼ CH3 Br ð g Þ þ HBr ð g ÞDH0r;298K ¼  28:0 kJ=mol

(7)

The halogenation of methane has the advantages of high selectivity and easy dehalogenation but suffers from a large amount of by-product haloid acid in the product and a poor halogen atom efficiency. Alternatively, methane is converted to CH3Br and CH2Br2 via oxy-halogenation (Eqs. 8 and 9), because of the high selectivity of methyl bromide and easy hydrolysis.41 The reaction of O2 and

332

Inorganic catalysis for methane conversion to chemicals

Fig. 5 The mechanism for Pt2þ catalyzed oxidation of methane. Periana, R.A.; Taube, D.J.; Gamble, S.; Taube, H.; Satoh, T., et al., Science 1998, 280 (5363), 560–4.

HBr forming Br2 under the catalysis of oxides of Ru, Ir, Ce, etc., followed by the reaction with methane to complete the reaction.42 HBr is essentially recycled in the whole process.42 2CH4 ðgÞ þ 2O2 ðgÞ þ HBr ðgÞ ¼ CH3 Br ðgÞ þ COðgÞ þ 3H2 OðgÞ

DH0r;298K ¼ 686 kJ=mol 3CH4 ðgÞ þ 4O2 ðgÞ þ 2HBr ðgÞ ¼ CH2 Br 2 ðgÞ þ 2COðgÞ þ 6H2 OðgÞ

DH0r;298K ¼ 382:1 kJ=mol

(8)

(9)

The best yield was achieved using CeO2 as the catalyst but accompanied by the formation of a large amount of COx in the presence of O2.42 Lin et al. reported FePO4/SiO2 catalysts give a methane conversion of 50%, the total selectivity of CH3Br with CO over 95% is indicated. The catalyst was also tested for 200 h.43 Paunovic et al. reported that the by-products (CO2) can be suppressed

Fig. 6

Schematic of the Sulfonation of methane process. Diaz-Urrutia, C.; Ott, T., Science 2019, 363 (6433), 1326–1329.

Inorganic catalysis for methane conversion to chemicals

333

using vanadium phosphate (VPO) as the catalyst.42 Methane conversion kept around 25% with CH3Br selectivity  62% and CH2Br2  20% at 753 K in 100 h stability test. They further predicted theoretically that the optimal catalytic material should sit at the apex of the pyramid in Fig. 7.42 The above studies show that highly selective methane sulfonation can be obtained and the resulting methanesulfonic acid can be converted to methanol via hydrolysis, which can be further converted to value-added chemicals. However, the reaction is usually carried out in batch reactors. In addition, extreme conditions of strong acid and oxidant impose great challenges for industrial applications. Although oxy-halogenation can readily lead to the formation of CH3Br, it is also faced with additional challenges of selectivity as well as corrosive and toxic conditions. 6.15.3.1.2.1.3 Methane borylation Methane borylation was reported using Ru,44 Ir,45 Rh,46 as catalysts and the major products were CH3Bpin and CH2(Bpin)2 in the presence of bis-pinacolborane (B2pin2) (Fig. 8A).47 This methodology was also found to be applicable for ethane although giving a lower selectivity. Smith et al.48 improved the chemoselectivity for monoborylated versus diborylated methane using iridium phenanthroline complex and a yield as high as 52% and turnover number of 100 was obtained under optimized conditions of 323 K and 3.5 MPa (Fig. 8B). Liu et al.49 reported that a metal-free N-heterocyclic carbene-stabilized borenium complex [Ph3C]þ[B(C6F5)4] solved in o-C6H4F2 can catalyze the direct conversion of methane under ambient pressure with high selectivity toward methane monoborylation. The yield was as high as 61% at 383 K after recrystallization (Fig. 8C). The authors suggested the s-bond metathesis mechanism of the CeH bond. 6.15.3.1.2.1.4 Selective oxidation of methane to methanol (SOM) Selective oxidation of methane has been considered as “dream reactions”50 and was first reported by Lance and Elworthy as early as 1906.51 Trace amounts of methanol/formaldehyde/formic acid were detected with the assistant of the FeS catalyst. Then selective oxidation of methane to methanol became one of the research hot topics in this field since the 1970s. Table 2 lists methane oxidation processes and the corresponding Gibbs free energies. One sees that deep oxidation of methane is the most favorable thermodynamically. For the synthesis of partially oxidized products such as methanol, formaldehyde, and hydrocarbons such as ethylene and ethane, the challenge lies in the selectivity control because these products are more reactive than methane in the presence of oxygen. Over-oxidation would lead to the formation of CO and CO2 and hence reducing the carbon utilization efficiency. This is particularly difficult when the reaction temperature exceeds 773 K in order to enhance the reaction kinetics. In response to the above-mentioned challenges, extensive efforts have been made, which led to significant progress in the last few decades.14,16,17,52–56 One elegant example is the research inspired by methane monooxygenase (MMO), which catalyzes methane conversion selectively to methanol in the presence of oxygen in nature. The mononuclear and dinuclear copper sites were identified and the dinuclear sites were generally considered as the active sites for methane activation. Mimicking the active sites of MMO, several transition metal doped zeolites were synthesized for methane oxidation, as summarized by Zhao et al. (Table 3).73 For

Fig. 7 Performance descriptors for the design of selective methane oxybromination catalysts. Paunovic, V.; Zichittella, G.; Moser, M.; Amrute, A. P.; Perez-Ramirez, J., Nat. Chem. 2016, 8 (8), 803–9.

334

Inorganic catalysis for methane conversion to chemicals

Fig. 8 (A) Proposed methane C-H borylation reaction. (B) Ir reagent Ligand is used in the borylation of methane. (C) Conversion of methane into MeBcat initiated by borenium2. (A) Cook, A.K.; Schimler, S.D.; Matzger, A.J.; Sanford, M.S., Science 2016, 351 (6280), 1421–4, (B) Smith, K.T.; Berritt, S.; Gonzalez-Moreiras, M.; Ahn, S.; Smith, M.R., 3rd, et al., Science 2016, 351 (6280), 1424–7, (C) Liu, Y.; Dong, W.; Li, Z.H.; Wang, H., Chem 2021, 7 (7), 1843–1851.

instance, Panov and his coworker reported FeZSM-5 for methane oxidation using N2O as the oxidant to weaken the overoxidation.57 A methanol selectivity of 87% was further obtained.60 Michalkiewicz54 reported that a Fe/HZSM-5 (Si/Fe ¼ 22) catalyst containing mono-nuclear or di-nuclear active center Fe]O, similar to those in MMO, gave methane conversion of 31.51% but at the expenses of deep oxidation of CH3OH and CH3OH selectivity was only 10.79%, HCHO 17.06% and CO2 72.14% at 903 K. In comparison, the over-oxidation could be suppressed at a lower temperature, for instance at 623 K, CH3OH selectivity reached 63.47%, HCHO 14.84%, and CO2 only 21.68%, but methane conversion was only 0.10%.54 Therefore, much more active catalysts are highly desired to alleviate and even avoid high temperature over-oxidation. To avoid strong oxidant O2 and also avoid using expensive greenhouse gas N2O as the oxidant for SOM, efforts were made using in situ generated H2O2 and H2O as the oxidant. For example, Xiao and his coworker recently reported an interesting strategy by

Inorganic catalysis for methane conversion to chemicals Table 2

335

Selective oxidation process and the Gibbs free energy (temperature unit: K).

DGr (kJ mol 1) No.

Reaction

298

650

700

750

800

1000

R1 R2 R3 R4 R5 R6 R7 R8 R9

CH4 þ 1/2O2 / CH3OH CH4 þ O2 / HCOOH þ H2O CH4 þ 1.5O2 / CO þ H2O CH4 þ 2O2 / CO2 þ H2O CH4 þ 1/2O2 / 1/2C2H4 þ H2O CH4 þ 1/4O2 / 1/2C2H6 þ 1/2H2O CH4 þ 1/2O2 / CO þ 2H2 CH4 þ H2O / CO þ 3H2 CH4 þ CO2 / 2CO þ 2H2

111 288 544 801 144 80 86 142 171

93 294 573 800 147 69 152 60 73

 91  294  578  799  147  67  162 48 61

88 295 582 799 147 076 172 36 47

86 296 586 799 147 63 182 23 33

76 298 693 798 147 55 222 27 23

Zhang, Q.; He, D.; Zhu, Q., J. Nat. Gas Chem. 2003, 12, 9.

immobilizing AuPd alloy nanoparticles in aluminosilicate zeolite crystals for the generation of hydrogen peroxide by reacting H2 and O2.17 They further treated hydrophobically the outer surface of the zeolite with organosilane. Thus, the in-situ generated hydrogen peroxide at 70  C could be kept close to the active sites of AuPd, where methane was selectively oxidized to methanol. Interestingly, over-oxidation was not observed and a methanol selectivity as high as 92% can be obtained at a methane conversion of 17.3%, equivalent to a methanol formation rate of 82 mmol g 1 h 1. This yield was three times higher than the catalyst without hydrophobic treatment.17 Van Bokhoven and coworkers reported mordenite supported Cu as catalyst and water as the oxygen source for the direct conversion of methane to methanol.16,56 A methanol selectivity of 97% was obtained and it produced 0.204 mol CH3OH per mole copper in zeolite at 473 K. By in situ XAS, IR, and DFT simulation, they suggested a two-step mechanism. Methane was first oxidized at the Cu(II) site, and these copper species were reduced to Cu(I), which was re-oxidized by water in the subsequent step to complete the catalytic cycle.16,56 Agarwal et al. reported a methanol selectivity of 92% with polyvinylpyrrolidone (PVP) stabilized Au-Pd nanoparticle colloidal catalyst, H2O2, and O2 were used as oxidants together.74 The amount of released oxygenates was greater than the H2O2 consumption, indicating the incorporation of molecular oxygen in this process. Thus, molecular oxygen can be used in the selective oxidation of methane for practical applications if the catalyst is carefully designed.74

Table 3

Catalysts for methane oxidation.

Catalyst

Temp. (K)

CH3OH concentration (mmol/gcat)

CH3OH concentration (mmol/gmetal)

CH3OH selectivity (%)

References

Fe-ZSM-5

RT RT RT 433 433 448 473 423 423 423 423 473 473 423 473 473 423 473 423 150 175

5.0 20.3 23.0 213.2 160.0 238.7 265.2 16 1.6 39 31 28.1 12.0 31.0 170 59.4 97.0 11.5 0.39 1.0 2

– 214 64.4 597.0 448.0 668.4 742.3 30 4.7 50 60 50 18.0 40.0 340 160 310 37.5 2.1 1.4 5.9

80 – 45 87 76 77 68 – – – –

57 58 59 60 61 60 60 62 63 62 62 64 65 62 66 67 68 69 70 71 72

Cu-ZSM-5 Cu-SSZ Cu-MOR

Cu-SiO2 Co-ZSM-5 Ni-ZSM-5

Zhao, G.; Drewery, M.; Mackie, J.; Oliver, T.; Kennedy, E. M., et al., Energ. Technol. 2020, 8 (8), 1900665.

– – – – – 38.0 75 –

336

Inorganic catalysis for methane conversion to chemicals

6.15.3.1.2.2 High temperature methane conversion 6.15.3.1.2.2.1 Oxidative coupling of methane to C2 hydrocarbon (OCM) OCM was first discovered by Keller and Bhasin in 1982 and has been under study for nearly 40 years.75 Methane reacts with oxygen via OCM leading to the formation of C2 hydrocarbons. Extensive studies showed that the acidity and basicity of the catalyst surface play an essential role in the catalytic activity of OCM.76,77 Zavyalova et al. summarized 1850 catalysts reported in more than 1000 papers published since 1982 and compared their activities, as displayed in Fig. 9.78 One sees that almost all catalysts with higher activity exhibit strong basicity, such as Li/MgO, one representative catalyst. Studies demonstrated that the adsorption and dissociation of methane readily happen on the alkaline surface. Therefore, rare earth La metal oxides76,79–84 and alkaline metal oxides77,85–92 are considered to be the most promising catalysts for OCM. In addition, doping of alkali metals (Li, Cs, and Na) and alkaline earth metals (Sr and Ba) can increase the C2 selectivity, while doping of Mn and W can facilitate the conversion. The state of art OCM catalysts give a C2 selectivity up to 72–82%, and a yield of C2 hydrocarbons 16–26%.93 Among the above-mentioned catalysts, Li/MgO, Sr/La2O3, Na2WO4/SiO2 catalysts are mostly studied. Lunsford et al. reported a methane conversion of 37%, C2 yield 20% at 913 K over the Li/MgO catalyst.94 Sollier et al. studied Sr doped La2O3 nanofibers, which contained more basic sites and different oxygen species on the surface.83 It was found to significantly improved the OCM activity and C2 selectivity, giving a C2 yield as high as 20% at 923 K. Furthermore, the catalyst containing 8.6% Sr showed a rather good stability against sintering and coke formation.83 Ji et al. reported the catalyst Mn/Na2WO4/SiO2, which led to methane conversion 20% and C2 selectivity 80%, which also exhibited a rather good stability within a 97 h stability test.95 In addition, interaction between oxide and support,82,83,96–98 concentration of oxygen vacancies,99,100 and morphology of the catalyst have been investigated in this system.80,101 It is noteworthy that Fan and coworkers recently combined La2O3 and Na2WO4/SiO2 as a bifunctional catalyst, which exhibits excellent performance in OCM.102 It was suggested that methane was first activated over La2O3 forming methyl radicals, which subsequently went through C-C coupling on the surface of Na2WO4/SiO2. The authors reported that the presence of 5NaWSi could suppress over-oxidation and consequently the selectivity of C2 was enhanced by three times at 823 K. The yield of C2 hydrocarbons as high as 10.9% was obtained at 823 K, which was the highest reported so far at a reaction temperature below 873 K. In addition to catalysts, the oxidants have been demonstrated to play an important role in OCM, particularly on product selectivity. Gaseous reagents such as N2O, CO2, and S2 have been investigated, intending to weaken the over-oxidation and safety issues considering that they are milder oxidants. Hutchings et al. showed that the total C2 selectivity over a 5%Li/Sm2O3 catalyst was 71.8% using N2O as the oxidant in comparison to about 52.9% using O2 as the oxidant at similar CH4 conversions.103 Langfeld et al. compared the OCM performance of a series of literature Li/MgO and NaWSi based catalysts and catalysts prepared by cellulose templating, as well as the oxidants of N2O and O2.104 As shown in Fig. 10, with NaWSiOx catalyst, C2 selectivity can reach as high as 82% using N2O whereas only 13% using O2 as the oxidant although methane conversion was only  5%. The study of reaction mechanism indicated that N2O can provide only monoatomic oxygen species, such as O species.105 In comparison, using O2 as the oxidant, peroxide species would form during OCM, which readily led to CO2 formation. Sulfur was also studied as the oxidant for OCM (SOCM) by Peter et al.106 It was demonstrated that methane conversion was correlated with the M (metal)eS bond energy. A weaker M-S bond was beneficial for methane activation, but a stronger MeS

Fig. 9

OCM catalysts and their catalytic performance. Zavyalova, U.; Holena, M.; Schlögl, R.; Baerns, M., ChemCatChem 2011, 3 (12), 1935–1947.

Inorganic catalysis for methane conversion to chemicals

337

Fig. 10 Selectivity versus conversion of OCM reaction at 1073 K using O2 (black) and N2O (red) as oxidants. Langfeld, K.; Frank, B.; Strempel, V.E.; Berger-Karin, C.; Weinberg, G., et al., Appl. Catal. A. Gen. 2012, 417, 145–152.

bond can inhibit the deep oxidation process, thereby increasing the ethylene selectivity (Fig. 11). 10% Pd/Fe3O4 can achieve 3%– 4% methane conversion and higher than 30% ethylene selectivity at 1323 K.106 However, it still inevitably produced by-products such as CS2 and H2S. 6.15.3.1.2.2.2 Non-oxidative methane dehydroaromatization to aromatics (MDA) In 1993, Wang et al. reported that Mo/ZSM-5 catalyst can activate CH4 at 973 K giving a conversion of 7–8%.107 Benzene was the major hydrocarbon product. Thereafter, extensive studies were carried out worldwide. After nearly 30 years of development, significant progress has been made in both mechanistic understanding and optimization of catalytic performance.13,108–123 In addition to Mo-based catalysts, other metals such as Fe,121,123,124 Co,125 Ga,125 Cr,126 Zn,125 W,127 V,126 Re,117,128 Cu,129 Mn,119,130, etc., have been studied. Ma et al. compared the catalytic activity of Mo, Zn, W. Re, Co-Ga, Fe, Mn, V, Cr (Table 4), and concluded that the Mobased catalyst showed the highest reaction activity.120 Compared with SOM and OCM reactions, the absence of oxygen in the reaction avoided over-oxidation and the selectivity of hydrocarbons was higher. However, a high temperature is usually required to achieve a higher methane conversion because MDA is an endothermic reaction. Furthermore, zeolites are generally used as the support. These features frequently lead to severe carbon deposition and hence deactivation of catalysts. This has been remaining a major obstacle for the industrialization of MDA reaction. Therefore, extensive efforts have been made to understand the mechanism of carbon deposition, to explore efficient techniques to suppress carbon deposition, and techniques to regenerate catalysts in order to bring this technology closer to commercialization. H2 and O2 have been employed to remove the deposited coke and regenerate the catalyst. O2 can readily burn away the coke, but at the same time led to the migration of Mo species upon extended exposure in O2. The advantage of H2 regeneration was that it could not only restore the activity directly but also brought little damage to the structure of the catalyst after repeated regenerations. Lu et al. compared the regeneration of Mo/HZSM-5 catalyst in O2 at 823 K and in H2 at 1173 K.131 Shu et al. developed a periodically switching strategy between the reactant gas methane and the regeneration gas H2 or CO2 for 30–30 min.132 The results in

Fig. 11 Scheme of SOCM. The inset figure shows the SOCM stability test of 10 wt% Pd/Fe3O4 and Fe3O4 under the conditions of 1223 K, 0.785 h 1 and CH4/S ¼ 7. Peter, M.; Marks, T.J., J. Am. Chem. Soc. 2015, 137 (48), 15234–40.

338

Inorganic catalysis for methane conversion to chemicals

Table 4

Comparison of MDA reaction performance with different metals. Selectivity (%)

Reaction conditions Active metals

Temperature ( C)

Flow (mL gcat 1 h 1)

CH4 conversion (%)

Benzene

Naphthalene

Mo Zn W Re Co-Ga Fe Mn V Cr

730 700 800 750 700 750 800 750 750

1500 1500 1500 1440 1500 800b 1600 800b 800b

16.7 1.0 13.3 9.3 12.8 4.1 6.9 3.2 1.1

60.4 69.9 52.0 52 66.5 73.4 75.6 32.6 72.0

8.1 –a –a 0 7.2 16.1 11.9 6.3 3.7

a

Not reported. GHSV/h 1. Ma, S.; Guo, X.; Zhao, L.; Scott, S.; Bao, X., J. Energy Chem. 2013, 22 (1), 1–20. b

Fig. 12 indicated that the regeneration of H2 was better than that of CO2 as far as the benzene formation rate was concerned. Sun et al. demonstrated a 1000 h stability test with periodic switching of CH4:H2 ¼ 15 min:45 min using pure hydrogen as the regeneration gas.111 Methane conversion was maintained at 15–19%, and the yield of benzene was higher than 12% (Fig. 13).111 Since MDA is a thermodynamically limited reaction, a membrane reactor was proposed in order to enhance methane conversion.133,134 For example, Morejudo et al. combined a perovskite BaZr0.7Ce0.2Y0.1O3  x oxygen-permeable membrane reactor with Mo/H-MCM-22 catalyst, as displayed in Fig. 14A.135 Oxide ions were transported to the reaction side and reacted with H2 forming steam, which further reacted with coke forming CO and H2. Thus, the carbon deposits were successfully eliminated by the water-gasshift reaction and consequently improved the carbon atom utilization efficiency. Fig. 14B and C show that the yield of aromatics was maintained over 10%, which was much more stable than that in the conventional fixed bed reactor.135 Kosinov et al. demonstrated that the reaction pressure can also significantly impact the reaction, particularly, the coke selectivity, as shown in Fig. 15.8 For example, the space-time yield of aromatics on 2% Mo/ZSM-5 was improved by nearly an order of magnitude by raising the reaction pressure to 15 bar and at the same time the coke selectivity decreased from 40% to 10%.8 Interestingly, the improvement of the catalytic performance at increased pressure was independent of the Mo loading, reaction temperature and methane space velocity. Operando X-ray absorption spectroscopy study evidenced that the Mo-oxide precursor was easier to reduce under elevated pressures and the structure of the active sites remained independent of reaction pressure. The improved performance can be attributed to faster coke hydrogenation at increased pressure.8 Bifunctional catalytic mechanism is widely accepted for MDA. Methane first dehydrogenates and couples to form C2Hx intermediates on the molybdenum carbide active sites. Then, the C2Hx intermediates go through aromatization forming aromatics on Brönsted acid sites of zeolite supports.13,112 Lunsford et al. discovered the induction period in the MDA reaction. During the induction period, MoO3 was first carbonized forming Mo2C as the active sites in both the inner and outer surface of the zeolite.126 Ma et al. studied the induction period with TPSR and 1H MAS NMR and revealed a three-step process.11 In the range of 820–960 K, methane was mainly converted to CO2 and H2O. With the temperature increasing to the range of 960–1050 K, the consumption of methane increased sharply, producing CO2, H2O, H2 and CO. When the temperature reached above 1050 K, benzene became the dominant product.11 In addition to the knowledge of Mo2C as the active site, Zheng et al. reported that the partially reduced and carbonized MoCxOy species on the Brönsted acid site were the active center for MDA reaction, evidenced by the high field 95Mo NMR spectroscopy.108 As shown in Fig. 16A, static 95Mo NMR spectra of Mo/HSZM-5 were well fitted and two types of molybdenum were suggested, namely MoO3 (purple) and exchanged Mo species (green) on zeolite. Fig. 16B suggests a good correlation between the amount of the exchanged molybdenum species and aromatics formation rate.108 Gonzalez et al. studied the evolution of Mo/ZSM-5 catalysts during the MDA reaction using operando Mo K-edge HERFD-XANES, XRD, and XES.113 The results indicated that isolated MoOx species formed after high-temperature calcination and were subsequently reduced to MoCxOy by methane, which were proposed as the active sites.113 While the bifunctional mechanism is more generally accepted for MDA over Mo/zeolite, a mechanism of monofunctional mechanism was also proposed. Mériaudeau et al. concluded that MDA did not go via bifunctional mechanism because the rate of benzene formation starting from C2H4/H2/N2 was higher on Mo/H-ZSM-5 compared to H-ZSM-5.136 Kosinov et al. further reported that the aromatization was an intrinsic ability of molybdenum carbides dispersed in the 10-membered-ring micropores of MFI zeolite.137 The authors concluded that Brönsted acid sites was able to promote the well dispersion of the Mo oxide precursor and, accordingly, the formation of the active Mo carbide phase in the micropores of HZSM-5.137 The above studies show that

Benzene formation rate / nmol g-cat–1 s–1

Inorganic catalysis for methane conversion to chemicals

339

12000 (A) 12600 ml g–1h–1, CH4 30min-H2 30min

10000 9000 ml g–1h–1, CH4 30min-H2 30min

8000

6000 9000 ml g–1h–1, without switching treatment

4000 2700 ml g–1h–1, CH4 30min-H2 30min

2000 2700 ml g–1h–1, without switching treatment

0 60

120

180

240

300

360

Hydrogen formation rate / nmol g-cat–1 s–1

Time on stream / min 25000

(B) 12600 ml g–1h–1, CH4 30min-H2 30min

20000 9000 ml g–1h–1, CH4 30min-H2 30min

15000

9000 ml g–1h–1, without switching treatment

10000

2700 ml g–1h–1, CH4 30min-H2 30min

5000 2700 ml g–1h–1, without switching treatment

0 60

120

180

240

300

360

Time on stream / min Fig. 12 MDA reaction under different modes of periodic switching of reactant and regeneration feed gas. Shu, Y. Y.; Ma, H. T.; Ohnishi, R.; Ichikawa, M., Chem. Commun. 2003, (1), 86–87.

the mono- or bi-functional mechanism, the active sites either MoOxCy or MoCx species, as well as the mechanism for the formation of primary C2 intermediates still need more sophisticated investigation.

340

Inorganic catalysis for methane conversion to chemicals

Fig. 13 1000 h stability test over Mo/ZSM-5 catalyst under1033–1073 K, tCH4: tH2 ¼ 15 min: 45 min. Sun, C.; Fang, G.; Guo, X.; Hu, Y.; Ma, S., et al., J. Energy Chem. 2015, 24 (3), 257–263.

Fig. 14 Current-controlled co-ionic membrane reactor. (A) The scheme of whole process. (B) Aromatics yield versus time. (C) Deactivation rate constant as a function of H2 extracted and O2 injected. Morejudo, S. H.; Zanon, R.; Escolastico, S.; Yuste-Tirados, I.; Malerod-Fjeld, H., et al., Science 2016, 353 (6299), 563–6.

Inorganic catalysis for methane conversion to chemicals

341

Fig. 15 The hydrocarbon selectivity and methane conversion rate as a function of pressure. Reaction conditions: 973 K, GHSV 15000 h 1, 900 min. Kosinov, N.; Uslamin, E. A.; Meng, L.; Parastaev, A.; Liu, Y., et al., Angew. Chem. Int. Ed. Engl. 2019, 58 (21), 7068–7072.

6.15.3.1.2.2.3 Non-oxidative methane conversion to olefins, aromatics and hydrogen (MTOAH) Guo et al. reported a silica lattice confined single-site iron (FeÓSiO2) catalyst can activate methane directly and intensive carbon deposition can be avoided (Fig. 17).138 The reaction gave mainly ethylene, benzene and naphthalene as products. In situ characterization of the catalyst with in situ XAFS and DFT calculation indicated that the catalyst was activated first forming a single site structure with iron bonded with two carbon atoms and one silicon atom (Si-Fe-2C). Thus, methane was activated forming methyl groups. The evolution of the catalytic active sites was later validated by Liu et al.139 Since there were no adjacent Fe sites, these methyl groups did not go through catalytic C-C coupling on the surface but desorbed into the gas phase to generate methyl radicals, which were experimentally detected by in situ vacuum ultraviolet single-photon dissociation mass spectrometry.138 These methyl radicals then went through C-C coupling in the gas phase, forming thermodynamically stable products of ethylene, benzene and

95

Mo NMR Spectrum of Mo/zeolite catalyst at 21.1 T

(A)

C6H6 + H2

CH4

2000

1000

0 (ppm)

–1000 –2000

(B)

Aromatics Formation Rate (nmol/gcat/s)

180 150 120 90 60

Total Mo Species MoO3 Crystallites

30

Exchanged Mo Species 0

0

150

300

450

600

750

900

Amount of Mo Species (umol/gcat) Fig. 16 (A) Direct observation of the active center for methane dehydroaromatization using an ultrahigh field 95Mo NMR spectroscopy. (B) Correlating the aromatics formation rate with different molybdenum species. Zheng, H.; Ma, D.; Bao, X.; Hu, J. Z.; Kwak, J. H., et al., J. Am. Chem. Soc. 2008, 130 (12), 3722–3.

Inorganic catalysis for methane conversion to chemicals

25

100 Ethylene Ethane Benzene Naphthalene Coke

Products selectivity (%)

90 80 70

20

15

60 50

10

40 30

5

20

Methane conversion (%)

342

10 0

0 Blank

0.5Fe/ZSM

0.5Fe/SiO2

0.8Fe/SiO2 0.5Fe©SiO2

0.2Fe/SiC

Catalysts Fig. 17 Direct conversion of methane to olefins, aromatics and hydrogen over Fe©SiO2. Guo, X.; Fang, G.; Li, G.; Ma, H.; Fan, H., et al., Science 2014, 344 (6184), 616–9.

naphthalene. At 1223 K, CH4 conversion was 8.1%, while selectivity of ethylene, benzene and naphthalene was 48%, 26% and 26%, respectively.138 Sakbodin et al. validated the catalytic activity of FeÓSiO2 single sites for methane activation and conversion.140 They further combined the catalyst with a tubular SrCe0.7Zr0.2Eu0.1O3  d hydrogen-permeable membrane, intending to remove the produced hydrogen out of the reaction system.140 Consequently, methane conversion was significantly enhanced according to Le Chatelier’s principle. A 60 h stability test indicated that the membrane reactor was rather stable, i.e., methane conversion was maintained at about 20%, the total selectivity of C2 hydrocarbons, benzene and naphthalene exceeded 90%.140 Coupling of membrane separation with catalytic reactions provides an elegant method to enhance the thermodynamically limited reactions by removing one of the products out of the reaction system. Han et al. embedded iron species into cristobalite, which was active for MTOAH.141 It gave 6.9– 5.8% methane conversion and 86.2% C2 selectivity within a 100 h stability test with cofeeding of 50% H2 at 1353 K. EXAFS showed that highly dispersed Fe carbide with FeeSi coordination was formed in the catalyst. Combined with theoretical calculations, the authors showed that these lattice confined highly dispersed Fe sites were more favorable for methyl radical formation and a high coke resistance than Fe3C clusters.141 In addition to FeÓSiO2 catalyst, Xie et al. reported that CeO2 confined single-site Pt catalyst was active for MTOAH reaction.142 It exhibited a superior catalytic activity with 14.4% methane conversion and 74.6% selectivity

Fig. 18 Scheme of millisecond catalytic wall reactor. Oh, S. C.; Schulman, E.; Zhang, J.; Fan, J.; Pan, Y., et al. Angew. Chem. Int. Ed. Engl. 2019, 58 (21), 7083-7086.

Inorganic catalysis for methane conversion to chemicals

343

Fig. 19 Detection of H radicals generated during MTOAH using the H atom Rydberg Tagging Time-of-Flight spectroscopy and its role in MTOAH. (A) Concentration of H radicals as a function of the reaction temperature; (B) rate of methane conversion as a function of H radical production rate. Hao, J.; Schwach, P.; Fang, G.; Guo, X.; Zhang, H., et al. ACS Catal 2019, 9 (10), 9045–9050. Hao, J. Q.; Schwach, P.; Li, L. L.; Guo, X. G.; Weng, J. B., et al. J Energ Chem 2021, 52, 372-376.

toward C2 products (ethane, ethylene, and acetylene) at 1148 K. Moreover, a rather good stability was also observed within a 40 h test.142 Liu and coworkers further developed a millisecond catalytic wall reactor with FeÓSiO2 catalyst coated on the inner side of the quartz reactor (Fig. 18).143 By carrying out the process simulation by Aspen Plus, they showed that the energy input was estimated to be reduced by six times by appropriate process optimization, e.g., by coupling the endothermic MTOAH on one side with the exothermic combustion of coke on the other side. Thus, high carbon and thermal efficiencies and low cost of the reactor materials can be realized. Therefore, this technology could be techno-economically viable.143 Hao et al. also studied methane conversion to olefins and aromatics in the FeÓSiO2 coated reactor.144,145 By employing the H atom Rydberg Tagging Time-of-Flight technique, direct experimental evidence was observed for the formation of hydrogen radicals during MTOAH reaction over a catalytic quartz wall reactor containing embedded iron species (Fig. 19A).145 It was further demonstrated that methane conversion can be enhanced by hydrogen radicals, provided by the thermal decomposition of hydrogen-donor molecules, such as 1,2,3,4tetrahydronaphthalene (THN) and benzene. The methane conversion rate was observed to be correlated with the rate of hydrogen radical production (Fig. 19B). For instance, methane conversion was enhanced to 25.5% from 19.7% in the presence of 1.41% THN in methane at 1323 K.144 Huang et al. recently carried out an economic assessment study of MTOAH by developing a flexible approach that allowed evaluation of the impact of various process variables on the economics of the process (Fig. 20).146 They showed that methane one-pass conversion impacted most significantly on the overall process economics and the reaction temperature had a small impact because the excessive heat can be recovered for power generation via process design. They concluded that MTOAH would be economically feasible only if coke formation can be restricted to be lower than 20% and a minimum conversion to products of 25% was obtained.146 Therefore, this process exhibits a good potential for applications upon further optimization of the catalyst and the process.

6.15.3.2 6.15.3.2.1

Electrocatalytic conversion of methane Introduction

Electrocatalytic conversion of methane to valuable chemicals and fuels can be driven by electricity derived from renewable energy sources such as solar energy and wind power, achieving sustainable methane conversion and utilization (Fig. 21).147–149 Electrocatalysis deals with electrochemical reactions in the interconversion of chemical energy and electrical energy, and may overcome some limitations of thermos-catalytic conversion of methane. When coupled with oxygen reduction reaction at the cathode in a fuel-cell-like reactor, electricity can be also produced as a byproduct from the electrocatalytic conversion of methane.150 The activity and selectivity of methane conversion at low temperatures (< 373 K) can be systemically tuned by controlling the chemical potential of the catalyst surface by varying the electrode potentials.151 The oxidant in this approach is the active oxygen species in situ generated from water oxidation under reaction conditions, free of external strong oxidants. Taking methane conversion to methanol for example (Eqs. 10–13), using water as the oxidant the partial oxidation of methane to methanol is 9 kJ/mol higher than that of complete oxidation to CO2 while the complete oxidation of methane is much more favorable in the case of molecular oxygen as the oxidant.148 The thermodynamic analysis indicates that using water as the oxidant under mild reaction conditions may avoid the overoxidation of products. Moreover, water can further stabilize the formed oxygenate products via solvation effects.152

344

Inorganic catalysis for methane conversion to chemicals

Fig. 20 Impact of key process variables on the economics of MTOAH process referred to the net present value (NPV). (A) Reaction temperature; (B) hydrocarbon selectivity/coke selectivity; (C) ethylene selectivity in hydrocarbons. Huang, K.; Miller, J. B.; Huber, G. W.; Dumesic, J. A.; Maravelias, C. T., Joule 2018, 2 (2), 349–365.

Inorganic catalysis for methane conversion to chemicals

345

These potential advantages make electrocatalytic conversion of methane to chemicals and fuels an interesting alternative to be explored for methane utilization. CH4 ðgÞ þ 1=2O2 ðgÞ/CH3 OHðlÞDG0 ¼  115 kJ mol1

(10)

CH4 ðgÞ þ 2O2 ðgÞ/CO2 ðgÞ þ 2H2 OðlÞDG0 ¼  818 kJ mol1

(11)

CH4 ðgÞ þ H2 OðlÞ/CH3 OHðlÞ þ H2 ð g ÞDG0 ¼ 121 kJ mol1

(12)

CH4 ðgÞ þ 2H2 OðgÞ/CO2 ðgÞ þ 4H2 ðgÞDG0 ¼ 130 kJ mol1

(13)

Early studies on electrocatalytic conversion of methane focused on the direct methane fuel cells using Pt and Pd as anode catalysts with CO2 as the major product.153–155 Later on, a high methanol selectivity (88%) was achieved in such fuel cell systems using a V2O5/SnO2 anode that is more inert than Pt.150 Recently researchers tried to conduct the reaction in the electrolysis mode, but it suffered from a low yield of oxygenates due to the low solubility of methane in aqueous electrolytes.156,157 While a large number of catalysts and electrochemical devices have been discovered,149,154,158–161 currently electrocatalytic conversion of methane is still in its early stage. Techno-economic analysis of a methane-to-methanol fuel cell mode indicated the minimum performance metrics for the economic viability of the process, namely, 70% methanol selectivity and  0.5 V cell voltage at a current density of 0.5 A cm 2 in the temperature range of 373–423 K.162 The state-of-the-art performance is far below this requirement but is anticipated to be promising upon further optimization. Therefore, we introduce the fundamentals of electrocatalytic conversion of methane, and then overview the recent progress in electrocatalytic conversion of methane in the following section. Photoelectrocatalytic conversion of methane, an emerging direction is also discussed briefly.

6.15.3.2.2

Fundamentals of electrocatalytic conversion of methane

Table 5 shows some possible reactions related to electrocatalytic conversions of methane in aqueous solution and the related redox potentials versus reversible hydrogen electrode (RHE) at 298 K.149 The valuable partial oxidation products include methanol,

Table 5

Equations and equilibrium potentials of electrocatalytic conversion of methane.

Anode reaction

E (V vs. RHE)

CH4(g) þ H2O(l) / CH3OH(a) þ 2Hþ(a) þ 2e CH4(g) þ H2O(l) / HCHO(a) þ 4Hþ(a) þ 4e CH4(g) þ 2H2O(l) / HCOOH(a) þ 6Hþ(a) þ 6e CH4(g) þ H2O(l) / CO(g) þ 6Hþ(a) þ 6e CH4(g) þ 2H2O(l) / CO2(g) þ 8Hþ(a) þ 8e 2CH4(g) / C2H6(g) þ 2Hþ(a) þ 2e 2CH4(g) / C2H4(g) þ 4Hþ(a) þ 4e

0.58 0.46 0.26 0.26 0.17 0.35 0.44

Xie, S.; Lin, S.; Zhang, Q.; Tian, Z.; Wang, Y., J. Energy Chem. 2018, 27 (6), 1629–1636.

Fig. 21 Renewable energy-powered selective electrocatalytic conversion of methane to liquid fuels and chemicals. Xie, S.; Lin, S.; Zhang, Q.; Tian, Z.; Wang, Y., J. Energy Chem. 2018, 27 (6), 1629–1636.

346

Inorganic catalysis for methane conversion to chemicals

(A)

(B)

O2

CH4 e–

e–

e–

e– H2

Cathode

Anode

Cathode

Electrolyte

H2O

CH4

Anode

Electrolyte

CH3OH H 2O

Alcohols

Fig. 22 Reactors for electrocatalytic conversion of methane: (A) fuel cell mode, (B) electrolysis mode. (A) Yuan, S.; Li, Y.; Peng, J.; QuestellSantiago, Y. M.; Akkiraju, K., et al., Adv. Energy Mater. 2020, 10 (40), 2002154, (B) Ma, M.; Jin, B. J.; Li, P.; Jung, M. S.; Kim, J. I., et al., Adv. Sci. (Weinheim, Ger.) 2017, 4 (12), 1700379.

formaldehyde, formic acid, ethylene, and ethane. Since all the products are more reactive than methane, the reaction conditions that activate methane easily cause overoxidation to form the most stable product, CO2. Protecting products from overoxidation is the key to improve selectivity at a higher methane conversion. The C-H dissociation of methane is a critical step in methane activation and oxidation, and its mechanism can be either dehydrogenation or deprotonation.151 In the dehydrogenation mechanism, a hydrogen atom ($ H) is abstracted from CH4 by electrophilic oxygen atoms (M-O sites) to generate $ CH3 and $ OH radicals in the transition state. The $ CH3 radicals weakly interact with catalytic centers through $ OH sites. The deprotonation mechanism starts by the dissociative adsorption of CH4 to form þ  þ CH 3 and H . The CH3 coordinates to catalyst surface to generate an M-C s-bond while the H is accepted by metal and O species on the surfaces.163 The active oxygen species are either electrophilic or nucleophilic. Electrophilic oxygen species (O, O 2 ) are electron-deficient, which are generated by strong oxidants (O2, N2O) and can promote the dehydrogenation of CeH bonds. Nucleophilic oxygen is electron saturated species (O2), which often act as Hþ acceptors and catalyze the deprotonation of CeH bonds.164 Electrocatalytic methane oxidation is driven by an anodic potential, which can be conducted in either fuel cell mode (Fig. 22A) or electrolysis mode (Fig. 22B). The former mode originates from solid electrolyte direct methane fuel cells in which methane and oxygen are supplied to anode and cathode, respectively, and methane is combusted to produce CO2 and power. With appropriate anode catalysts that are not very active for the complete oxidation of methane, it is able to achieve the partial oxidation of methane to oxygenates and C2 hydrocarbons.150,165 In the case of electrolysis mode, hydrogen evolution via water reduction serves as a counterreaction, instead of oxygen reduction reaction. The electrolysis mode is mainly carried out in aqueous electrolyte solutions. The electrolysis current density is often restricted by the very low solubility of methane in water ( 1.5 mM at 298 K, 1 atm166), in addition to the selectivity issue caused by overoxidation. This issue could be overcome by applying continuous-flow electrolyzers with gas diffusion electrode (GDE) configuration developed recently in other electrocatalytic reactions.167 The reactor configuration has remarkable effects on electrocatalytic methane oxidation. We will discuss recent advances in electrocatalytic conversion of methane in terms of specific products, as listed in Table 6.

6.15.3.2.3

Electrocatalytic conversion of methane to fuels and chemicals

The selective partial oxidation of methane to desired products such as hydrocarbons, alcohols, acids and ketones is very challenging, as all these products are facile to be overoxidized to CO2. Many efforts have been devoted to this transformation in past decades. 6.15.3.2.3.1 Electrocatalytic conversion of methane to ethylene and ethane Tang et al. reported a solid oxide fuel cells (SOFCs) tubular membrane reactor integrated with a Mn-Ce-Na2WO4/SiO2 catalyst for OCM.165 It exhibited a high methane conversion of 60.7%, with a C2 þ selectivity of 41.6%, an ethylene-to-ethane ratio of 5.8 and an ethylene yield of 19.4%. The catalyst showed significant advantages over conventional OCM process. Xie et al. designed a solid

Inorganic catalysis for methane conversion to chemicals Table 6

347

Performance of electrocatalytic conversion of methane over metal and metal oxide catalysts.

Major product

Catalyst

Reaction conditions

Methanol

V2O5/TiO2/RuO2/PTFE gas diffusion electrodes

Single Yield 297 mg/L/cm2, Faradaic compartment efficiency 57% cell Dry CH4 at anode at 50% Humidified CH4 at anode at 373 K Fuel cell Faradaic efficiency 61%, selectivity 88.4% CH4 at anode, 2.2 V vs. SCE, Single Yield 220 mg/L/cm2, Faradaic 0.1 M Na2SO4 compartment efficiency 22% cell 1%H2O, 1%O2, 10%CH4 at anode, Micro/nano Selectivity 100% 623 K, 0.5 bar SOFC CH4 at anode, 1.5 V vs. Pt, Two-electrode Yield 752.9 mmol/gcat/h, 0.5 M Na2CO3 for 6 h Selectivity 79% system CH4 at anode, 2.3 V vs. RHE, H-cell Yield 364 mmol/gcat/h, CH3Cl/ 0.5 M Na2SO4 CO2 selectivity >400 Diluted CH4 at anode at 1073 K SOFC Yield 19.4%, C2 selectivity 41.6% Dry CH4 at anode at 1123 K SOEC Yield 12.1%, C2 selectivity 81.2% CH4 at anode, 1.40 V vs. RHE, H-cell Faradaic efficiency 89%, Yield 1 1 h 0.1 M NaOH 25 mmol$gNiO CH4 at anode, 2.0 V vs. Pt, 0.5 M Two-electrode Carbon efficiency 60% Na2CO3 for 12 h system CH4 at anode, 2.0 V vs. Pt, 0.5 M Two-electrode Carbon efficiency 47.5% Na2CO3 for 20 h system

GDC-Li2CO3/La-SrTiO3 V2O5/SnO2 (TiO2/RuO2)/PTFE Pd-Au-Cu-C CuO/CeO2 CH3Cl

CoNi2Ox

Ethylene Ethane

Mn-Ce-Na2WO4/SiO2 Sr2Fe1.5Mo0.5O6-d

Ethanol

NiO/Ni

Propanol

Co3O4/ZrO2

Propionic acid, acetone, ZrO2:NiCo2O4 acetic acid

Dry CH4 at anode, 2.0 V, 0.2 M Na2SO4

Reactor

Performance

References 168 169 150 170 171 172 173 165 174 156 157 175

oxide electrolyzer by combining methane oxidation as an anode reaction and CO2 reduction as a cathode reaction.174 The O2 ions extracted from CO2 transfer through the oxide-conducting electrolyte to anode and thus oxidize methane. It is able to achieve simultaneous conversion of both methane to ethylene and CO2 to CO in a single cell. They in situ constructed metal-oxide interfaces on porous redox-reversible layered perovskite Sr2Fe1.5Mo0.5O6  d (SFMO) scaffold and tailored the amount of excess Fe species to grow Fe nanoparticles on the SFMO scaffold via phase decomposition in reducing atmospheres. The highest C2 product selectivity reached 81.2% and the highest C2 product concentration in output gas reached 16.7% (12.1% ethylene and 4.6% ethane) in the outlet gas at ambient pressure and 1123 K. A high methane conversion as high as 41% was achieved in the initial pass, and no obvious degradation was observed after 100 h operation and 10 redox cycles at high temperature (Fig. 23).

Fig. 23 (A) Product distribution in anode outlet gas. (B) The long-term stability of 0.075Fe-SFMO-SDC electrode for methane oxidation coupled with CO2 electrolysis at 1123 K. Zhu, C.; Hou, S.; Hu, X.; Lu, J.; Chen, F., et al., Nat. Commun. 2019, 10 (1), 1173.

Inorganic catalysis for methane conversion to chemicals

(A)

0.4

0.01

Electrochemical activation (200mA)

0.3

100 Sn0.9In0.1P2O7+Pd-Au-Cu/C Pd-Au-Cu/C (only yield)

Methanol yield / %

Catalytic activation (4.5% hydrogen)

R(CH3OH) 0.2 /µmol h–1 cm–2 0.1

0.0075

80

Sn0.91n0. 1P2O7 (only yield)

60 0.005 40 0.0025

20

Methanol Selectivity / %

348

0 50

100

150

200

250

0

Temperature / °C

(C) 0.05

0 0

(D)

200 300 400 Temperature / °C

400°C 20

0.8 wt%

Concentration (%)

0.95 wt% 1.0 wt%

0.03

1.05 wt% 1.2 wt%

0.02

1.5 wt% 100 wt%

350°C

15

10 600°C

0.01

0 –800

600

300°C

CH3OH

0.5wt%

0.04

500

25

0 wt%

CH3OH concentration / %

100

5 CO

–400

0 800 1200 400 Electrode potential / mV

1600

0 0

5

10

15

20

25

Time (min)

Fig. 24 Methanol production in SOFC over: (A) PdAu/C, (B) Sn0.9In0.1P2O7 mixed with PdAuCu/C catalyst, (C) V2O5/SnO2 with different V2O5 contents, and (D) NiO/La-doped SrTiO3. Mohamed, A. G. A.; Naqviab, S. A. Z.; Wang, Y. B., ChemCatChem 2021, 13 (3), 787–805.

6.15.3.2.3.2 Electrocatalytic conversion of methane to methanol Most of the studies on the electrocatalytic conversion of methane focus on the selective production of methanol.148,151,159 Since methanol is not stable at high temperature, the reactors for producing methanol usually work at lower temperatures. The Hibino group did a variety of works using proton-conducting solid oxide fuel cells (P-SOFCs) for low/intermediate temperatures conditions. At the cathode, oxygen reduction reaction could produce oxygen species which is ready to react with methane toward methanol generation. With this idea, Hibino et al. fed a mixture of oxygen and methane to the cathode of a hydrogen-oxygen fuel cell between 323 and 523 K using Sn0.9In0.1P2O7 proton conductor as electrolyte.176 The highest formation rate of methanol reached 0.38 mmol h 1 cm 2 over a PdAu/C catalyst at an applied current of 400 mA cm 2 (523 K). The methanol selectivity increased as the temperature decreased, e.g., 6.03% at methane conversion 0.38% at 523 K; 10.8% at methane conversion 0.14% at 423 K, and 60% at 323 K but at the expenses of further reduced methane conversion 0.012% (Fig. 24A). The PdAu/C catalyst was further modified by adding Cu and mixed with proton-conducting Sn0.9In0.1P2O7 particles to provide electrolyte–electrode interfaces as reaction sites for methane oxidation. Meanwhile, the cathode feed gasses changed from hydrogen to water vapor.171 With such a catalytic system, the methanol selectivity was improved to around 100% at 523–623 K (Fig. 24B), but the yield of methanol was only 0.003% at 573 K and the Faradaic efficiency of methanol was only 0.013%. This low yield of methanol was attributed to the fact that oxygen reduction was the predominant reaction. Interestingly, when the mixture of methane and water vapor were fed to anode and air to the cathode, a V2O5/SnO2 anode was able to drive the production of methanol significantly at 373 K (Fig. 24C) leading to a Faradaic efficiency as high as 61.4% and the methanol selectivity 88.4%.150 Characterization indicated that the active oxygen species formed over V2O5 were key for the efficient conversion of methane to methanol. Recently, Torabi et al.169 demonstrated oxide-conducting solid oxide fuel cells (O-SOFCs) for methanol generation. The methanol selectivity was about 90% over NiO/La-doped SrTiO3 perovskite anode using a dense GDC-Li2CO3 electrolyte at 573–673 K (Fig. 24D). The selective oxidation of methane to methanol was also conducted in electrolysis mode in liquid electrolytes. While the mass transport was limited by the low solubility of methane in water, in some cases the electrolyte can play a role as an oxidizing agent. For example, Spinner et al.177 reported a NiO-ZrO2 bifunctional catalyst for direct oxygen species delivery using CO32 as an oxygen-donating species through ZrO2 parallel to methane activation on NiO. Methane was adsorbed and activated by NiO while CO32 was adsorbed onto non-conducting ZrO2. Oxygen was then abstracted and donated from CO32 to electrocatalytically active

Inorganic catalysis for methane conversion to chemicals

349

Fig. 25 (A) Oxygenate/methanol production rates and (B) oxygenate selectivity over CuO/CeO2 catalysts with various compositions. (C) Catalytic cycle for electrophilic methane oxidation using stoichiometric oxidant YO and redox potentials of methane oxidation catalysts. (D) Proposed mechanism for electrochemical methane functionalization by a putative Pd2III, III intermediate. (A and B) Lee, J.; Yang, J.; Moon, J. H., ACS Energy Lett. 2021, 6 (3), 893–899, (C and D) O’Reilly, M. E.; Kim, R. S.; Oh, S.; Surendranath, Y., ACS Cent. Sci. 2017, 3 (11), 1174–1179.

sites to form new CeO or OeH bonds when using 0.1 M Na2CO3 electrolyte. This proposed mechanism may provide a new pathway for the synthesis of complex hydrocarbons and oxygenates. Following this idea, Moon et al.172 demonstrated that CuO/CeO2 as catalyst for the room-temperature conversion of methane to methanol in the presence of CO32. The highest production rate of methanol 1 1 h (6 h reaction) at ambient pressure and the methanol selectivity was 79% over CuO/CeO2 (Cu:Ce ¼ 6:4) was 752.9 mmol gcat among the oxygenates (Fig. 25A and B). Isotopic labeling experiments confirmed that methane was activated by active oxygen species formed on catalysts surface directly from CO32. By integrating the reaction with a solar cell, they achieved a methanol production of 1 1 at ambient pressure and 21,986.6 mmol gcat at 10 bar over a course of 12 h. 7165 mmol gcat Methane can be activated by electrophiles in concentrated acidic media like CF3COOH and H2SO4. The electrophilic ions can mediate two-electron oxidation of methane to methyl bisulfate (CH3OSO3H), which was protected from over-oxidation by the electron-withdrawing effect of CF3COO and HSO4 groups. These methyl esters can be hydrolyzed to methanol. Surendranath et al.178 developed a versatile new electrochemical oxidation strategy for accessing electrophilic high-valent metal ions. A Pd2III,III species electrogenerated by electrochemically oxidizing PdSO4 in concentrated H2SO4 can activate methane to generate methanol precursors methyl bisulfate (CH3OSO3H) and methanesulfonic acid (CH3SO3H) through concurrent faradaic and nonfaradaic reaction pathways (Fig. 25C and D). Very recently, Zheng et al.173 developed an efficient methane conversion approach using intermediate chlorine species that were electrochemically generated and stabilized on mixed Co-Ni spinels. The CoNi2Ox catalyst 1 1 exhibited a high CH3Cl yield of 364 mmol gcat h and a high CH3Cl/CO2 selectivity of more than 400. GDE is also applied to the electrocatalytic conversion of methane to methanol, in order to overcome the mass transport of methane in water. Bertazzoli et al.170 designed a (TiO2/RuO2)/PTFE GDE in the methane conversion to methanol using 0.1 M

350

Inorganic catalysis for methane conversion to chemicals

Na2SO4 as the supporting electrolyte. The methanol concentration achieved 220 mg L 1 cm 2 at 2.2 V vs. SCE and the methanol Faradaic efficiency was 30%. Further adding V2O5 to the composition of GDE promoted the methanol selectivity at low current densities while the formation of formic acid and formaldehyde was suppressed, thus allowing for higher Faradaic efficiencies, up to 57% at 2.0 V vs. SCE.168 6.15.3.2.3.3 Electrocatalytic conversion of methane to C2 þ oxygenates In addition to methanol, the partial oxidation of methane in the liquid electrolyte is also possible to produce higher oxygenates, including alcohols, acids and ketones. Metal oxide or mixed metal oxide catalysts are usually used for the generation of these C2 þ products.156,157,175,179 Sun et al.156 reported a NiO/Ni interface constructed by calcination as the active site for the electrooxidation of methane to ethanol. The optimized NiO/Ni interface catalyst achieved an 89% Faradaic efficiency for ethanol production with a yield of 25 mmol gNiO 1 h 1 at 1.40 V vs. RHE. Theoretical calculations demonstrated that the NiO/Ni interface was the active site for the initial C-H activation and the subsequent C-C coupling. Park et al.157 designed a Co3O4/ZrO2 nanocomposite for the electrochemical oxidation of methane using Na2CO3 as an electrolyte at room temperature. The generation of 2-propanol and 1-propanol over the given catalyst was achieved with a production efficiency of over 60%. The formation of stable products, such as higher alcohols was considered due to the strong surface adsorption ability of Co3O4 and the participation of carbonate delivered by ZrO2. The same group further modified the Co3O4/ZrO2 with Ni cations to form a ZrO2:NiCo2O4 quasi-solid solution catalyst.175 The integrated Ni atoms were considered to lower the methane activation barrier and thus facilitate the formation of methyl radical species. Over this catalyst, methane was oxidized to propionic acid, acetic acid, and acetone with 47.5% conversion. These products may derive from the further oxidation of the intermediate products from partial methane oxidation, 1-propanol, acetaldehyde, and 2-propanol.

6.15.3.2.4

Photoelectrocatalytic conversion of methane

Photoelectrochemical reactions occur at the semiconductor photoelectrode/electrolyte interface under light irradiation with an applied bias, combining the advantages of electrocatalysis and photocatalysis. The externally applied potential does not directly act on reactants, but creates an electric field to facilitate charge separation between photogenerated electrons and holes, thus significantly increasing the rates of photocatalytic reactions. As for methane conversion, the photoelectrocatalytic conditions may suppress the overoxidation issue since a much less positive applied potential is needed to drive the methane activation process compared to electrocatalytic conditions. Wang et al.180 demonstrated the highly selective production of CO by methane oxidation using a photoelectrocatalytic approach (Fig. 26). The anode substrate they used was TiO2 grown by an atomic layer deposition (ALD) method, which has a higher concentration of Ti3þ species than P25 or commercial anatase TiO2 substrates. The Faradaic efficiency and selectivity of CO strongly depend on the applied potentials. Under the optimized reaction conditions, the highest CO Faradaic efficiency was 50% at 0.4 V vs. RHE, with a carbon selectivity of 81.9% (Fig. 26). Amano et al.181 conducted a proof-ofconcept work studying the homocoupling of methane to ethane in the gas phase at room temperature under visible light irradiation. The used photoelectrochemical membrane flow reactor was equipped with an optical window to allow visible light irradiation and a tungsten trioxide (WO3) gas-diffusion photoanode coated with a proton-conducting ionomer. In such a system, methane was efficiently converted to ethane with selectivity of 54% and incident photon-to-current conversion efficiency of 11% under blue light at an applied voltage of 1.2 V. However, the methane conversion is low, only 0.1% and might be increased by improving the accessibility of methane on the photoanode surface. Recently, with WO3 photoanode, Xiong et al.182 further upgraded methane to more complex ethylene glycol via the photoelectrocatalytic conversion approach. The efficient production of ethylene glycol was achieved by tuning the reactivity of hydroxyl radicals ($ OH) over WO3 with specific facets. Theoretical simulation and radical trapping test indicated the highest reactivity of surface-bound $ OH on (010) facets. Performance measurement showed that the WO3 with the

Fig. 26 (Left) Schematic illustration of selective methane oxidation to CO on a TiO2 photoanode. (Right) CO efficiency and selectivity as a function of applied potential. Li, W.; He, D.; Hu, G.; Li, X.; Banerjee, G., et al., ACS Cent. Sci. 2018, 4 (5), 631–637.

Inorganic catalysis for methane conversion to chemicals

351

highest (010) facet ratio reached an ethylene glycol production rate of 0.47 mmol cm 2 h 1. In situ diffuse reflectance infrared Fourier-transform spectroscopy results confirmed that the methanol, which could be attacked by $ OH to form hydroxymethyl radicals, was the main intermediate for the production of ethylene glycol. These examples have shown great promise of the photoelectrocatalytic approach in methane conversion, but many efforts are still needed in the future to make this process more practical.

6.15.3.3

Summary

Catalytic conversion of natural gas to chemicals presents a promising approach of clean utilization of the abundant resources. The indirect conversion route via syngas intermediate stage has been commercialized as the key for GTL technology while the direct conversion route is still under development, including thermocatalytic conversion in a wide range of temperatures and emerging electrocatalytic and photoelectrocatalytic conversion. The above analysis demonstrates the significant progress that has been made, the great challenges ahead for industrial applications but also vast opportunities. Controlled activation of CeH bond and selective CeC coupling are challenging now and in the future. For oxidative reaction, the key is to avoid deep oxidation leading to CO2 while maintaining a high conversion while for non-oxidative reaction, it is essential to enhance the conversion and at the same time to prohibit deep dehydrogenation leading to coke deposition. The electrocatalytic and photoelectrocatalytic conversion of methane have their unique benefits different from thermo-catalytic processes. However, the research of this field just starts. Intensive efforts are still required to understand the reaction mechanism, to optimize the catalysts and the process, so as to improve methane conversion and product selectivity, as well as long term stability. Future development will require deeper understanding of the mechanism of the known reactions to guide further optimization of the catalysts, which in turn relies on the use of advanced operando characterizations as well as the theoretical modeling. Furthermore, innovations in catalyst and reactor design will also play a vital role in the successful development of these technologies.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

Schwach, P.; Pan, X.; Bao, X. Chem. Rev. 2017, 117 (13), 8497–8520. BP plc, BP Statistical Review of World Energy; BP plc, 2020. Energy Information Administration of US, Natural Gas Annual 2019, 2019. https://www.eia.gov/naturalgas/annual/. p 49. International Energy Agency. CO2 Emissions From Fuel Combustion 2019, 2019. Gunsalus, N. J.; Koppaka, A.; Park, S. H.; Bischof, S. M.; Hashiguchi, B. G.; et al. Chem. Rev. 2017, 117 (13), 8521–8573. Yang, M.; Fan, D.; Wei, Y.; Tian, P.; Liu, Z. Adv. Mater. 2019, 31 (50), e1902181. Pan, X.; Jiao, F.; Miao, D.; Bao, X. Chem. Rev. 2021, 121 (11), 6588–6609. Kosinov, N.; Uslamin, E. A.; Meng, L.; Parastaev, A.; Liu, Y.; et al. Angew. Chem. Int. Ed. Engl. 2019, 58 (21), 7068–7072. Spivey, J. J.; Hutchings, G. Chem. Soc. Rev. 2014, 43 (3), 792–803. Liu, H. M.; Bao, X. H.; Xu, Y. D. J. Catal. 2006, 239 (2), 441–450. Ma, D.; Shu, Y. Y.; Cheng, M. J.; Xu, Y. D.; Bao, X. H. J. Catal. 2000, 194 (1), 105–114. Ma, D.; Shu, Y. Y.; Bao, X. H.; Xu, Y. D. J. Catal. 2000, 189 (2), 314–325. Liu, S. T.; Wang, L.; Ohnishi, R.; Ichikawa, M. J. Catal. 1999, 181 (2), 175–188. Chinchen, G. C.; Mansfield, K.; Spencer, M. S. ChemTech 1990, 20 (11), 692–699. Tang, P.; Zhu, Q.; Wu, Z.; Ma, D. Energ. Environ. Sci. 2014, 7 (8), 2580–2591. Sushkevich, V. L.; Palagin, D.; Ranocchiari, M.; van Bokhoven, J. A. Science 2017, 356 (6337), 523–527. Jin, Z.; Wang, L.; Zuidema, E.; Mondal, K.; Zhang, M.; et al. Science 2020, 367 (6474), 193–197. Pashchenko, D. Fuel 2019, 236, 686–694. Liu, Z.; Gao, F.; Zhu, Y. A.; Liu, Z.; Zhu, K.; et al. Chem. Commun. (Camb.) 2020, 56 (88), 13536–13539. Lu, Q.; Hou, Y.; Laraib, S. R.; Khalifa, O.; Li, K.; et al. Fuel Process. Technol. 2019, 192, 57–64. Kalamaras, C. M.; Efstathiou, A. M., Hydrogen production technologies: Current state and future developments. In Conference Papers in Energy 2013, 2013; Vol. 2013, p 690627. Song, Y.; Ozdemir, E.; Ramesh, S.; Adishev, A.; Subramanian, S.; et al. Science 2020, 367 (6479), 777–781. Akri, M.; Zhao, S.; Li, X.; Zang, K.; Lee, A. F.; et al. Nat. Commun. 2019, 10 (1), 5181. Zhang, J.-C.; Ge, B.-H.; Liu, T.-F.; Yang, Y.-Z.; Li, B.; et al. ACS Catal. 2019, 10 (1), 783–791. Marinho, A. L. A.; Rabelo-Neto, R. C.; Epron, F.; Bion, N.; Toniolo, F. S.; et al. Appl. Catal. B Environ. 2020, 268, 118387. Vakili, R.; Gholami, R.; Stere, C. E.; Chansai, S.; Chen, H.; et al. Appl. Catal. B Environ. 2020, 260, 118159. Wang, C.; Qiu, Y.; Zhang, X.; Zhang, Y.; Sun, N.; et al. Cat. Sci. Technol. 2018, 8 (19), 4877–4890. Tian, J.; Ma, B.; Bu, S.; Yuan, Q.; Zhao, C. Chem. Commun. (Camb.) 2018, 54 (99), 13993–13996. Liu, H. F.; Liu, R. S.; Liew, K. Y.; Johnson, R. E.; Lunsford, J. H. J. Am. Chem. Soc. 2002, 106 (15), 4117–4121. Cortés Ortiz, W. G.; Delgado, D.; Guerrero Fajardo, C. A.; Agouram, S.; Sanchís, R.; et al. Mol. Catal. 2020, 491, 110982. Ma, Y.; Ma, Y.; Chen, Y.; Ma, S.; Li, Q.; et al. Fuel 2020, 265, 116861. Dinh, K. T.; Sullivan, M. M.; Narsimhan, K.; Serna, P.; Meyer, R. J.; et al. J. Am. Chem. Soc. 2019, 141 (29), 11641–11650. Goldshleger, N. F.; Shteinman, A. A.; Shilov, A. E.; Eskova, V. V. Zh. Fiz. Khim. 1972, 46 (5), 1353. Snyder, J. C.; Grosse, A. V. Reaction of Methane With Sulfur Trioxide, 1950. US2493038A. Periana, R. A.; Taube, D. J.; Evitt, E. R.; Loffler, D. G.; Wentrcek, P. R.; et al. Science 1993, 259 (5093), 340–343. Lavrushko, V. V.; Lermontov, S. A.; Shilov, A. E. React. Kinet. Catal. Lett. 1980, 15 (2), 269–272. Periana, R. A.; Taube, D. J.; Gamble, S.; Taube, H.; Fuji, H. High Yield, Low Temperature Oxidation of Methane to Methanol. In Catalytic Activation and Functionalisation of Light Alkanes, Springer, 1998; pp 297–310. Periana, R. A.; Taube, D. J.; Gamble, S.; Taube, H.; Satoh, T.; et al. Science 1998, 280 (5363), 560–564. Diaz-Urrutia, C.; Ott, T. Science 2019, 363 (6433), 1326–1329. Degirmenci, V.; Yilmaz, A.; Uner, D. Catal. Today 2009, 142 (1–2), 30–33.

352 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113.

Inorganic catalysis for methane conversion to chemicals Ding, K.; Metiu, H.; Stucky, G. D. ACS Catal. 2013, 3 (3), 474–477. Paunovic, V.; Zichittella, G.; Moser, M.; Amrute, A. P.; Perez-Ramirez, J. Nat. Chem. 2016, 8 (8), 803–809. Lin, R.; Ding, Y.; Gong, L.; Dong, W.; Chen, W.; et al. Catal. Today 2011, 164 (1), 34–39. Murphy, J. M.; Lawrence, J. D.; Kawamura, K.; Incarvito, C.; Hartwig, J. F. J. Am. Chem. Soc. 2006, 128 (42), 13684–13685. Mkhalid, I. A.; Barnard, J. H.; Marder, T. B.; Murphy, J. M.; Hartwig, J. F. Chem. Rev. 2010, 110 (2), 890–931. Wei, C. S.; Jimenez-Hoyos, C. A.; Videa, M. F.; Hartwig, J. F.; Hall, M. B. J. Am. Chem. Soc. 2010, 132 (9), 3078–3091. Cook, A. K.; Schimler, S. D.; Matzger, A. J.; Sanford, M. S. Science 2016, 351 (6280), 1421–1424. Smith, K. T.; Berritt, S.; Gonzalez-Moreiras, M.; Ahn, S.; Smith, M. R., 3rd; et al. Science 2016, 351 (6280), 1424–1427. Liu, Y.; Dong, W.; Li, Z. H.; Wang, H. Chem 2021, 7 (7), 1843–1851. Horn, R.; Schlögl, R. Catal. Lett. 2015, 145 (1), 23–39. Lance, D.; Elworthy, E. G. Br. J. Exp. Pathol. 1906. GB190607297A. Zhang, Q.; He, D.; Zhu, Q. J. Nat. Gas Chem. 2003, 12, 9. Han, S.; Martenak, D.; Palermo, R.; Pearson, J.; Walsh, D. J. Catal. 1992, 136 (2), 578–583. Michalkiewicz, B. Appl. Catal. A. Gen. 2004, 277 (1–2), 147–153. Chang, C. J. Catal. 1977, 47 (2), 249–259. Alayon, E. M.; Nachtegaal, M.; Ranocchiari, M.; van Bokhoven, J. A. Chem. Commun. (Camb.) 2012, 48 (3), 404–406. Sobolev, V. I.; Dubkov, K. A.; Panna, O. V.; Panov, G. I. Catal. Today 1995, 24 (3), 251–252. Panov, G. I.; Uriarte, A. K.; Rodkin, M. A.; Sobolev, V. I. Catal. Today 1998, 41 (4), 365–385. Starokon, E. V.; Parfenov, M. V.; Pirutko, L. V.; Abornev, S. I.; Panov, G. I. J. Phys. Chem. C 2011, 115 (5), 2155–2161. Parfenov, M. V.; Starokon, E. V.; Pirutko, L. V.; Panov, G. I. J. Catal. 2014, 318, 14–21. Starokon, E. V.; Parfenov, M. V.; Arzumanov, S. S.; Pirutko, L. V.; Stepanov, A. G.; et al. J. Catal. 2013, 300, 47–54. Wulfers, M. J.; Teketel, S.; Ipek, B.; Lobo, R. F. Chem. Commun. (Camb.) 2015, 51 (21), 4447–4450. Beznis, N. V.; Weckhuysen, B. M.; Bitter, J. H. Catal. Lett. 2010, 138 (1–2), 14–22. Ipek, B.; Wulfers, M. J.; Kim, H.; Göltl, F.; Hermans, I.; et al. ACS Catal. 2017, 7 (7), 4291–4303. Bozbag, S. E.; Alayon, E. M. C.; Pechácek, J.; Nachtegaal, M.; Ranocchiari, M.; et al. Cat. Sci. Technol. 2016, 6 (13), 5011–5022. Grundner, S.; Markovits, M. A.; Li, G.; Tromp, M.; Pidko, E. A.; et al. Nat. Commun. 2015, 6 (1), 7546. Le, H. V.; Parishan, S.; Sagaltchik, A.; Göbel, C.; Schlesiger, C.; et al. ACS Catal. 2017, 7 (2), 1403–1412. Kim, Y.; Kim, T. Y.; Lee, H.; Yi, J. Chem. Commun. (Camb.) 2017, 53 (29), 4116–4119. Bozbag, S. E.; Sot, P.; Nachtegaal, M.; Ranocchiari, M.; van Bokhoven, J. A.; et al. ACS Catal. 2018, 8 (7), 5721–5731. Beznis, N. V.; Weckhuysen, B. M.; Bitter, J. H. Catal. Lett. 2009, 136 (1–2), 52–56. Beznis, N. Microporous Mesoporous Mater. 2011, 138, 176–183. Shan, J.; Huang, W.; Nguyen, L.; Yu, Y.; Zhang, S.; et al. Langmuir 2014, 30 (28), 8558–8569. Zhao, G.; Drewery, M.; Mackie, J.; Oliver, T.; Kennedy, E. M.; et al. Energ. Technol. 2020, 8 (8), 1900665. Agarwal, N.; Freakley, S. J.; McVicker, R. U.; Althahban, S. M.; Dimitratos, N.; et al. Science 2017, 358 (6360), 223–227. Keller, G.; Bhasin, M. J. Catal. 1982, 73 (1), 9–19. Choudhary, V. R.; Mulla, S. A. R.; Rane, V. H. J. Chem. Technol. Biotechnol. 1998, 72 (2), 125–130. Choudhary, V. R.; Mulla, S. A. R.; Uphade, B. S. J. Chem. Technol. Biotechnol. 1998, 72 (2), 99–104. Zavyalova, U.; Holena, M.; Schlögl, R.; Baerns, M. ChemCatChem 2011, 3 (12), 1935–1947. Ferreira, V. J.; Tavares, P.; Figueiredo, J. L.; Faria, J. L. Catal. Commun. 2013, 42, 50–53. Huang, P.; Zhao, Y.; Zhang, J.; Zhu, Y.; Sun, Y. Nanoscale 2013, 5 (22), 10844–10848. Noon, D.; Zohour, B.; Senkan, S. J. Nat. Gas Sci. Eng. 2014, 18 (1), 406–411. Song, J.; Sun, Y.; Ba, R.; Huang, S.; Zhao, Y.; et al. Nanoscale 2015, 7 (6), 2260–2264. Sollier, B. M.; Gómez, L. E.; Boix, A. V.; Miró, E. E. Appl. Catal. A. Gen. 2017, 532, 65–76. Liu, Z.; Ho Li, J. P.; Vovk, E.; Zhu, Y.; Li, S.; et al. ACS Catal. 2018, 8 (12), 11761–11772. Ito, T.; Wang, J. X.; Lin, C. H.; Lunsford, J. H. J. Am. Chem. Soc. 1985, 107 (18), 5062–5068. Morales, E. J. Catal. 1989, 118 (1), 255–265. Driscoll, D. J.; Martir, W.; Wang, J. X.; Lunsford, J. H. J. Am. Chem. Soc. 2002, 107 (1), 58–63. Maksimov, N. G.; Selyutin, G. E.; Anshits, A. G.; Kondratenko, E. V.; Roguleva, V. G. Catal. Today 1998, 42 (3), 279–281. Arndt, S.; Laugel, G.; Levchenko, S.; Horn, R.; Baerns, M.; et al. Catal. Rev. 2011, 53 (4), 424–514. Tang, L.; Yamaguchi, D.; Wong, L.; Burke, N.; Chiang, K. Catal. Today 2011, 178 (1), 172–180. Simon, U.; Arndt, S.; Otremba, T.; Schlingmann, T.; Görke, O.; et al. Catal. Commun. 2012, 18, 132–136. Schwach, P.; Willinger, M. G.; Trunschke, A.; Schlogl, R. Angew. Chem. Int. Ed. Engl. 2013, 52 (43), 11381–11384. Palermo, A.; Holgadovazquez, J.; Lee, A.; Tikhov, M.; Lambert, R. J. Catal. 1998, 177 (2), 259–266. Hinson, P. G.; Clearfield, A.; Lunsford, J. H. J. Chem. Soc. Chem. Commun. 1991, 20, 1430–1432. Ji, S. J. Catal. 2003, 220 (1), 47–56. Yildiz, M.; Aksu, Y.; Simon, U.; Kailasam, K.; Goerke, O.; et al. Chem. Commun. (Camb.) 2014, 50 (92), 14440–14442. Colmenares, M. G.; Simon, U.; Yildiz, M.; Arndt, S.; Schomaecker, R.; et al. Catal. Commun. 2016, 85, 75–78. Yildiz, M.; Simon, U.; Otremba, T.; Aksu, Y.; Kailasam, K.; et al. Catal. Today 2014, 228, 5–14. Cheng, Z.; Qin, L.; Guo, M.; Xu, M.; Fan, J. A.; et al. Phys. Chem. Chem. Phys. 2016, 18 (47), 32418–32428. Cheng, Z.; Qin, L.; Guo, M.; Fan, J. A.; Xu, D.; et al. Phys. Chem. Chem. Phys. 2016, 18 (24), 16423–16435. Farsi, A.; Mansouri, S. S. Arab. J. Chem. 2016, 9, S28–S34. Zou, S.; Li, Z.; Zhou, Q.; Pan, Y.; Yuan, W.; et al. Chin. J. Catal. 2021, 42 (7), 1117–1125. Hutchings, G.; Scurrell, M.; Woodhouse, J. Chem. Soc. Rev. 1989, 18, 251. Langfeld, K.; Frank, B.; Strempel, V. E.; Berger-Karin, C.; Weinberg, G.; et al. Appl. Catal. A. Gen. 2012, 417, 145–152. Yakovlev, A. L.; Zhidomirov, G. M.; van Santen, R. A. Catal. Lett. 2001, 75 (1), 45–48. Peter, M.; Marks, T. J. J. Am. Chem. Soc. 2015, 137 (48), 15234–15240. Wang, L.; Tao, L.; Xie, M.; Xu, G.; Huang, J.; et al. Catal. Lett. 1993, 21 (1–2), 35–41. Zheng, H.; Ma, D.; Bao, X.; Hu, J. Z.; Kwak, J. H.; et al. J. Am. Chem. Soc. 2008, 130 (12), 3722–3723. Xu, Y.; Wang, J.; Suzuki, Y.; Zhang, Z.-G. Appl. Catal., A 2011, 409-410, 181–193. Xu, Y.; Lu, J.; Wang, J.; Suzuki, Y.; Zhang, Z.-G. Chem. Eng. J. 2011, 168 (1), 390–402. Sun, C.; Fang, G.; Guo, X.; Hu, Y.; Ma, S.; et al. J. Energy Chem. 2015, 24 (3), 257–263. Liu, S.; Wang, L.; Ohnishi, R.; Lchikawa, M. Kinet. Catal. 2000, 41 (1), 132–144. Lezcano-Gonzalez, I.; Oord, R.; Rovezzi, M.; Glatzel, P.; Botchway, S. W.; et al. Angew. Chem. Int. Ed. Engl. 2016, 55 (17), 5215–5219.

Inorganic catalysis for methane conversion to chemicals 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182.

353

Honda, K.; Yoshida, T.; Zhang, Z.-G. Catal. Commun. 2003, 4 (1), 21–26. Chu, N.; Yang, J.; Wang, J.; Yu, S.; Lu, J.; et al. Catal. Commun. 2010, 11 (6), 513–517. Chu, N.; Yang, J.; Li, C.; Cui, J.; Zhao, Q.; et al. Microporous Mesoporous Mater. 2009, 118 (1–3), 169–175. Wang, L. S.; Ohnishi, R.; Ichikawa, M. J. Catal. 2000, 190 (2), 276–283. Vollmer, I.; Mondal, A.; Yarulina, I.; Abou-Hamad, E.; Kapteijn, F.; et al. Appl. Catal. A. Gen. 2019, 574, 144–150. Tshabalala, T. E.; Coville, N. J.; Scurrell, M. S. Catal. Commun. 2016, 78, 37–43. Ma, S.; Guo, X.; Zhao, L.; Scott, S.; Bao, X. J. Energy Chem. 2013, 22 (1), 1–20. Lai, Y.; Veser, G. Cat. Sci. Technol. 2016, 6 (14), 5440–5452. Kosinov, N.; Coumans, F. J. A. G.; Li, G.; Uslamin, E.; Mezari, B.; et al. J. Catal. 2017, 346, 125–133. Denardin, F.; Perez-Lopez, O. W. Fuel 2019, 236, 1293–1300. Weckhuysen, B. M.; Wang, D. J.; Rosynek, M. P.; Lunsford, J. H. Angew. Chem. Int. Ed. Engl. 1997, 36 (21), 2374–2376. Dehydro-aromatization of CH4 over W-Mn(or Zn, Ga, Mo, Co)/HZSM-5(or MCM-22) catalysts. In Natural Gas Conversion VII, Proceedings of the 7th Natural Gas Conversion Symposium; Huang, L. Q., Yuan, Y. Z., Zhang, H. B., Xiong, Z. T., Zeng, J. L., ; et al.Bao, X., Xu, Y., Eds.; vol. 147; Elsevier, 2004; pp 565–570. Weckhuysen, B. M.; Wang, D.; Rosynek, M. P.; Lunsford, J. H. J. Catal. 1998, 175 (2), 338–346. Zeng, J. L.; Xiong, Z. T.; Zhang, H. B.; Lin, G. D.; Tsai, K. R. Catal. Lett. 1998, 53 (1/2), 119–124. Wang, L. S.; Ohnishi, R.; Ichikawa, M. Catal. Lett. 1999, 62 (1), 29–33. Li, S.; Zhang, C.; Kan, Q.; Wang, D.; Wu, T.; et al. Appl. Catal. A. Gen. 1999, 187 (2), 199–206. Tan, P. L.; Au, C. T.; Lai, S. Y. Catal. Lett. 2006, 112 (3–4), 239–245. Lu, Y.; Xu, Z.; Tian, Z.; Zhang, T.; Lin, L. Catal. Lett. 1999, 62, 215–220. Shu, Y. Y.; Ma, H. T.; Ohnishi, R.; Ichikawa, M. Chem. Commun. 2003, (1), 86–87. Cao, Z.; Jiang, H.; Luo, H.; Baumann, S.; Meulenberg, W. A.; et al. Angew. Chem. Int. Ed. Engl. 2013, 52 (51), 13794–13797. Xue, J.; Chen, Y.; Wei, Y. Y.; Feldhoff, A.; Wang, H. H.; et al. ACS Catal. 2016, 6 (4), 2448–2451. Morejudo, S. H.; Zanon, R.; Escolastico, S.; Yuste-Tirados, I.; Malerod-Fjeld, H.; et al. Science 2016, 353 (6299), 563–566. Mériaudeau, P.; Ha, V. T. T.; Tiep, L. V. Catal. Lett. 2000, 64 (1), 49–51. Kosinov, N.; Coumans, F. J. A. G.; Uslamin, E. A.; Wijpkema, A. S. G.; Mezari, B.; et al. ACS Catal. 2016, 7 (1), 520–529. Guo, X.; Fang, G.; Li, G.; Ma, H.; Fan, H.; et al. Science 2014, 344 (6184), 616–619. Liu, Y.; Liu, J. C.; Li, T. H.; Duan, Z. H.; Zhang, T. Y.; et al. Angew. Chem. Int. Ed. Engl. 2020, 59 (42), 18586–18590. Sakbodin, M.; Wu, Y.; Oh, S. C.; Wachsman, E. D.; Liu, D. Angew. Chem. Int. Ed. Engl. 2016, 55 (52), 16149–16152. Han, S. J.; Lee, S. W.; Kim, H. W.; Kim, S. K.; Kim, Y. T. ACS Catal. 2019, 9 (9), 7984–7997. Xie, P.; Pu, T.; Nie, A.; Hwang, S.; Purdy, S. C.; et al. ACS Catal. 2018, 8 (5), 4044–4048. Oh, S. C.; Schulman, E.; Zhang, J.; Fan, J.; Pan, Y.; et al. Angew. Chem. Int. Ed. Engl. 2019, 58 (21), 7083–7086. Hao, J.; Schwach, P.; Fang, G.; Guo, X.; Zhang, H.; et al. ACS Catal. 2019, 9 (10), 9045–9050. Hao, J. Q.; Schwach, P.; Li, L. L.; Guo, X. G.; Weng, J. B.; et al. J. Energy Chem. 2021, 52, 372–376. Huang, K.; Miller, J. B.; Huber, G. W.; Dumesic, J. A.; Maravelias, C. T. Joule 2018, 2 (2), 349–365. Jang, J.; Shen, K.; Morales-Guio, C. G. Joule 2019, 3 (11), 2589–2593. Fornaciari, J. C.; Primc, D.; Kawashima, K.; Wygant, B. R.; Verma, S.; et al. ACS Energy Lett. 2020, 5 (9), 2954–2963. Xie, S.; Lin, S.; Zhang, Q.; Tian, Z.; Wang, Y. J. Energy Chem. 2018, 27 (6), 1629–1636. Lee, B.; Hibino, T. J. Catal. 2011, 279 (2), 233–240. Yuan, S.; Li, Y.; Peng, J.; Questell-Santiago, Y. M.; Akkiraju, K.; et al. Adv. Energy Mater. 2020, 10 (40), 2002154. Latimer, A. A.; Kakekhani, A.; Kulkarni, A. R.; Nørskov, J. K. ACS Catal. 2018, 8 (8), 6894–6907. Grubb, W. T.; Michalske, C. J. Nature 1964, 201 (4916), 287–288. Stoukides, M. J. Appl. Electrochem. 1995, 25 (10), 899–912. Joglekar, M.; Nguyen, V.; Pylypenko, S.; Ngo, C.; Li, Q.; et al. J. Am. Chem. Soc. 2016, 138 (1), 116–125. Song, Y.; Zhao, Y.; Nan, G.; Chen, W.; Guo, Z.; et al. Appl. Catal. B Environ 2020, 270, 118888. Ma, M.; Jin, B. J.; Li, P.; Jung, M. S.; Kim, J. I.; et al. Adv. Sci. (Weinheim, Ger.) 2017, 4 (12), 1700379. Sher Shah, M. S. A.; Oh, C.; Park, H.; Hwang, Y. J.; Ma, M.; et al. Adv. Sci. (Weinheim, Ger.) 2020, 7 (23), 2001946. Bagherzadeh Mostaghimi, A. H.; Al-Attas, T. A.; Kibria, M. G.; Siahrostami, S. J. Mater. Chem. A 2020, 8 (31), 15575–15590. Mohamed, A. G. A.; Naqviab, S. A. Z.; Wang, Y. B. ChemCatChem 2021, 13 (3), 787–805. Sargeant, E.; Kolodziej, A.; Le Duff, C. S.; Rodriguez, P. ACS Catal. 2020, 10 (14), 7464–7474. Promoppatum, P.; Viswanathan, V. ACS Sustain. Chem. Eng. 2016, 4 (3), 1736–1745. Latimer, A. A.; Kulkarni, A. R.; Aljama, H.; Montoya, J. H.; Yoo, J. S.; et al. Nat. Mater. 2017, 16 (2), 225–229. Guo, Z.; Liu, B.; Zhang, Q.; Deng, W.; Wang, Y.; et al. Chem. Soc. Rev. 2014, 43 (10), 3480–3524. Liu, K.; Zhao, J.; Zhu, D.; Meng, F.; Kong, F.; et al. Catal. Commun. 2017, 96, 23–27. Carroll, J. J.; Jou, F.-Y.; Mather, A. E.; Otto, F. D. Can. J. Chem. Eng. 1998, 76 (5), 945–951. Gao, D.; Wei, P.; Li, H.; Lin, L.; Wang, G.; et al. Acta Phys. -Chim. Sin. 2021, 37 (5), 2009021. Rocha, R. S.; Reis, R. M.; Lanza, M. R. V.; Bertazzoli, R. Electrochim. Acta 2013, 87, 606–610. Torabi, A.; Barton, J.; Willman, C.; Ghezel-Ayagh, H.; Li, N.; et al. ECS Trans. 2016, 72 (7), 193–199. Rocha, R. S.; Camargo, L. M.; Lanza, M. R. V.; Bertazzoli, R. Electrocatalysis 2010, 1 (4), 224–229. Lee, B.; Sakamoto, Y.; Hirabayashi, D.; Suzuki, K.; Hibino, T. J. Catal. 2010, 271 (2), 195–200. Lee, J.; Yang, J.; Moon, J. H. ACS Energy Lett. 2021, 6 (3), 893–899. Wang, Q.; Li, T.; Yang, C.; Chen, M.; Guan, A.; et al. Angew. Chem. Int. Ed. Engl. 2021, 60 (32), 17398–17403. Zhu, C.; Hou, S.; Hu, X.; Lu, J.; Chen, F.; et al. Nat. Commun. 2019, 10 (1), 1173. Ma, M.; Oh, C.; Kim, J.; Moon, J. H.; Park, J. H. Appl. Catal. B Environ. 2019, 259, 118095. Tomita, A.; Nakajima, J.; Hibino, T. Angew. Chem. Int. Ed. Engl. 2008, 47 (8), 1462–1464. Spinner, N.; Mustain, W. E. J. Electrochem. Soc. 2013, 160 (11), F1275–F1281. O’Reilly, M. E.; Kim, R. S.; Oh, S.; Surendranath, Y. ACS Cent. Sci. 2017, 3 (11), 1174–1179. Xu, N.; Coco, C. A.; Wang, Y.; Su, T.; Wang, Y.; et al. Appl. Catal. B Environ. 2021, 282, 119572. Li, W.; He, D.; Hu, G.; Li, X.; Banerjee, G.; et al. ACS Cent. Sci. 2018, 4 (5), 631–637. Amano, F.; Shintani, A.; Tsurui, K.; Mukohara, H.; Ohno, T.; et al. ACS Energy Lett. 2019, 4 (2), 502–507. Ma, J.; Mao, K.; Low, J.; Wang, Z.; Xi, D.; et al. Angew. Chem. Int. Ed. Engl. 2021, 60 (17), 9357–9361.

6.16

Promoted Fischer-Tropsch catalysts

Paul B. Webba and Ivo A.W. Filotb, a University of St Andrews, School of Chemistry, St Andrews, Fife, United Kingdom; and b Eindhoven University of Technology, Inorganic Materials and Catalysis, Eindhoven, The Netherlands © 2023 Elsevier Ltd. All rights reserved.

6.16.1 6.16.1.1 6.16.1.2 6.16.1.2.1 6.16.1.2.2 6.16.1.2.3 6.16.1.3 6.16.1.3.1 6.16.2 6.16.2.1 6.16.2.2 6.16.2.3 6.16.2.4 6.16.2.5 6.16.2.6 6.16.2.6.1 6.16.2.6.2 6.16.3 6.16.4 6.16.4.1 6.16.4.2 6.16.5 6.16.5.1 6.16.5.2 6.16.6 References

Fischer-Tropsch synthesis: Past, present and future The past, an historical perspective Present commercial operations Fischer-Tropsch catalysts and processes Fe catalysts Co catalysts The future role of FT in achieving net zero carbon Power-to-liquids, a future beyond biomass? Mechanism of the Fischer-Tropsch reaction Reactant adsorption Monomer formation and chain initiation Chain growth Chain termination and product desorption Readsorption and further reaction Mechanistic models Carbide mechanism CO-insertion mechanism Structure sensitivity Kinetics of the Fischer-Tropsch reaction Kinetic modeling Mechanistic insights Fischer-Tropsch to chemicals (Chem FT) Lower olefin synthesis Higher alcohol synthesis Outlook

354 355 356 357 358 358 360 360 361 361 362 362 363 363 363 364 364 366 368 368 369 371 372 374 375 376

Abstract Fischer-Tropsch synthesis (FTS) is the conversion of synthesis gas, a mixture of carbon monoxide and hydrogen, into a broad range of hydrocarbon products, including olefins, paraffins and oxygenates. The process is essentially a polymerization reaction involving the stepwise coupling of C1 monomeric units that are generated in situ from CO and H2. Careful design of catalyst and process parameters allows for selectivity control and the production of a versatile range of fuels and chemicals. In this chapter, a brief historical overview will be given, from the early discoveries of Frans Fischer and Hans Tropsch up to present day state-of-the-art commercial processes. Features of catalyst structure, preparation and performance will be discussed in detail and followed by a contemporary overview of the kinetic mechanism and structure-sensitivity of the reaction. These mechanistic insights provide the basis for a “smart design” approach to next generation catalysts. Concepts for tuning selectively to high-value chemicals, such as olefins and higher alcohols, will be discussed and the chapter will conclude with a brief outlook on the future role of FTS in a post-fossil fuel world.

6.16.1

Fischer-Tropsch synthesis: Past, present and future

Fischer-Tropsch synthesis (FTS) is the conversion of synthesis gas (CO þ H2) to aliphatic hydrocarbons and water. Developed by Frans Fischer and Hans Tropsch at the Kaiser Wilhelm Institute for Coal Research in Mülheim, the main products are linear a-olefins (1) and n-paraffins (2) as summarized in the simplified reaction equations below.1

354

2nH2 þ nCO/Cn H2n þ nH2 O

(1)

ð2n þ 1ÞH2 þ nCO/Cn H2nþ2 þ nH2 O

(2)

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00034-0

Promoted Fischer-Tropsch catalysts

355

Branched hydrocarbons (mainly monomethyl branching) and oxygenates (alcohols, aldehydes, ketones and acids) are also formed as minority species. Fischer-Tropsch synthesis has been studied extensively for almost a century with renewed interest in recent years driven largely by the potential for converting a diverse range of biomass and waste streams into liquid products. It is a vast topic that has been the subject of many excellent reviews.2–5 In this chapter, a brief history of FTS and the current status of commercial processes is first provided. In keeping with the purpose and intended scope of this book, emphasis will then be placed on the reaction mechanism, the nature of catalytically active sites, the challenges of manipulating product selectivity and the potential role of FT in achieving global net zero targets.

6.16.1.1

The past, an historical perspective

Fischer-Tropsch synthesis has a long and fascinating history that is largely beyond the scope of this text. It would be remiss, however, not to include a brief account here and the reader is referred to several articles for a more detailed overview.6–9 The development of FT technology was born out of research exploring possible routes to hydrocarbons from abundant German coal reserves. This drive for energy independence led to Bergius’ invention of direct coal liquefaction, a high-pressure (200 bar), high-temperature (400  C) hydrogenation process that yields petroleum-like liquids.6 This was the first successful production of synthetic hydrocarbons from coal. Developments in indirect coal liquefaction were first patented by Baden Aniline and Soda Factory (BASF) in 1913.6 This alternative route to hydrocarbons, via the catalytic hydrogenation of carbon monoxide, was not progressed further by BASF owing to a prioritization of research into ammonia and methanol synthesis. Building on the earlier BASF work, Fischer and Tropsch disclosed the production of “Synthol,” an oily mixture of mostly oxygenates produced from coal-derived synthesis gas using an alkali–iron catalyst at 400–450  C and 100–150 bar.10 Further developments using a Co–Fe catalyst operating at 250–300  C and normal pressure largely eliminated oxygenate production, yielding desired olefins and paraffins. Other notable contributions to early developments were made by Otto Roelen at Ruhrchemie.6 His highly active Co-ThO2-kieselguhr catalysts were capable of operation at significantly lower temperatures and became the benchmark for further optimization. Following years of catalyst and process development, four Ruhrchemie licensed FT plants, with a total annual capacity of  100,000 metric tonnes, were under construction by late 1935. With the completion of five additional plants and increased annual capacity of the original units to 300,000 metric tonnes (2.17 million barrels (bbl); 1bbl ¼ 159 liters), total capacity peaked at 576,000 metric tonnes (4.1 million bbl) in 1944.6 The processes operated at 180–200  C with either atmospheric or medium pressure (5–15 bar) synthesis gas using Cobased catalysts. Synthesis gas was generated by reacting coke with steam, the overall process enabling the production of petroleum from coal supplies to establish the world’s first synthetic fuel industry. Up to circa 1955, an increasing demand for petroleum and less than optimistic oil forecasts fueled continued developments in FTS within Germany, the United States and South Africa. With the discovery of major oil fields in the Middle East, Alaska and the North Sea, the future outlook for FT changed dramatically and global interest largely declined. At this time, activities continued in South Africa, which benefitted from cheap domestic coal and state policy. The oil crisis of the 1970s then initiated a new wave of activity in FT research and led ultimately to a number of new plants being constructed. Those still in operation today are summarized in Table 1. Table 1

Commercial FT processes currently in operation.

Company (licensor)

Location

Process

Design

Catalyst

Start-up date

Sasol

Sasolburg, South Africa

GTL

Precipitated Fe/K

Sasol

Secunda, South Africa

CTL/GTL

Shell PetroSA Sasol-QP Lu’an (Synfuels China) Yitai (Synfuels China) Shenua Group (Synfuels China) Shell Sasol-Chevron Shenua Group (Synfuels China) NiQuan Energy (Emerging Fuels) Fulcrum Bioenergy (BP and JM)

Bintulu, Malaysia Mossel Bay, South Africa Ras Laffan, Qatar Shanxi, China

GTL GTL GTL CTL

LTFT fixed bed LTFT slurry phase HTFT CFB HTFT SAS LTFT fixed bed HTFT fluidized bed LTFT slurry phase MTFT slurry phase

Co-TiO2 Fused Fe/K Co-Al2O3 Fe

1955 1993 1980 1995 1993 1993 2006 2009

Inner Mongolia, China

CTL

MTFT slurry phase

Precipitated Fe/K

2010

Inner Mongolia, China

CTL

MTFT slurry phase

Fe

2010

Ras Laffan, Qatar Escravos, Nigeria Ningxia, China

GTL GTL CTL

LTFT fixed bed LTFT slurry phase MTFT slurry phase

Co-TiO2 Co-Al2O3 Fe

2011 2013 2016

Pointe-à-Pierre, Trinidad & Tobago

GTL

LTFT fixed bed

Co-Al2O3

2021

Nevada, United States

WTL

LTFT CANS™

Co-TiO2

2021

Fused Fe/K

Based on data available in the public domain. CFB, circulating fluidized bed; SAS, sasol advanced synthol; CANSÔ, modular fixed bed technology; LTFT, low temperature FischerTropsch; MTFT, medium temperature Fischer-Tropsch; HTFT, high temperature Fischer-Tropsch.

356

Promoted Fischer-Tropsch catalysts

Detailed analysis of Fischer-Tropsch products reveals a high degree of order with respect to the molecular weight distribution. To a first approximation, the molar amount of individual carbon number fractions declines exponentially with increasing carbon number. This led to an early recognition of the reaction being a polymerization process involving the stepwise addition of a C1 monomeric unit that is generated in situ from CO and H2 (Fig. 1).11,12 Considering the simplified reaction sequence shown in Fig. 1, the C1 monomeric unit can either desorb as product methane or couple with another C1 unit to generate a C2* surface species. The probability of chain propagation occurring is defined by the parameter a. A mathematical treatment of the product distribution becomes possible with the basic assumptions that all Cn products arise only from sequential chain growth, i.e. thermal cracking of products does not occur, and the chain growth probability, a, is independent of chain length. Starting with a surface population of C1* ¼ 1, the fraction of product desorbed as methane will be given by C1 ¼ 1(1  a). Analogous expressions for longer chain products can be derived by applying a steady state approximation where the rate of formation of any surface species Cn* is equal to the rate of its consumption, (dCn*/dt ¼ 0). It can then be shown that: Cn ¼ nð1  aÞan1

(3)

The total amount of product formed can be determined by summing over all values of Cn and the convergent series obtained can be simplified using a power series: N X n¼1

Cn ¼

N X

nð1  aÞan1 ¼

n¼1

1 1a

(4)

It follows that the selectivity to a specific product, Sn, can be represented by combining Eqs. (3) and (4): Cn Sn ¼ PN

n¼1 Cn

¼ nð1  aÞ2 an1

and conversion of Eq. (5) to logarithmic form gives: !   Sn ð1  aÞ2 ¼ nlnðaÞ þ ln ln n a

(5)

(6)

For an ideal product distribution, which meets the assumptions of chain length independent a, a plot of ln(Sn/n) against carbon number n will give a linear response from which a is easily determined (Fig. 2A). Typically referred to as an Anderson-Schulz-Flory (ASF) distribution,11,13,14 this approach is often used to characterize FTS product slates. The idealized kinetics used to derive the ASF distribution place constraints on the theoretically achievable product spectrum (Fig. 2B). With a chain growth probability of 0, methane is the only product that can be formed with 100% selectivity (C1 ¼ 1(1  a) ¼ 1) and as the value of a increases so does the carbon number range of products being formed. In practice, an ideal ASF distribution is rarely obtained with methane production often higher than predicted and C2 formation lower. These observations can be attributed to additional pathways to methane production and a stronger binding and facile chain growth of ethene compared to higher olefins.15 The “kink” in the ASF plot often observed at low reaction temperatures (Fig. 2A) can be explained by the readsorption of lighter olefins and their further reincorporation into the growing chain. This chain length dependence in a reflects a change in the solubility of olefins in the liquid phase, which increases the residence time of readsorbed species. Other physical phenomena, such as diffusivity and physisorption, may also play a role.16–20

6.16.1.2

Present commercial operations

Fischer-Tropsch technology was developed initially for the conversion of coal to hydrocarbons but diverging oil and gas prices have more recently provided economic incentive for the use of natural gas as a feedstock. A resurgent interest in FT chemistry in recent

Fig. 1 A simplified model of hydrocarbon production via the stepwise coupling of C1* building blocks. The molecular weight distribution of products in this polymerization process is defined by the chain growth probability, a.

Promoted Fischer-Tropsch catalysts

357

Fig. 2 Fischer-Tropsch product distribution. (A) Mole fraction versus carbon number showing the fitted Anderson-Schulz-Flory distribution for low temperature (LTFT) and high temperature (HTFT) operation. Note the C1 and C2 anomalies. (B) Weight fractions of various Cn mixtures as function of the chain-growth parameter assuming an ideal ASF distribution.

years has been motivated also by the potential for converting sources of so-called above ground carbon (biomass and waste streams) into synthetic crude. Commercial processes are typically referred to by the nature of the carbon feedstock i.e. coal-to-liquids (CTL), gas-to-liquids (GTL), biomass-to-liquids (BTL) and waste-to-liquids (WTL) or collectively as anything to liquids (XTL). The generation of synthesis gas (“syngas” for short) from these hydrocarbonaceous feedstocks is a key first step in all commercial FT plants. Solid raw materials are processed by gasification in the presence of oxygen and steam. Partial oxidation with oxygen supplies the heat energy for this endothermic process and the steam provides a source of hydrogen. Following gasification, impurities such as sulfur and nitrogen compounds, and entrained solid particles, are removed to obtain pure syngas. Natural gas can also be adiabatically reformed in the presence of oxygen and steam by partial oxidation or autothermal reforming. Alternatively, methane can be reformed in the absence of oxygen with the required heat energy being applied externally. Examples include steam methane reforming (CH4 þ H2O), dry reforming (CH4 þ CO2), bi reforming (CH4 þ CO2 þ H2O) and heat exchange reforming. Ideally, the generated syngas ratio should match the usage ratio, the stoichiometric requirements of H2 and CO for hydrocarbon production. For saturated hydrocarbons the usage ratio is 2. Hydrogen lean feedstocks can yield H2:CO ratios as low as 0.7 and this is particularly problematic for coal gasification. Additional gas processing using water gas shift reactors to adjust the H2:CO ratio is then required and will be discussed further in the context of Fe-based catalysts (Section 6.16.1.2.2).

6.16.1.2.1

Fischer-Tropsch catalysts and processes

Commercial FTS has been adopted globally with an estimated annual production 450,000 bbl per day in 2021. For context, this capacity is dwarfed by the annual production of crude oil, currently standing at 95–100 million bbl per day. The largest coal to liquids and gas to liquids operations are Sasol’s Secunda facility in South Africa and the Shell-Qatar Petroleum’s Pearl facility at Ras Laffan, Qatar, each producing over 140,000 bbl per day. Current commercial processes, and those in the final stages of startup at the time of writing, are summarized in Table 1. Commercial operations can be categorized firstly by operating temperature regimes. Historically, two principal modes of operation were recognized; low-temperature FT (LTFT, 200–250  C) and high-temperature FT (HTFT, 320–350  C), but more recent developments by Synfuels China has established a third, mid-temperature process (MTFT, 260–300  C). Raising the reaction temperature shifts the product distribution to lower carbon numbers. High temperature operation, in nearly isothermal fluidized bed reactors, comes close to yielding products with an ideal ASF behavior (Fig. 2A). The much lighter product spectrum is suited to the production of gasoline (Fig. 2B) but a high selectivity toward linear a-olefins and oxygenates allows also for chemicals extraction. Of particular value are ethylene and propylene, the building blocks of polyethylene and polypropylene production, 1-hexene and 1-octene, co-monomers used in the manufacture of linear low-density polyethylene, and alcohols, aldehydes and ketones that find use as solvents. At these higher temperatures the lighter product remains in the gas phase, which is critical to the operation of this gas-solid reaction using fluidized bed technology.21 Low temperature operation generates a much heavier product slate that extends into waxes. This mode of operation is suitable for the production of middle distillate range products (C9–C22), such as diesel (C9–C22) and jet fuel (C9–C15). According to Fig. 2B, middle distillate yields peak at an a value of 0.85, but will account for only 40% of the entire product spectrum under these conditions. To maximize middle distillate production, LTFT processes operated by Shell and Sasol are designed to rather target wax (a ¼ 0.95), which is then hydrocracked to the desired hydrocarbon range with much greater overall selectivity.22,23 Another distinguishing operational parameter is catalyst type. The hydrogenation of CO can be catalyzed by group VIIIB metals, but only Co and Fe find use in commercial applications. The majority of group VIIIB metals are prohibitively expensive and the methanation activity of Ni is simply too high to be of practical use for the production of long chain hydrocarbons.24 There are three

358

Promoted Fischer-Tropsch catalysts

catalyst classes in commercial use today; fused iron for HTFT, precipitated iron for LTFT and supported Co catalysts for LTFT. Salient features of catalyst design and preparation are outlined in the following sections.

6.16.1.2.2

Fe catalysts

Iron catalysts find application in processes spanning the entire range of commercial operating temperatures. Catalyst design is dictated largely by the mode of operation and reactor configuration.25 Sasol’s Advanced Synthol process (HTFT) employs fluidized bed technology using a fused, bulk iron catalyst. Fluidization places mechanical demands on the catalyst and particle strength is therefore a critical design feature together with a high density required for effective fluidization; both criteria are met by bulk Fe. The catalyst is prepared by fusing a raw iron oxide material, such as raw mill-scale, together with promoter elements, primarily potassium (K2O), in an arc furnace at temperatures close to 1500  C. The molten magnetite is then cast into ingots and cooled, producing a low surface area material (< 10 m2 g 1). The ingots are finally crushed and sieved to a particle size range designed for optimum fluidization. The addition of alkali promoters was exploited in the early developments of Fischer and Tropsch and remains commonplace today. For Fe-based catalysts, alkali promotion is needed to attain high activity, but also enhances the selectivity toward olefins and longer chain products with a concomitant decrease in methane formation. A low methane selectivity is highly desirable, but there is a practical upper limit to the amount of potassium that can be added without causing operational problems arising from the formation of heavy hydrocarbons that are insufficiently volatile to remain in the gas phase. Other undesirable consequences associated with a high alkali content are an increased rate of carbon formation on the catalyst and increased levels of organic acids in the hydrocarbon products. Further discussion on the role of alkali promoters is provided in Section 6.16.5. Sasol’s Fe-catalyzed LTFT process is used to produce high molecular weight products for the wax industry.21 The process was operated originally in multi-tubular fixed-bed reactors, known as ARGE-reactors, developed jointly by Lurgi and Ruhrchemie.26 For operation in fixed-beds, minimization of mass transfer limitations must be taken into consideration, and here catalyst strength is less important than catalyst shape and form. Later developments saw the introduction of slurry-bed reactors and a phasing out of fixed-bed operation. To compensate for the decreased reaction rates at lower operating temperature, catalysts used in LTFT must exhibit a far greater surface area than the bulk iron used in the HTFT process. The Fe LTFT catalyst is therefore prepared by the precipitation of iron oxides and oxy-hydroxides from iron(III) nitrate solutions upon reaction with a suitable base (e.g., sodium carbonate). Following precipitation, the slurry is filtered and washed to remove soluble nitrate salts. Structural promoters (silica, alumina or magnesia), which provide mechanical strength and stabilize the active phase, are then added by impregnation of the reslurried mixture. Alkali promotion is again important and copper is also added early in the synthesis to enhance the reduction of iron oxide in a subsequent catalyst reduction/conditioning step. For slurry bed applications the catalyst is spray dried to produce spherical particles.27 A final calcination step removes water and other volatile compounds such as NOx and improves mechanical strength, which becomes important again for slurry-bed operation. The as prepared iron oxides and (oxy)hydroxides do not catalyze the FT reaction and must first be reduced to an active form. This activation step can be achieved in hydrogen, CO or a mixture of both. In the SAS process, the magnetite powder is pre-reduced with hydrogen at 350–450  C before being loaded into the synthesis reactor.21 This strategy has also been applied in the low temperature process.28 The reduction of iron oxides in hydrogen generates a-Fe metal but under FTS conditions the structure evolves to a mixture of carbides. Several iron carbide phases, predominantly Hägg-carbide (c-Fe5C2), and magnetite can co-exist during steady-state operation.29–33 Iron carbide formation is needed for high activity and it is widely accepted that carbides constitute the active phase.4,29,32,34–37 An important feature of Fe-based processes is the water gas shift (WGS) reaction, believed to be catalyzed by Fe3O4 that exists in dynamic equilibrium with carbides during FTS.38–41 This results in a high proportion of oxygen being removed as CO2 as well as H2O. CO þ H2 O/H2 þ CO2 ;

DH ¼  41 kJ=mol

(7)

While WGS chemistry has a deleterious impact on carbon utilization, it provides a pathway to additional hydrogen that can be beneficial in coal-to-liquids processes. The H2/CO ratio of coal-derived syngas can be as low as 0.7–1.0, but this can be compensated by the WGS activity of Fe, which circumvents the need for a separate, upstream shift reactor to provide the necessary H2/CO ratio.32,36 This intrinsic ability to concurrently catalyze both FTS and WGS reactions is one of the main reasons, besides low cost, for the predominance of Fe catalysts in CTL processes. For high temperature operation the WGS reaction is close to equilibrium. The addition of CO2 to synthesis gas can therefore eliminate its production during FTS, shifting the mechanism of O adatom removal entirely toward the formation of water. The recycle of CO2 with fresh feed gas can therefore dramatically improve carbon efficiency.42,43 On a final note, the activity of iron catalysts is inhibited by product water and this restricts the achievable degree of conversion.26 Lower per pass conversions require a higher gas recycle, a feature that is partly responsible for an increased interest in Co-based catalysts, which are less susceptible to high partial pressures of water.

6.16.1.2.3

Co catalysts

Cobalt-based materials are now the preferred catalysts for middle distillate production, owing to their high activity, exceptional C5þ selectivity and long operational lifetimes. Unlike iron, cobalt catalysts are limited to low temperature operation; under HTFT conditions mostly methane is formed on cobalt. Other notable differences to Fe-based catalysts include a higher hydrogenating capacity, which gives rise to a more saturated product slate, and negligible WGS activity. The absence of WGS chemistry ensures that CO is not

Promoted Fischer-Tropsch catalysts

359

sacrificed as CO2 and instead the majority of oxygen is removed as by-product water. Cobalt is therefore ideally suited to GTL processes where the H2/CO ratio is already at or above the usage ratio and there is no need to produce additional hydrogen. The considerably higher cost of cobalt precludes its use as a bulk metal. Developments in catalyst design have therefore focused primarily on the goal of achieving a high surface area of active metal. Firstly, cobalt is dispersed on an oxidic carrier, such as silica, alumina or titania (Fig. 3). The basic requirements of the support material are a high surface area, high hydrothermal stability and, when applied in slurry reactors, attrition resistance. There are a number of synthetic routes that can be employed in the preparation of highly dispersed cobalt, but only precipitation and wet impregnation methods have been applied in the large-scale manufacture of commercial catalysts.44 The first cobalt catalysts (Co-ThO2-kieselguhr) were prepared by co-precipitation of cobalt and thorium from nitrate solutions, in which the kieselguhr had been suspended, by reaction with molar equivalents of base (aqueous ammonia in this case).8 Cobalt precipitates as a hydroxide and is converted to the oxide upon calcination. The radioactivity of thoria, albeit weak, led to its replacement by other non-reducible oxides such as zirconia and titania. This preparative route can be adapted easily for the synthesis of other catalyst compositions including Co–MnO, Co–MgO–SiO2 and Co–ZnO2.45–47 A second, more commonly employed approach, is to impregnate a preformed support material with a soluble cobalt precursor. Using incipient wetness or slurry (wet) impregnation techniques, the support material is first contacted with an aqueous solution of a cobalt salt, typically cobalt nitrate.44 Following impregnation, the material is then calcined in air at elevated temperature to decompose the cobalt precursor and produce highly dispersed Co3O4 crystallites. In a final activation step, the catalyst is reduced in hydrogen at high temperature. Reduction occurs in two consecutive steps passing through Co(II) oxide. To achieve high activity, all of the supported cobalt nanoparticles should ideally be larger than 8–10 nm in diameter (see Section 6.16.3), but this is difficult to achieve in practice and a broad crystallite size distribution (CSD) is often obtained using scalable methods of catalyst preparation. The support can, however, play a role in controlling CSD to some extent.48–51 Studies have shown that CSD is intimately linked to support porosity; the preparation of 8–10 nm cobalt nanoparticles requires a pore size of at least 12–16 nm.24 Intuitively, a larger pore size might be expected to allow the formation of larger crystallites, but there appears to be an upper limit. If pore size is too large, the final reduction step can lead to break-up and smaller crystallites are ultimately formed.48 The distribution and nature of surface hydroxyls on the support, its point of zero charge, the nature of the cobalt precursor and the method of calcination are among a host of other factors that can also influence dispersion. Further details can be found in several review articles.44,52 A high degree of catalyst reduction is also essential but can be difficult to achieve for highly dispersed metal oxide particles in the nanometer-sized range, particularly when strong metal-support interactions are present. These interactions impede the formation of active metal surface area during reduction in hydrogen. To counteract this, the cobalt-support interaction can be moderated through the incorporation of an additional, irreducible transition metal oxide, such as ZrO2, La2O3 or MnO.5,53,54 These oxides can provide a “protective layer” that prevents the formation of a strong interaction between cobalt and the support (Fig. 3). The control of metalsupport interactions requires a fine balancing act. If the interactions are too weak an agglomeration of cobalt particles can occur, and if the interaction is too strong, metal surface area will be lost to the formation of support compounds such e.g. silicates or aluminates. Both cases will result in a lower CO conversion on a per gram catalyst basis. To further maximize metal surface area, a second transition metal is often introduced as a reduction promoter (Fig. 3). The choice of promoter element is dictated by the surface energy of the FT metal; the promoter must exhibit a higher surface energy otherwise surface enrichment can occur.55 Thus, while copper is used as a reduction promoter of iron-based catalysts, its lower surface energy compared to cobalt leads to surface enrichment of the FT metal and the blocking of active sites.5,56,57 An effective promotion of cobalt catalysts with higher surface energy noble metals, including platinum, rhenium and iridium, has been demonstrated.5,53,56,58–60 The reduction promoter activates hydrogen, which can then spillover to the cobalt oxide and facilitate the

Fig. 3 Cobalt catalysts used in current, commercial processes (Table 1) share common design features and are not too dissimilar to those developed by Otto Roelen in the 1930s. Cobalt metal (10–30 wt%), shown in blue, is dispersed on an oxidic carrier (white). A second irreducible oxide (e.g., MnO, ZrO2 at 1–10 wt%), shown in green, controls the metal support interaction and enables a high dispersion of cobalt to be obtained. A second transition metal (e.g., Pt, Pd, Ru, Re at 0.1–0.5 wt%), shown in red, acts as a reduction promoter, facilitating the transformation of as prepared Co3O4 crystallites to active Co metal.

360

Promoted Fischer-Tropsch catalysts

transformation to Co metal.5,53,59,61,62 Spillover hydrogen can migrate across the support and so a direct contact between cobalt and promoter is not essential to obtaining the promoting effect.58,61,63 Reduction promoters are typically incorporated with a loading of 0.1–0.5 wt%, levels sufficient to promote hydrogen activation without influencing hydrocarbon production and keeping the costs of catalyst production as low as possible. The additional reduction promoter can be impregnated together with the initial cobalt solution or in a series of separate steps following calcination. The noble metal reduction promoter and surface modifying transition metal oxides are sometimes referred to collectively as structural promoters, their main function being to affect the formation and stability of the active phase.53 Further details on the preparation of supported cobalt catalysts and the role of support and structural promoters in catalyst design can be found in the open literature.5,53,54,64 A significant portion of product is liquid under LTFT reaction conditions giving rise to a three-phase solid-liquid-gas system that is handled commercially using either slurry bubble column or multi-tubular, fixed bed reactors. Early LTFT processes employed only fixed bed reactors, but this design suffers a number of drawbacks. The removal of heat from the highly exothermic FT reaction is a major issue for fixed bed operation and productivity per unit reactor volume is compromised as a consequence. Maintaining a flat temperature profile across the catalyst bed (adiabatic operation) places limits on the per pass conversion. A further constraint relates to catalyst particle size. To avoid pressure drops across the reactor, catalyst particles must be in the millimeter-sized range. Intraparticle diffusion limitations then restricts catalyst utilization and also impacts selectivity, owing to differences in rates of diffusion for H2 and CO. Slurry bubble column reactors were developed to overcome some of the aforementioned limitations.65 In slurry reactors, syngas is contacted in a bubble column with the catalyst suspended in liquid products. Efficient mixing yields superior heat and mass transfer characteristics that affords near isothermal operation and the catalyst particles are sufficiently small (typically < 100 mm) to avoid intraparticle diffusion limitations. More recent reactor developments by Johnson Matthey Davy Technologies Ltd combine the benefits of fixed bed and slurry bed reactors.66 A modular catalyst carrier device, called CANSÔ (see Table 1), is stacked inside a tubular reactor creating a series of mini adiabatic radial-flow reactors with interbed cooling. And while the reactor tube itself is 10–15 m long, radial flow through the modular units ensures that the effective catalyst bed length is only a fraction of this. Smaller catalyst particles (sub-millimeter) can therefore be used to improve activity and selectivity without incurring the pressure drop that would normally be observed in a fixed bed. The system is claimed to deliver three times the productivity of a conventional multitubular fixed bed reactor.67 Another configuration worth noting is a microchannel reactor, which also exhibits high heat and mass transport rates and is particularly suited to smaller scale applications.68 The reader is referred to the literature for further details on reactor design.5,69–72

6.16.1.3

The future role of FT in achieving net zero carbon

In the coming decades an ambitious and sustained global response to the threat of climate change is needed to protect fragile ecosystems and vulnerable people and societies. Governments, industries and organizations worldwide are pledging to adopt practices aimed at averting acceleration of global warming and avoiding a 2  C rise in global average temperature. Stabilizing atmospheric levels of CO2 will eventually require deep reductions in anthropogenic emissions from all sectors of the economy, primarily through a rapid phasing out of fossil fuel consumption. A reliance on a steady supply of carbon is one of the major challenges facing the chemical industry, which accounted for 14% of global oil demand in 2018. To reduce emissions, the chemicals sector must begin to use sources of above ground carbon, but this transition brings its own unique set of challenges relating to the marked compositional difference between fossil fuels and renewable feed streams. Furthermore, meeting climate targets within the legislated timescale can only be achieved by harnessing existing technology platforms operating at an appropriate scale. All of these challenges can be met by establishing a synthesis gas economy in which CO and H2 are used as the basic chemical building blocks for fuels and chemicals production. Primary syngas conversion processes are of course FT, but also methanol synthesis, which has a current global capacity of 100 Mt per year from approximately 100 sites operating worldwide. These processes are agnostic to the source of synthesis gas and in principle this allows for a smooth transition from the established use of fossil fuels to a future that is entirely dependent on a broad range of sustainable biomass and waste streams. The world’s first commercial-scale waste-to-liquids plant, using FT technology, became fully operational in 2021 and highlights the importance of syngas-based chemistry for the conversion of complex, carbonaceous feedstocks into useful chemical products. The new Sierra BioFuels Plant is operated by Fulcrum Bioenergy using BP and Johnson Matthey licensed technology (Table 1).67,73 The Sierra plant will convert 175,000 tonnes of municipal solid waste into 11 million gallons of low-carbon transportation fuel annually.

6.16.1.3.1

Power-to-liquids, a future beyond biomass?

The transition to a low-carbon future will be predicated on decarbonizing the production and use of energy, currently responsible for over two thirds of anthropogenic greenhouse gas emissions.74–76 To meet the Paris Agreement and Sustainable Development goals, energy supply would need to fully decarbonize by 2050, if not before. The transition from fossil fuel consumption will rely on mechanisms for dealing with the intermittency of renewable energy, such as storage in batteries or supercapacitors. Another approach is to use surplus energy for the production of hydrogen, as an energy vector, via the electrolysis of water (Power-toHydrogen). This production of green hydrogen can also play a key role in the important task of decarbonizing the chemical industry. When combined with sources of above ground carbon, the use of green hydrogen enables the electrification of fuels and chemicals production, a strategy referred to as Power-to-Liquids.77,78 And by removing carbon from the natural carbon cycle, through the

Promoted Fischer-Tropsch catalysts

361

capture and reaction of biogenic CO2 with green hydrogen, the route to eFuels and eChemicals summarized in Fig. 4 may even provide carbon-negative technologies based on an established synthesis gas platform. Synthesis gas can be generated through separate CO2 and H2O electrolysis or the co-electrolysis of CO2/H2O mixtures.79–84 There is also increasing interest in the direct hydrogenation of CO2 and significant progress has already been made in methanol production. Carbon Recycling International (CRI) constructed the first, and only (at the time of writing), commercial CO2 hydrogenation plant in 2011.78,85–87 The George Olah Renewable Methanol Plant has a production capacity of 5 million liters per year and uses geothermal electricity to produce hydrogen from the electrolysis of water (Power-to-Methanol). The construction of a 110,000 tonnes per year CO2-to-methanol plant based on CRI’s technology is now underway in Anyang, in the Henan province of China.88 The development of carbon negative processes may play a pivotal role in reaching a net zero future by offsetting the emissions of systems and processes that are more difficult to decarbonize. Combining the Power-to-Liquids strategy with the direct capture of CO2 from air opens the possibility of decentralized chemical synthesis opportunities with global reach and by-passes the need to generate biomass as a carbon feedstock, alleviating concerns about land use and food security. When paired with new solutions for sequestering carbon, this approach can do more than just reduce emissions, it may provide a crucial carbon sink. This future scenario is not without its challenges. The global adoption of a CO2-based chemicals platform will be dictated by the ability to generate low-cost hydrogen from water and a significant reduction in the energy requirements and associated costs of CO2 capture, which is currently prohibitive. An established hydrogen economy will overcome the first obstacle but the second requires a rethink in carbon capture technology.

6.16.2

Mechanism of the Fischer-Tropsch reaction

Since its discovery over a century ago, a large number of reaction mechanisms describing FT chemistry have been postulated and the topic remains one of intense debate.89 The reaction is comprised of a vast and complex chemo-kinetic network involving a large number of intermediates and elementary reaction steps that can be ascribed to one of the following five categories.2,90 1. 2. 3. 4. 5.

Reactant adsorption Monomer formation and generation of a chain initiator Chain growth Chain termination and product desorption Readsorption and further reaction

In the following sections, an overview of elementary steps and possible reaction pathways is given.

6.16.2.1

Reactant adsorption

The initial step in FTS involves the adsorption of reactant(s) on the catalytic surface, a process that is fundamental to all heterogeneously catalyzed reactions. Carbon monoxide can adsorb both molecularly and in a dissociative manner. Transition metals toward the left of the periodic table, e.g. Cr, Mn and Fe, tend to favor dissociative adsorption with C–O bond scission occurring even at room temperature. Conversely, molecular adsorption, for which the C–O bond remains intact, occurs on the later transition metals such as Co, Ni, Ru and Rh. This process is highly exothermic (160–220 kJ/mol) and from an energetic standpoint, adsorbed CO is typically found to be one of the lowest thermodynamic states.91

Fig. 4 Hydrogenative routes to CO2 conversion will rely on the availability of cheap, green hydrogen generated via the electrolysis of water. Carbon dioxide can be used directly or first converted to CO via electrolysis, providing access to established synthesis gas chemistry for electro-fuels and electro-chemicals production using only CO2 and H2O as basic feedstocks.

362

Promoted Fischer-Tropsch catalysts

For most metals, hydrogen adsorbs dissociatively via a weakly physisorbed state of dihydrogen. The facile dissociation of hydrogen is attributed to the metal–H bond being much stronger than the H–H bond. In general, the heat of hydrogen adsorption, relative to CO, increases on moving toward the late transition metals (with the exception of Fe) and this is commensurate with trends in methanation activity.2 The relatively small adsorption enthalpy provides a much greater mobility of hydrogen over the catalytic surface, resulting in spillover from the metal nanoparticles to the support, which can then act as a hydrogen reservoir.61,63 Reaction selectivity is governed partly by the concentrations of surface intermediates and this is strongly influenced by the relative heats of adsorption of CO and H. For example, an excess of hydrogen at the surface will promote the synthesis of methane over long chain hydrocarbons.

6.16.2.2

Monomer formation and chain initiation

Monomer formation and chain initiation occur via the dissociation of carbon-monoxide, which can proceed either directly, as described above, or through a hydrogen-assisted pathway as illustrated in Fig. 5. In the direct pathway, C–O bond scission occurs after adsorption of CO onto the metal, resulting in a carbon and oxygen fragment on the catalytic surface. If the barrier for direct dissociation is too high, an alternative hydrogen-assisted pathway, via intermediate HCO or COH, can be traversed. Following subsequent C–O bond scission, HCO will generate CH and O fragments, whereas intermediate COH will lead to the formation of C and OH fragments. A third possible pathway involves C–C bond formation via the adsorption of CO at an existing carbonaceous species on the catalytic surface. Carbon monoxide can, for example, bind to a co-adsorbate, such as a growing hydrocarbon chain or a chain-starter (vide infra), but also to lattice carbon that is prevalent in iron-carbides for example. Formation of this new C–C bond leads to an activation of the C–O bond that promotes bond scission. Following CO dissociation, the resulting C1 fragments (atomic carbon (C) or methylidyne (CH)) can be hydrogenated to methylene (CH2), methyl (CH3) or methane (CH4). With the exception of methane, all of these C1 species can couple to another CHx (x ¼ 0–3) fragment or directly to CO forming a C2 moiety as shown in Fig. 6. The C2 intermediates (center row, Fig. 6) can then undergo a series of reactions to form a C2 chain starter (bottom row, Fig. 6). For the C2 oxygenates, this step involves a C–O bond scission, which can be direct or hydrogen-assisted, to yield a CCHx moiety that is further hydrogenated to a CHxCH3 species. The remaining C2 species, of the general form CHxCHy, generate CHxCH3 through a series of dehydrogenation and hydrogenation steps. At this point, further chain growth can occur via the coupling of CHxCH3 with either another CHx monomer or CO to form the C3 analogs of species shown in the third row. We will refer to the C2 species shown in the third row as chain-starters.

6.16.2.3

Chain growth

After formation of a chain-starter, chain growth proceeds by further insertion of a C1 monomer, either CHx (x ¼ 0–3) or CO. Both carbon atoms of the newly formed C–C bond are usually bound to the catalytic surface. For the next chain propagation step to occur, the growing hydrocarbon chain must undergo one or more (de-)hydrogenation steps to achieve a level of coordinative saturation capable of accepting another C1 monomer. To maintain an active surface, a critical balance between the rate of C–C bond formation and the rate of hydrogenation of the growing hydrocarbon chains must be attained. If the rate of hydrogenation is too fast, the carbon atoms at the reactive center become coordinatively saturated and will leave the catalytic surface, in other words chain growth will be inhibited. If, on the other hand, the rate of hydrogenation is too slow, the catalytic surface can become carburized leading to deactivation via either surface reconstruction or catalyst poisoning.92

Fig. 5 CO activation pathways. (Top) Hydrogen-assisted CO activation via either HCO or COH as intermediates. (Middle) Direct CO dissociation. (Bottom) Carbon-assisted CO activation. Here, the carbon atom can be either part of the catalyst (e.g. as in iron-carbides) or a co-adsorbate (i.e. COinsertion mechanism).

Promoted Fischer-Tropsch catalysts

363

Fig. 6 Each of the CHx (x ¼ 0–3) intermediates (top row) resulting from CO dissociation can connect to another CHx (x ¼ 0–3) or directly to adsorbed CO to generate a C2 species (center rows). In a cascade of (de-)hydrogenation and C–O bond scission reactions, these C2 species can form C2 growing hydrocarbon chains (bottom row). Note that CH4 (top row) and CH3CH3 (middle row) are saturated hydrocarbons that will leave the catalytic surface. Ethylene (CH2CH2) could also potentially leave the surface but, owing to a sufficiently strong interaction with many transition metals, typically remains adsorbed and further reacts.

6.16.2.4

Chain termination and product desorption

The primary products of the Fischer-Tropsch reaction are a-olefins and n-paraffins that are formed in a series of hydrogen abstraction and/or addition reactions of growing chains. The route to product is dependent on the nature of the growing hydrocarbon chain. For example, in the case of a surface CR species attached to the catalytic surface by the a carbon alone, an n-paraffin can be formed by threefold hydrogenation of the a carbon. If the last step in this sequence is a b-hydrogen abstraction, rather than an a-hydrogenation, an a-olefin can be formed. These competing pathways are illustrated in Fig. 7.

6.16.2.5

Readsorption and further reaction

To add further complexity, linear olefins can readsorb at the catalytic surface and further react. These secondary reactions can have a profound impact on the product distribution giving rise to the aforementioned deviations from ideal ASF distributions (Fig. 3A). Studies exploring the effect of space velocity on product selectivity have shown that mostly a-olefins are formed at very short residence times and are therefore the dominant primary product. With increasing residence times the initially formed linear olefins are further hydrogenated to a saturated hydrocarbon that can no longer readsorb.93–95 The impact of readsorption on product selectivity will be discussed further in Section 6.16.5.

6.16.2.6

Mechanistic models

The elementary reaction steps described in the preceding section can be combined to construct a mechanistic cycle. The large number of mechanisms postulated to date can be broadly categorized as belonging to either the carbide (Biloen-Sachtler) mechanism or the CO-insertion (Pichler-Schulz) mechanism (Fig. 8).93,96 These two distinct pathways are discriminated by the mode of C–O bond activation and the nature of the subsequent chain-growth step. In the carbide mechanism, monomer formation proceeds via C–O bond scission and occurs prior to C–C bond formation in the chain-growth step. In contrast, for the CO-insertion

364

Promoted Fischer-Tropsch catalysts

Fig. 7 Primary chain-termination reactions give rise to a-olefins or n-paraffins. A growing hydrocarbon chain C–CH2–R can undergo a two-fold hydrogenation at its a-position. Thereafter, further a-hydrogenation will produce the saturated product while b-dehydrogenation will yield the corresponding a-olefin.

mechanism, adsorbed CO acts as the primary monomer and directly couples to the growing hydrocarbon chain. The C–O bond is then broken after the C–C bonding forming reaction.

6.16.2.6.1

Carbide mechanism

The carbide mechanism, originally proposed by Fischer and Tropsch, is named for the existence of surface carbides, which act as the monomeric building blocks in the polymerization process.1 A large body of experimental and theoretical work further supports the carbide mechanism and the now prevailing view is that it accounts for the main growth pathway on Co, Fe and Ru catalysts.97–102 Mechanistically, it proceeds by the formation of a monomeric C1 carbide species formed via the direct dissociation of CO. The resulting C atom can be hydrogenated or remain as atomic carbon on the surface. This CHx species (x ¼ 0–3) can then couple to a growing hydrocarbon chain. In Fig. 8, this is illustrated for a CH monomer unit coupling to a CR species representing the growing hydrocarbon chain. After C–C coupling (Fig. 8, step 1), both carbon atoms of the growing hydrocarbon chain, as well as the monomer species that has coupled to it, remain bound to the catalytic surface. For the growing hydrocarbon chain to accept the next monomer unit, it is assumed that a series of hydrogenation and dehydrogenation steps must first occur. For the specific example shown, the b-carbon needs to be hydrogenated twice (steps 2 and 3) after which it becomes coordinatively saturated and the M–C bond is broken (step 4). This process is driven enthalpically by the favorable van der Waals interaction between the hydrocarbon chain and the wax layer that has formed over the catalytic surface, and entropically by the sizeable gain in configurational freedom. Finally, the a-carbon is dehydrogenated once (step 5) to form a CR species that allows for another CH þ CR coupling step to occur and the chain growth to continue. It should be emphasized that the example given is just one of many possible chain-propagation routes that could potentially proceed according to the carbide mechanism. The principal monomeric unit (CHx, where x ¼ 0–3) and the growing hydrocarbon chain (CHxR, where x ¼ 0–2) can be one of several structures with varying C:H ratios, giving rise to a complex reaction network. Theoretical studies also indicate that the chain-propagation mechanism depends on the length of the growing hydrocarbon chain as subtle differences in the distribution of electron density as a function of the dangling chain length are known to affect the height of the transition states for C–C coupling in a profound manner.15 For example, it is found that CH þ CH coupling has a higher activation barrier than CH þ CCH3 coupling. The substitution of hydrogen for a good s-donating methyl group leads to an enhanced electron density between the two fragments that facilitates bond formation. The propensity toward bond formation can also be tuned by modifying the electronic nature of the catalytic surface by means of adding promoters. For example, an experimental study has shown that by careful incorporation of Na and S promotors in an Fe catalyst, CH4 and C5 þ selectivity can be decreased in favor of producing C2–C4 olefins.103 The reaction pathways discussed thus far describe the production of linear hydrocarbons, but it is known that some branching (predominantly monomethyl branching) also occurs. Branched product can be accounted for by the coupling of a C1 monomeric unit with a CHxCHyR species (x ¼ 0–2, y ¼ 0–1 and R is a saturated hydrocarbon chain). If the C1 monomer connects to the a carbon, (CHx), subsequent hydrogenation will produce internal olefins. Conversely, coupling of the C1 monomer to the b carbon, (CHy) will lead to the formation of branched hydrocarbons in the product mixture. Although such products are observed, they are minority species and so the probability of these reactions occurring is relatively small.

6.16.2.6.2

CO-insertion mechanism

The CO insertion mechanism proposed by Pichler and Schulz (Fig. 8B) was inspired by known CO-insertion processes from organometallic chemistry.93 The primary mode of C–C coupling is the insertion of molecular CO into a growing hydrocarbon chain. Chain initiation will proceed via either direct or (H,C)-assisted CO dissociation as adsorbed CO species are not found to couple directly. After CO insertion (step 1 in Fig. 8B), C–O bond scission occurs (step 2). To accept the next CO monomer in the chain-propagation step the b-carbon must first be hydrogenated twice (steps 3 and 4). Finally, the a-carbon is dehydrogenated once (step 5) to form a CR species that allows for another CR þ CO coupling step to occur and so the chain continues to grow.

Promoted Fischer-Tropsch catalysts

365

Fig. 8 (A) Monomer-insertion and activation according to the carbide mechanism. Following direct CO dissociation, the resulting C1 species inserts into a growing hydrocarbon chain. (B) According to the CO-insertion mechanism, CO couples directly to a growing hydrocarbon chain and only thereafter is the C–O bond broken.

366

Promoted Fischer-Tropsch catalysts

The example shown in Fig. 8 is one of several possible pathways by which the CO-insertion mechanism can proceed. Here, the growing hydrocarbon is represented by CR, but in principle any CHxR (x ¼ 0–2) species can accept a CO molecule.

6.16.3

Structure sensitivity

The concept of an active site was introduced almost a century ago, a term that refers to a specific contiguous array of surface atoms that act as atomic-scale docking sites, where reactants bind temporarily, and physical or chemical transformations are promoted. The importance of defect sites, formed from low-coordinated surface atoms with unsaturated valencies, was first postulated by Taylor.104 Commonly observed variations in catalyst activity with particle size, which led to the classification of “surface-sensitive” reactions, provided evidence for the role of such defects. Ertl’s ground-breaking studies of NO dissociation on Ru(0001) found in favor of Taylor’s model.105 The now prevailing view recognizes the requirement of the active site to be composed of a combination of undercoordinated edge and more highly coordinated lower step atoms for facile scission of diatomic molecules (i.e., the transition state straddles the step site from edge atoms to bottom atoms). A growing body of evidence reveals the existence of such structure-sensitivity in the FT reaction. Data obtained from numerous experimental studies show a strong link between intrinsic catalytic activity and crystallite size for cobalt-catalyzed syngas conversion (Fig. 9). The turnover frequency (TOF) is found to increase sharply with increasing nanoparticle size up to  5–10 nm after which the activity reaches a plateau. Crystallite size is typically measured prior to catalyst reduction and therefore does not reflect the true state of the active surface. Pre-treatment of the oxidic catalyst precursor in hydrogen is an essential step in generating the active catalyst, but can lead to a variable degree of reduction, the formation of irreducible mixed oxides, such as aluminates or silicates, and the coverage of active metal surface by suboxides in so-called strong metal support interactions. These support effects contribute more greatly to the structure of very small nanoparticles and can therefore hinder the investigation of structure-sensitivity effects because a significant irreducible component creates uncertainty about the true particle size.115 The dramatic drop in catalytic activity below  5–10 nm may simply be a consequence of small crystallites being more difficult to reduce. Fundamental studies aimed at addressing intrinsic particle size effects therefore require model materials in which contributions from the support are minimized.106,115 The correlation between particle size and activity shown in Fig. 9 is also observed for systems that are specifically designed to minimize interactions between metal and support,106,115 including the use of inert carbon as a carrier.117 This important result indicates that the general trend shown in Fig. 9 is not caused by metal-support interactions.100 Early evidence of structure sensitivity in the Fischer-Tropsch reaction was provided by studies of CO dissociation on well-defined  single crystals.118–121 Under FT-relevant conditions, the activation of CO was found to be more facile on the stepped Coð1012Þ 121 surface compared to that of the close-packed Co(0001) surface. Spectroscopic data also revealed a very low CO stretching frequency (1285 cm 1) indicating a significant lengthening of the CO bond; this was assigned to CO adsorbed in a step site on  surface. Insights into the exact nature of sites responsible for C–O bond scission can be traced to early investigations the Coð1012Þ of the crystallite size dependence of infrared active N2 adsorption on Ni, Pd and Pt. Dinitrogen is infrared inactive in the gas phase, but sufficient polarization of the N^N bond occurs upon adsorption on a metal surface and an N^N vibrational mode can be observed, but seemingly only for larger nanoparticles.122,123 To explore the origin of this apparent size effect, models of nanoparticles were constructed using marbles with the object of counting the number of different surface sites as a function of crystallite

Fig. 9 The turnover frequency (TOF) for CO consumption as a function of cobalt nanoparticle size for the Fischer-Tropsch reaction. The dashed line serves to guide the eye.106–116

Promoted Fischer-Tropsch catalysts

367

size.122 The models revealed that small crystallites cannot sustain B5 sites, a term that refers to a unique topology created by five metal atoms localized at step edges (Fig. 10A). On this basis, the absence of the N^N vibrational mode on materials comprising small crystallites was attributed to the absence of B5 sites. Many years later, surface science and computational studies confirmed the role of B5 sites in the dissociation of N2 and NO.105,124 This structure dependence of dissociative adsorption is now recognized as a common activation pathway for p-bonded diatomic molecules. Evidence for the activation of CO at B5 sites is provided by computational studies, which reveal a significantly lower activation energy of 65 kJ/mol for dissociation on the corrugated  surface of Ru compared with 210 kJ/mol at the Ru(0001) terrace.125 Similar results are found on Co surfaces where direct ð1121Þ  surface than the most facile pathway on the flat C–O bond scission is about 100 kJ/mol lower in energy for the corrugated Coð1121Þ 125–128 Co(0001) surface. In light of these findings, an inability of smaller nanoparticles to sustain sites capable of CO activation provides a logical interpretation of the observed tail off in FTS activity shown in Fig. 9. The nature of active sites on the surface of a nanoparticle critically depends on its size but quantifying the surface density of these sites is challenging. Whereas pioneering studies into this topic used marble models to enumerate the surface density of active sites,122 modern approaches use physically based force fields. This force field approach is far superior as it accounts for chemical bonding principles in addition to geometric factors. Several recent studies have clearly demonstrated that the surface density of a step-edge site, necessary for facile CO dissociation, increases with particle size up to a critical point, after which the surface density remains constant (Fig. 11).129–132 The correlation between the data presented in Fig. 11 and experimental observations shown in Fig. 9 is remarkable.106–116 The computational studies further reveal that the observed trend reflects the ability of metallic

Fig. 10 (A) CO adsorbed on a B5 step-edge, highlighted in orange, showing a tilt that is predicted by density functional theory calculations. (B) Co nanoparticle of 6 nm generated by simulated annealing. The colors indicate different active sites: red atoms are threefold fcc(111)-type nanoislands, blue atoms are fcc(100) type nano-islands and green for the center atom of highly-active step-edge sites. A B5 site near the center of the nanoparticle is highlighted in orange (only a single B5 site is shown for clarity). Such sites are located at the edges of the nano-islands.

Fig. 11 Number density of the surface atoms comprising a step-edge site as a function of the nanoparticle size. Several published models predict an increased surface fraction of step-edge sites with increasing nanoparticle size up to  5–6 nm. Thereafter, the number density reaches a plateau and is invariant with particle size.

368

Promoted Fischer-Tropsch catalysts

nanoparticles to stabilize nano-islands (see Fig. 10B), whose edges are the loci of highly active sites.129 It is worth noting that the data presented so far considers a nanoparticle to be static, whereas experimental evidence suggests that under typical operating conditions the catalytic surface is dynamic and continuous restructuring is possible.133–138 The role of B5 sites in FTS has gained a lot of traction in recent years, but other perspectives on the origins of structure sensitivity do exist and are worth noting. An alternative view relates to inefficient H2 activation rather than an apparent lack of sites capable of activating CO.112,139 The dissociation of CO by a H-assisted process, originally proposed by King,140 first requires the generation of H* at surfaces that contain the contiguous arrays of atoms essential for hydrogen activation. The probability of finding such large ensembles of cooperative sites is expected to decrease rapidly for smaller particles. Another opposing view argues that defect sites, such as B5 sites, bind intermediates strongly, rendering them inactive at the high surface coverages present under catalytic conditions. As a consequence, CO activation must take place over planar surfaces via hydrogen assisted routes.141 Recent DFT calculations, however, have shown that even at high coverages the B5 site offers a more facile route for C–O bond activation than is achievable at terrace sites.142

6.16.4

Kinetics of the Fischer-Tropsch reaction

Gaining a fundamental understanding of the Fischer-Tropsch reaction is a critical step toward process optimization and the discovery of new catalytic materials. Unraveling the reaction mechanism can be achieved through detailed kinetic modeling, but this is far from trivial owing to the vast chemokinetic network comprised of a large number of elementary reactions and intermediates. Many kinetic models, based either on experimental results or derived from theory, have nevertheless been made.3,143,144 They differ greatly in their level of complexity and the amount of useful information that can be gleaned from them. In the following section a brief overview of the most common types of kinetic models is given, starting from the simplest empirical approaches and working toward the more complex microkinetic models. The insights provided toward our understanding of the kinetic mechanism is also discussed.

6.16.4.1

Kinetic modeling

The simplest kinetic models use an empirical power-law expression and consider only the rate of CO consumption, as a function of temperature, and the partial pressures of H2 and CO.91,145,146 A general rate expression for this approach is given by: y

r ¼ kðT ÞpxCO pH2

(8)

where r is the rate of CO consumption, k is a temperature-dependent rate constant (typically an Arrhenius-type expression), pCO and pH2 are the partial pressures of CO and H2 respectively, and x,y denote the reaction orders of the gaseous substrates. The coefficients k(T), x and y are determined from mathematical fitting of experimental data. In this empirical approach, the surface composition and the role of reaction intermediates is not explicitly taken into account. The relative simplicity of these kinetic expressions limits their application to small operating windows of temperature and pressure. More complex kinetic models are based on Langmuir-Hinshelwood-Hougen-Watson (LHHW) or Eley-Rideal type rate equations in which a rate-determining step is assumed and surface intermediates are modelled under a pseudo-equilibrium and mean-field approximation.147 Such models account for the temporary poisoning of the catalyst surface caused by adsorbates. If the rate equation is derived from a series of elementary reaction steps, the rate-determining step assumption allows for the construction of an analytical or closed-form expression, avoiding the need to perform more complicated numerical computations. In FTS, the ratedetermining step is often assumed to be monomer formation and the functional expression has the following form: r ¼ kðT Þ

KCO pCO x KH2 pH2 x P 1 þ Ki pi

(9)

where K corresponds to an equilibrium constant of an adsorption/desorption step. Expression (9) can be interpreted qualitatively as follows: the numerator reflects the chance of two surface intermediates participating in a reaction event, which scales linearly with their surface concentration. The denominator on the other hand, contains an inhibiting term that corresponds to the composition of the catalytic surface. For example, if the surface is completely empty, the denominator would be equal to unity. If the surface is instead covered to a large extent by intermediates that occupy sites required for bond breaking, this term will be much greater than unity giving a lower overall reaction rate, r. A detailed explanation of this type of modeling can be found in several textbooks.148,149 The most advanced kinetic models developed to date are microkinetic models. Rather than assuming one or more ratedetermining steps, which reduces the applicability of the kinetic expressions to a relatively small set of operating conditions, no such assumption is made and each elementary reaction step is explicitly taken into account.149 The rate constants of the elementary reaction steps are derived from transition state theory and given by the Eyring equation150:   kT DGact k ¼ b exp  (10) h kb T

Promoted Fischer-Tropsch catalysts

369

where kb is Boltzmann’s constant, h is Planck’s constant, DGact is the Gibbs free energy of activation and T the temperature. For a system of N elementary reaction steps, a set of 2N coupled ordinary differential equations (ODE) is obtained. Considering the vast differences in the magnitude of rate constants for the different elementary reaction steps involved in the mechanism, such sets of differential equations require special numerical solvers known as stiff solvers, such as the backward differentiation formula (BDF) method.151 The underlying parameters of microkinetic models, i.e. the Gibbs free energies of activation, can be found from kinetic experiments, but are typically derived from quantum chemical calculations using density functional theory, for example. Catalytic activities can be calculated by time-integrating the differential equations until a steady-state solution for the surface intermediates is found. This provides access to the rate of change of reactants and products as well as their surface concentration. Furthermore, by adjusting individual model parameters (e.g. rate constants, interaction energies of surface intermediates, pressure and temperature), a sensitivity analysis can be performed and affords invaluable insights into the critical aspects of the chemokinetic network and the adaptations of catalyst structure needed to improve activity and selectivity toward desired products.99,149,152 The chemokinetic network of the FT reaction is incredibly complex, requiring an evaluation of  500 differential equations containing approximately 1000 components.99

6.16.4.2

Mechanistic insights

Improvements in process efficiency, and the development of new FT processes that target specific fuels and chemical products, will be driven by a deeper understanding of the fundamental factors controlling reaction selectivity. Kinetic and microkinetic modeling can assist in elucidating the underlying reaction mechanism. However, accounting for the entire, complex chemokinetic network requires the generation of large data sets using prohibitively expensive quantum chemical calculations. One approach to simplifying this problem, and reducing the computational burden associated with these modeling efforts, is to consider sets of elementary steps and their associated reaction rates bundled together. A microkinetic model based on these so-called lumped-sum kinetic expressions can be used to identify the factors that define chain-length selectivity, i.e. the magnitude of a.97 These sets of lumped steps are shown schematically in Fig. 12 and consist of:

• • • •

Monomer formation to convert CO to CHx. Methane formation. Chain-growth via CHx þ CHy coupling and the (de-)hydrogenation steps of the growing hydrocarbon chain. Chain termination via the formation of olefins or paraffins.

With this approach the following criteria for the formation of longer chain hydrocarbons can be established: 1. CO activation and monomer formation needs to be much faster than methane production and chain-termination. This is a universal condition required to form longer hydrocarbons. 2. A high coverage of CO under reaction conditions may still occur if the rate of chain-growth is much faster than the rate of monomer formation. Conditions 1 and 2 define the monomer-formation limiting regime. 3. The rate of chain-propagation is slow compared to monomer formation and the surface is predominantly covered by growing hydrocarbon chains. In conjunction with condition (1), this is termed the growing-chain limiting regime.

Fig. 12 Lumped chemokinetic network of the Fischer-Tropsch reaction. All elementary reaction steps are gathered into four sets of kinetic expressions, involving (1) monomer formation, (2) chain-growth via CHx–CHy coupling and chain (de-)hydrogenation reactions, (3) methane formation and (4) chain termination by formation of paraffins or olefins.

370

Promoted Fischer-Tropsch catalysts

These conditions only ensure a high chain-growth probability. To further account for a high activity, i.e. a high rate of CO consumption, a low activation energy for monomer formation, in conjunction with an intermediate thermodynamic stability of carbonaceous intermediates, is required. Current generation Fischer-Tropsch catalysts operate in a monomer-formation limiting regime and exhibit a high chain-growth probability. There is, however, a clear opportunity to improve the rate of CO conversion while maintaining the high selectivity to longer chain hydrocarbons. This simplified model identifies avenues for catalyst and process improvement, but does not provide insights into how this can be achieved. Insights into the factors that govern a high CO conversion, while maintaining a high chain-growth probability, can be gained through microkinetic simulations based on calculated molecular reactivity data. Such models can establish the stability of surface intermediates and their rate of formation and consumption for complex reaction networks. However, building such elaborate microkinetic networks using density functional theory calculations is computationally demanding. It is possible, however, to calculate reaction and activation energies for a given transition metal using known values established for another transition metal by use of so-called scaling relations, assuming the two metals share a similar geometry for their active sites.99,153,154 Procedurally, one first calculates the kinetic parameters of the relevant elementary reaction steps for a single transition metal and a single configuration of the active site. An extrapolation to other transition metals can then be made using an appropriate set of scaling relations. The strength of the adsorbate-surface chemical bond governs the stability of reactive intermediates and controls their fate. Thus, the metal-carbon and metal-oxygen interaction energy, which describe the propensity of a metal to bind to surface intermediates, serve as ideal indicators of reactivity. Catalysis is a kinetic phenomenon and therefore a second relationship is necessary to determine how changes in the overall thermodynamics affect the kinetic barriers. Here the Brønsted-Evans-Polanyi relation can be applied. This well-established correlation is used in heterogeneous catalysis to associate changes in the relative difference between initial and final state to changes in the height of the transition state by assuming that the curvature of the potential energy around the initial and final state remains constant.153 Following this strategy, the turnover frequency and chain-growth probability can be determined. The results are shown in the two-dimensional volcano plot (Fig. 13), in which the optimum conditions for the formation of longer-chain hydrocarbons are found near the center and much poorer conditions are seen at the edges. The origin (0,0) of this heat map pertains to the  surface from which all other data points in the plot are derived. Fig. 13 also highlights three distinct reactivity regimes Ruð1121Þ that distinguish the rate determining step for the overall FT reaction. These three regimes are discussed below. 1. Monomer formation regime: The rate of formation of CHx monomers is much slower compared to all other elementary reaction steps. This occurs when the metal has a weak interaction with both carbon and oxygen, giving rise to a high barrier for C–O bond scission. Depending on the propensity of the metal to perform hydrogenation reactions, this regime can lead to the predominant formation of short chain oxygenates, methane or methanol. 2. Chain-growth regime: The rate of C–C bond formation or the rate of (de-)hydrogenation steps that allow for the insertion of the next monomer is ratelimiting. This occurs when the metal-carbon interaction is strong, but the metal-oxygen interaction is weak. Despite facile C–O bond

Fig. 13 Reaction rate for CO consumption (A) and chain-growth probability   (B) as function of the metal-carbon and metal-oxygen bond strength.  from which all other data points in the plot are scaled using the metalThe origin (0,0) coincides with the FTS activity and selectivity for Ru 1121 carbon and metal-oxygen bond strength. Adapted from Filot, I. A. W. A.; van Santen, R. A. A.; Hensen, E. J. M. J. Angew Chem. Int. Ed. Engl. 2014, 53 (47), 12746–12750.

Promoted Fischer-Tropsch catalysts

371

scission, the subsequent formation of C–C bonds is difficult as the metal favors carbon in its atomic form. Metals that operate within this regime have a tendency to form carbides. 3. Water formation regime: The rate of oxygen adatom hydrogenation to form water is rate-limiting. This happens when the metal-oxygen interaction is very strong and the metal therefore favors oxygen in its atomic form. This leads to oxygen poisoning (when oxygen is adsorbed on top) or oxidation of the metal surface (when oxygen occupies interstitial sites). It should be emphasized that the monomer formation and chain-growth regimes were already found using the simplified lumped-sum model whereas the full microkinetic model described here reveals the existence of a third regime associated with the rate of removal of oxygen adatoms, resulting from C–O bond scission, in the form of water. The behavior of different FT active metals can be compared on the basis of their metal-carbon and metal-oxygen bond strengths (Fig. 13). Metallic iron lies at the interface between the chain-growth limiting and oxygen removal regimes. It exhibits a very strong interaction with both C and O, giving rise to facile CO bond scission. However, subsequent CHx formation and the removal of O is difficult. Consequently, metallic iron tends to carburize under CO hydrogenation conditions and is transformed to an iron-carbide, which is broadly viewed as the active phase in Fe-based FTS. 4,29,32,34–37 Cobalt lies in the monomer-formation limiting regime. It displays a more difficult C–O bond scission when compared to the subsequent C–C coupling steps.101,126,140,155–157 Ruthenium on the other hand lies in the oxygen-removal limiting regime. It exhibits a much more facile C–O bond scission compared to Co, but removal of the resulting oxygen after CO activation is difficult. The 2D descriptor model also enables us to rationalize the behavior of metals such as Ni and Rh under CO hydrogenation conditions. Both of these metals lie in the monomer formation regime, but give rise to a different FT product slate. Lying to the right-hand side of the d-block, Ni has a relatively weak M–C and M–O interaction strength. It exhibits a fast rate for CO dissociation, but an even faster rate for CHx hydrogenation, which far exceeds its rate of C–C coupling. As such, Ni mainly produces methane. Rhodium lies between Ru and Pd in the periodic table, which have lower and higher barriers for CO dissociation, respectively. Consequently, Rh displays an intermediate rate for CO bond scission and a unique selectivity toward short-chain hydrocarbons and oxygenates. From Fig. 13 it can be seen that Co and Ru show the highest rate for CO conversion followed by Fe. Interestingly, the same figure also reveals that Co, Fe and Ru are suboptimal and opportunities therefore exist for improving catalyst performance. Consistent with the Sabatier principle, the optimal kinetic conditions for any reaction is governed by the competition of several elementary reaction steps rather than a single step. The model behind Fig. 13 sheds some light on the electronic modifications needed to mitigate the limiting conditions and shift the position of the transition metal closer to the theoretical optimum. Prominent examples of such modifications are the use of promoting elements or alloying. For example, it is known from experiment that the introduction of potassium to iron-based FTS catalysts results in an increased binding strength for CO and an increase in the turnover frequency for CO consumption.158,159 The enhanced binding of CO corresponds to a stronger M–C bond and a shift further to the left of the volcano plot. This brings Fe closer to the theoretical optimum and hence a higher activity is predicted. This microkinetic model has also been used to shed light on the long-standing debate surrounding which of the two prominent FTS mechanisms is dominant. Both the carbide and CO-insertion mechanism are accessible pathways in the overall chemo-kinetic network and could potentially occur simultaneously. Quantum chemical calculations have shown that the C–C coupling steps in the carbide pathway are energetically more favorable than the competing route in the CO insertion pathway by over 100 kJ/mol or by 11 orders of magnitude in terms of rate.160 This observation is commensurate with the results of kinetic modeling, which suggests that less than 1% of the total CO consumption proceeds through the CO insertion pathway.99 More recent microkinetic modeling based on extensive experimental data adds further credibility to the dominance of the carbide mechanism.101 Advances in computational catalysis are playing an increasingly important role in our understanding of complex reaction networks and how they operate as a function of surface structure. Microkinetics simulations that are based on calculated molecular reactivity data can establish the stability of surface intermediates and their rate of formation and consumption for complex reaction networks. It should be emphasized, however, that these models have inherent limitations. They assume that the catalytic surface is static and does not reorganize under catalytic conditions.149 Furthermore, because of the mean-field approximation, the exact topology of the adlayer is neglected.149,161,162 And finally, these models often disregard the effect of metal support interactions.163 It is expected that continued development of microkinetic simulations, with the aid of machine learning strategies, will overcome these limitations, opening up possibilities toward full multiscale modeling.

6.16.5

Fischer-Tropsch to chemicals (Chem FT)

Developments in commercial FT processes have focused largely on the production of liquid fuels, but further opportunities exist for the synthesis of high-value chemicals, in particular short-chain olefins and higher (C2þ) alcohols. Both are produced in current FT processes and the economic viability of their extraction is governed by volumes of production. The Sasol Synfuels HTFT process was designed to maximize petrol production and includes a C4 oligomerization unit that yields high octane number hydrocarbons.164 The FT tail gas is cryogenically separated to obtain this olefin feed stream and this also enables the extraction of C2 and C3 hydrocarbons. To capitalize on chemicals production, however, any process designed specifically for the targeted synthesis of light olefins would need to operate at a much higher C2–C4 selectivity. Fischer-Tropsch product distributions are bound by the kinetic constraints of a step-growth polymerization process and a greater control of selectivity therefore remains one of the biggest

372

Promoted Fischer-Tropsch catalysts

challenges in FT chemistry. In the following sections, a brief overview will be given of FT routes to lower olefins and higher alcohols from syngas. Particular emphasis will be placed on the technical challenges and strategies that can be employed to overcome them.

6.16.5.1

Lower olefin synthesis

Lower olefins (C2–C4), also referred to as light olefins, are major chemical feedstocks used in the manufacture of a broad range of products including polymers, solvents, pharmaceuticals and synthetic textiles. These commercially important platform chemicals are currently produced from the steam cracking of crude oil–derived naphtha, fluid catalytic cracking of heavy gas oil or alkane dehydrogenation.165–167 All of these processes are of course reliant on a steady supply of fossil fuel-derived hydrocarbons and alternative strategies are needed for the production of light olefins from sustainable feedstocks. Synthesis gas-based routes offer a viable alternative that also affords great flexibility with respect to the range of carbonaceous feedstocks that can be used. There are several possible pathways to lower olefins from syngas, the majority requiring the production of an intermediate such as methanol or dimethyl ether (Fig. 14). Fischer-Tropsch synthesis can, in principle, provide a direct route to lower olefins and requires the development of bespoke catalysts, coupled with optimized reaction conditions that can disrupt the “normal” ASF distribution. Olefin formation is known to dominate the primary pathways to FT products. However, readsorption leads to secondary reactions that destroy an initially high selectivity.18,95,168 Secondary reactions include hydrogenation to the corresponding paraffin, incorporation into growing chains and oxygenate formation. To maintain a high selectivity to olefins, an effective suppression of secondary reactions must be achieved. Several process parameters provide convenient handles for controlling the extent of readsorption and secondary reactions. At sufficiently high space velocity, a shortened residence time ensures that the selectivity defined by primary reactions is retained, but there is a trade-off with low per pass conversion and the requirement for large volume gas recycle limits the commercial viability of this approach. Product composition will be governed ultimately by the availability of surface intermediates and the precursors from which they are formed. Selectivity can therefore be controlled through the regulation of carbon and hydrogen supply within the catalytic system. The simplest approach is to lower the H2/CO feed ratio and starve the catalyst surface of hydrogen. This has a pronounced, positive impact on olefin selectivity, but there are practical limitations to the extent to which syngas ratios can be manipulated. Furthermore, at low partial pressures of hydrogen an excessive formation of hard carbon can cause severe swelling of a catalyst bed and/or lead to eventual deactivation. Promotion with alkali metals has a similar net effect to operation in low pH2 regimes. Potassium in particular is incorporated into the structure of commercially operated iron FT catalysts to achieve target activity.21,169–171 The alkali metal is an electronic promoter, it affects the local electronic structure of the FT metal, altering the chemisorption behavior and the surface coverage of reactants and intermediates formed from them.53,172 An enhanced adsorption and activation of CO, resulting from greater back

Fig. 14 Routes to lower (light) olefins from synthesis gas. The majority of possible routes are indirect, requiring the production of an intermediate that is transformed in a second step to olefins. The direct production of lower olefins is possible through the FT route using specially designed catalysts that maximize selectivity. This is often described as Fischer-Tropsch to Olefins or FTO.

Promoted Fischer-Tropsch catalysts

373

donation of electron density to the 2p* orbital, increases the coverage of surface carbon relative to the supply of hydrogen. This leads to an increased selectivity to olefins, a shift toward higher molecular weight products and a concomitant reduction in methane formation. An over-supply of carbon caused by the electronic promoter can also increase the extent of carbidization and carbon formation.30,173 Targeting the production of lower olefins is not simply a question of suppressing secondary reactions, it also requires an effective control of chain growth. Reaction temperature has the biggest impact on the molecular weight distribution of products (Fig. 2). In a low temperature operating regime, a chain growth probability of 0.95 is achievable, yielding mostly C21þ hydrocarbons. For high temperature operation, the chain growth probability drops to 0.70–0.75 and the lighter product formed is more suited to gasoline and chemicals production. According to an ideal ASF distribution, a maximum in C2–C4 yield of 58% can be attained with a chain growth probability of 0.4–0.5 (Fig. 2B) and would require even higher operating temperatures than those used commercially today. Increasing the reaction temperature is not a workable approach, however, owing to the concurrent increase in methane formation to just under 30% at the optimum chain growth probability for C2–C4 hydrocarbons.174 The fixed relationship between C1 and C2–C4 yields must therefore be disrupted and is the main hurdle to the development of a commercially viable FTO technology. To break the ASF distribution, much attention has focused on the effects of chemical promotion. Potassium is essential to achieving high turnover with Fe-based catalysts and while olefin formation is also promoted, the alkali metal causes an increase in the average molecular weight of hydrocarbon products rather than the desired decrease. A negative deviation from an ASF product slate can be obtained, however, through the combined effects of potassium (or sodium) and sulfur in Fe-based FT catalysts.16,34,174– 189 This might seem counterintuitive given that sulfur is usually viewed as a catalyst poison. Syngas prepared via the gasification of coal contains a host of sulfur-containing compounds including H2S and volatiles such as carbon disulfide, thiophene, mercaptans and thioethers.183 Early developments in FTS research were dedicated to understanding the impact of sulfur on catalyst performance.190 Fischer reported a practical upper limit of 1–2 mg/m3 for the concentration of sulfur in syngas and most commercial processes operate well below this.190 At levels above this threshold limit poisoning occurs as recognized by a loss of activity and shortened catalyst lifetime, necessitating rigorous clean-up of syngas feed in commercial processes.21,26,191 At sufficiently low levels of sulfur a promotional effect emerges and was first recognized in a patent filed by I.G. Farbenindustrie Aktiengesellschaft in 1929, which claimed that FT catalysts containing alkali promoters and small levels of sulfur extended catalyst lifetime, reduced the production of high boiling hydrocarbons and enhanced olefinic content.192 Several other early studies also revealed an enhanced activity in both fused and precipitated iron catalysts in the presence of low levels of sulfur.190,193–195 A large body of research has since been dedicated to the optimization of sulfur promotion for light olefins synthesis and to understanding the origins of the promotional effect.34,174–178,180,185–189 High olefin selectivities have been reported over the years, but should always be viewed in light of the reaction conditions employed, which are often selected to maximize the olefin/paraffin ratio and may not be practical for commercial operation. More recent work, exploiting high-throughput experimentation for catalyst discovery and optimization, has led to the development of an iron FTS catalyst that affords high C2–C4 olefin selectivity (68% of olefins in the C2–C4 fraction) and a simultaneous suppression of methane formation under industrially relevant operating conditions.174 The combined promotional effects of alkali metal/sulfur have been confirmed in numerous studies but the origin of shifts in selectivity remains unclear. Postulates tend to focus on describing the possible origins of a regulated hydrogen supply that will limit both hydrogenation and C–C propagation steps. These include the blocking of hydrogenation sites by sulfur, favoring chain growth termination via b-hydrogen abstraction,189 a modification of the binding geometry of the active iron carbide surface,196 and an attenuation of a hydrocarbonaceous overlayer that is believed to facilitate hydrogen supply to active sites.179,180,197 Sulfur may also influence the electronic characteristics of the catalytic surface34; while the addition of alkali elements enhances electron density, the presence of sulfur will have the opposite effect.183,198 The promotion of FT catalysts by alkali metals and sulfur is described in more detail in several review articles.183,199,200 Iron-based catalysts exhibit a lower hydrogenation capacity when compared to cobalt and have therefore featured predominantly in Chem-FT research. A recent study, however, has revealed that a high selectivity toward light olefins can also be achieved with cobalt-based materials.201,202 Manganese promoted Co2C nanoprisms exhibit high FTS activity with  60% C2–C4 olefin and only 5% CH4 selectivity under mild reaction conditions.201 This is a remarkable result when considering that cobalt carbide is normally viewed as being inactive for FTS203 and its formation has even been attributed to catalyst deactivation. The key here is particle shape, the preferentially exposed (101) and (020) facets of the nanoprisms might possess particular active sites for syngas conversion to olefins. This unexpected result highlights the importance of controlling surface structure in catalysis and provides a foundation for new avenues to be explored in the design of next generation FTO catalysts. Despite the large volume of research dedicated to FTO spanning several decades, there has been no commercial application for this process. High C2C4 olefin yields can be obtained at low per pass conversion and attempts to increase reaction rate destroy the desired selectivity. This may be an inherent constraint; if a regulation of hydrogen supply is key to selectivity control but also negatively impacts rate then achieving both high selectivity at high conversion may not be possible. The formation and build-up of carbon and a significant conversion of CO to CO2 observed in many cases are additional complicating factors. The main challenge in FTO process design remains; the decoupling of C1 and C2–C4 yields at commercially acceptable reaction rates. Tailoring the design of known FT catalysts is likely to afford only incremental improvements in performance rather than the much needed step-change. An entirely new strategy is necessary and there has been notable recent success in the development of multifunctional catalysts that integrate FT active sites with complimentary properties. The coupling of FTS with acid catalyzed cracking, for example, can provide an effective means of regulating MW distribution.85,188,204–206

374 6.16.5.2

Promoted Fischer-Tropsch catalysts Higher alcohol synthesis

Higher alcohols have a diverse range of applications including their use as solvents (C2–C5), either directly or in the preparation of derivatives (e.g. acetates, ethers and acrylates), and as raw materials for the production of plasticizers (C7–C13) and detergents (C12– C18). These commercially important compounds are prepared by the fermentation of sugars (ethanol, isobutanol), the esterification and hydrogenation of triglycerides obtained from natural production extraction (C12–C22) or via the hydroformylation of alkenes, a major route to C3–C15 alcohols. In principle, the entire range of valuable higher alcohols can be produced via FTS. Virtually any hydrocarbonaceous material can be converted to syngas, providing a platform for the sustainable production of alcohols using sources of above ground carbon and avoiding the need for fossil-derived alkenes. The initial reports of Fischer and Tropsch described the production of mainly oxygenates from syngas,207 but at the time these were unwanted products and further developments focused on shifting the selectivity to hydrocarbons.10,16 Catalysts in commercial operation today still produce oxygenates to some extent, primarily n-alcohols and n-aldehydes but also secondary alcohols, esters, acids and methyl ketones.208 The levels of these compounds can be easily under-estimated owing to the complexities of identifying their presence in olefin/paraffin rich mixtures.209 The CO insertion mechanism provides a direct route to oxygen-containing surface species of the general form R–CHx–OH (Fig. 15). Addition of a surface hydroxyl to an alkylidene species will produce the same R–CHx–OH surface intermediates, which can undergo hydrogenative or dehydrogenative desorption to generate alcohols and aldehydes respectively.210 A large body of experimental evidence confirms the production of these oxygenates via primary pathways.211 The hydroformylation of olefins (oxo process) provides an additional, secondary pathway to aldehydes. Indeed, it was Otto Roelen’s experiments on the recycle of olefins, to maximize liquid product formation in FTS, that led to the serendipitous discovery of the oxo process.212 This route cannot, however, account for the formation of C1 and C2 oxygenates, providing further evidence in support of alcohol production proceeding through primary pathways. Like alkenes, the initial selectivity to oxygenates is compromised by readsorption and further reactions, including hydrogenation to the corresponding paraffin and incorporation into growing chains.213–216 Alcohol yields can therefore be maximized by suppressing re-adsorption and in principle the same strategies used to inhibit secondary alkene reactions should also apply. The selectivity toward oxygen-containing compounds can be increased by operating at high space velocity, which reduces the residence time of primary products, and by increasing the partial pressure of carbon monoxide to promote oxo-synthesis over paraffin formation. The beneficial effect of a low H2/CO ratio is exploited in Shell’s hydroformylation process catalyzed by cobalt.217 This homogeneously catalyzed reaction is operated in two stages; a first reactor operates at low partial pressures of hydrogen to limit the hydrogenation of olefin feed and promote reductive carbonylation while a second reactor is operated at high partial pressures of hydrogen to force the hydrogenation of aldehyde to alcohol. The alkali content of a catalyst also has a profound impact on alcohol selectivity, in line with earlier discussions relating to olefin production. An important function of potassium promotion in iron-based catalysts is to force back secondary reactions involving hydrogen. Thus, whilst comparatively small amounts of oxygenates are formed with non-alkalized iron catalysts, most likely a consequence of extensive secondary reactions, oxygenate production is increased significantly with the addition of potassium. Alkali promotion can also shift the selectivity behavior of cobalt catalyzed FTS. An appreciable oxygenate selectivity (> 20%) can be obtained with highly dispersed cobalt, but this is augmented further with the addition of basic promoters (up to 60%).218,219 To follow the rationale in HAS catalyst design it is worth considering first the mechanistic challenge with respect to formation of the simplest alcohol, methanol, and the requirements of chain growth in FTS. The FT mechanism has already been described in detail and so our attention turns first to methanol synthesis (MS). The development of MS catalysts can be traced back to research conducted at BASF in the 1920s. Early catalysts were based on Cr-MnO with the more well-known, commercial catalysts used today

Fig. 15 Formation of oxygenates via primary pathways. Insertion of molecular CO or a hydroxyl species into a growing hydrocarbon chain results in the formation of an acyl species that can undergo hydrogenative or dehydrogenative desorption to generate alcohols or aldehydes, respectively.

Promoted Fischer-Tropsch catalysts

375

(Cu-ZnO-Al2O3) being introduced some 40 years later.220 In conventional MS, which operates with syngas feed, CO is adsorbed molecularly and is hydrogenated to the alcohol. The C1 source for this reaction is widely believed to be CO2, which exhibits an accelerated rate of hydrogenation compared to CO, and is generated in situ from the WGS reaction.221–223 For HAS, pathways involving CO hydrogenation are believed to be dominant, but a role for WGS and CO2 hydrogenation cannot be ruled out. The production of higher alcohols from syngas therefore requires both chain growth and a CO insertion step where the CO moiety remains intact i.e. the breaking and retention of CO bonds must therefore occur concurrently. If oxygenates are formed by CO insertion in a growing hydrocarbon chain followed by subsequent hydrogenation, this immediately leads to chain termination. Thus, the process of oxygenate formation directly competes with chain-growth, irrespective of the kinetic regime under which the reaction operates. A prime example of a catalyst that exhibits this behavior is Rh, which produces ethanol under FT conditions.224–227 If oxygenates are instead formed by the insertion of OH, the production of longer-chain oxygenates will benefit from a high OH coverage that is present when the catalyst operates under the oxygen removal limiting regime and is exposed to an excess of water. This has been demonstrated for aqueous-phase FTS using both Ru and Rh catalysts.228,229 The selectivity differences observed for CO hydrogenation on different metals can be rationalized on the basis of the potential energies for CO dissociation and of the metal–carbon bond.230 Methanol formation is catalyzed by Cu, which does not dissociate CO, whereas methane formation will dominate on metals that display low barriers for CO dissociation, such as Ni. For chain growth to occur, stronger metal–carbon bonds are needed and, not surprisingly, this property is displayed by the FT metals, Co and Fe. Alkali promoters on iron catalysts increase the electron density and will increase the CO adsorption strength.231 From these basic mechanistic insights a key design feature for HAS emerges, active sites that can promote CO dissociation and chain growth must coexist with sites responsible for CO insertion. For simple monometallic catalysts this can be achieved by maintaining a balance between active metal, responsible for chain growth, and an oxidic component that provides the CO insertion function. The origin of oxygenate selectivity displayed by highly dispersed cobalt/alkali catalysts now becomes clear, the presence of small crystallites is essential to the creation of an irreducible component responsible for CO insertion.218,219 The basic alkali promoter may perform an additional function by inhibiting catalyst reduction and ensuring that a fraction of the supported Co remains unreduced. There is now a general consensus that cobalt carbide (Co2C) is also highly active for CO insertion reactions.232 Cobalt carbide has been shown to catalyze the reductive carbonylation of alkenes,233 but is considered inert in FTS.203 The generation of higher alcohols therefore requires the coexistence of Co2C and Co metal. There are indications that the interface between these two phases is the active site for the FTS reaction. Molybdenum, iron and rhodium have also been explored as potential catalysts for HAS.223,234 Rhodium lies in the periodic table between elements that adsorb CO molecularly (Pd, Pt and Cu) and those capable of dissociative CO adsorption (Fe and Co). Coupled with its extensive use in hydroformylation chemistry, rhodium might therefore be expected to display ideal characteristics for HAS. However, rhodium produces mainly C2 oxygenates from syngas and must be combined with a second metal that provides the chain growth function.223,235–238 The high cost of Rh must also be taken into consideration. In the first quarter of 2021, the price of rhodium was  16 times greater than the price of gold239 making it prohibitively expensive for commercial use. More recent developments have shifted toward multi-metallic catalyst systems that combine, for example, components of FT and MS catalysts; Co–Cu and Co–Mo are among the most promising catalyst candidates reported to date.234 The introduction of a second metal provides much greater flexibility and control in catalyst design that has led to an extensive and varied array of catalyst formulations being investigated for HAS. Despite their compositional diversity, catalytic systems developed for HAS share common characteristics. For effective catalysis there must be a close interaction between a reduced phase, responsible for chain growth, and a second phase (e.g. oxides, suboxides, carbides, or another metal) that assists in the termination of chain growth via insertion of an O-containing species. To achieve intimate contact between metal and an oxidic component, the strategies highlighted in Section 6.16.1.2.3 for the control of dispersion, metal support interaction and extent of reduction can be adopted.240 In the design of a hydrocarbon producing FT catalyst, maximizing the reduced metal surface area is the primary goal, while the target structure for HAS catalysts contains the additional phase that promotes oxygenate formation. It is difficult to completely “switch off” interactions between a metal and the support or additional structural promoters, without compromising stability, and it is for this simple reason that catalysts designed to yield saturated product slates typically also produce some oxygenates. Alkali promotion has also become an important strategy for maximizing alcohol yields.223,234 The alkali promoter increases the dissociation rate of CO, regulating the availability of surface intermediates in a way that allows for both an increased chain growth probability, a, and an increase in the probability for chain termination by CO. Research into HAS can be traced back to the late 1960s and has been the focus of renewed interest in recent years. Yet despite the large volume of work, selectivity remains too low at commercially viable reaction rates. Furthermore, the majority of catalysts also produce significant amounts of CO2, which is often overlooked when reporting selectivity data. Further reading can be found in a number of review articles.223,234,241–243

6.16.6

Outlook

The commercialization of Fischer-Tropsch synthesis has been motivated largely by the production of transportation fuels from coal and natural gas. But with many countries planning to ban cars that are powered by internal combustion engines there will be a rapidly declining demand for petrol and diesel over the next two decades. The net zero transition will not lead to the demise of FT technology, however. While a mix of technology options (battery and fuel cell powered vehicles, hydrogen, electro-fuels

376

Promoted Fischer-Tropsch catalysts

and advanced biofuels) will define the future of road transportation, the power and energy capacity demands of heavy-duty transportation modes will continue to rely on combustible hydrocarbons as a source of fuel. For the aviation sector in particular, energy density requirements and stringent specifications will ensure that Jet-A1 fuel remains the industry norm for years to come. This does not mean there is complacency in an industry that is currently responsible for 2–3% of global anthropogenic greenhouse gas emissions.244 The Air Transport Action Group (ATAG) has set the ambitious aim of neutral carbon growth from 2020, a target that will reduce net carbon emissions to 50% of 2005 levels by 2050.244 Aircraft efficiency and optimized flight operations will form part of the strategy to achieving neutral carbon growth, but the largest contributor to emission reduction will be the large-scale adoption of renewable jet fuel (RJF), a drop-in alternative to fossil derived jet fuel. Synthetic Paraffinic Kerosene (SPK) is currently certified for use in the United States and international aviation fleets added up to 50% in a blend with conventional jet fuel. There are several potential routes to SPK; alcohol-to-jet, hydroprocessed esters and fatty acids, direct sugars to hydrocarbons and anything-to-liquids (XTL) processes.244 Of all possible routes, XTL offers the greatest flexibility on the simple basis that X can be virtually any source of carbonaceous material. The production of kerosene via FTS is already commercially practiced, but future opportunities will exist for the manufacture of fully synthetic jet fuel from renewable resources. Fischer-Tropsch synthesis can also play a leading role in the sustainable production of chemicals. One of the major challenges facing the chemical industry is its high dependence on a steady supply of carbon. And in a post-fossil fuel world the chemicals sector must harness sources of above ground carbon as chemical feedstocks. To reach net zero targets within legislated timescales, the switch to these alternative and compositionally very different feed streams will be dependent on technology platforms that are already demonstrated at an appropriate scale. An established synthesis gas economy, in which CO and H2 are used as the basic chemical building blocks for fuels and chemicals production, will provide a mechanism for achieving these goals. FischerTropsch synthesis is impartial to the origins of synthesis gas and this allows for a smooth transition from the established use of fossil fuels to a future that is entirely dependent on a broad range of sustainable biomass and waste streams using the same FT technology. Chain growth in FTS is controlled by surface polymerization kinetics that constrain our ability to manipulate carbon number distribution. Thus, the technical barrier that must be overcome in the targeted synthesis of chemicals or specific fuel types is an efficient disruption of the normal FT product distribution. This can be achieved to some extent through catalyst and process design, but tuning the structure and electronic properties of known FT catalysts is likely to deliver incremental improvements in performance rather than the necessary step-change. Significant progress has been made in reaction coupling strategies, where a conventional FT catalyst is integrated with a second catalyst component that can cleave C–C bonds, for example.85 This approach may enable the selective formation of specific ranges of hydrocarbons such kerosene for aviation fuel or light olefins for polymer production. Despite extensive studies spanning almost a century, there are still gaps in our understanding of the FT reaction. Attempts to unravel relationships between structure and catalytic response are often hindered by the complex and diverse array of surface structures displayed by FT catalysts. An ability to control interfaces between the various components of a FT catalyst is a potentially powerful design feature but progress is again hampered by a lack of structural information. Often ill-defined, and contributing only a fraction to the overall structure, these interfaces are incredibly difficult to characterize. An accurate, relevant relationship between structure and function can only be achieved through physical measurement of site density, and this is one of the great challenges in heterogeneous catalysis. To compound the complexities of structure determination, FT catalysts are known to undergo a process of self-organization under reaction conditions.245,246 They are dynamic in nature with respect to structure, composition and local electronic environment and it is the reconstructed surface that is responsible for active catalysis. In situ or in operando measurements are needed, but the majority of techniques afford information relating to the bulk properties of a catalyst. Thus, the scientific frontier that must be crossed is the ability to accurately define surface structure under operating conditions. The development of novel, sophisticated methods that combine model material design and synthesis with accurate structure determination under catalytically relevant conditions is essential to bridging these knowledge gaps.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Fischer, F.; Tropsch, H. Chem. Ber. 1926, 59 (4), 832–836. Rofer-DePoorter, C. K. Chem. Rev. 1981, 81 (5), 447–474. Van Der Laan, G. P.; Beenackers, A. A. C. M. Catal. Rev. Sci. Eng. 1999, 41 (3 & 4), 255–318. De Smit, E.; Weckhuysen, B. M. Chem. Soc. Rev. 2008, 37 (12), 2758–2781. Gholami, Z.; Tisler, Z.; Rubás, V. Catal. Rev. Sci. Eng. , 1–84. Stranges, A. N. Studies in Surface Science and Catalysis; vol. 163; Elsevier, 2007; pp 1–27. Anderson, R. B. In Studies in Surface Science and Catalysis. Catalysis on the Energy Scene, vol. 19, Kaliaguine, S., Mahay, A., Eds.; vol. 19; Elsevier: Amsterdam, 1984; pp 456–461. Roelen, O. Erdo¨l Kohle Erdgas Petrochem. 1978, 31, 524–529. Dry, M. E. Catal. Today 2002, 71 (3–4), 227–241. Fischer, F.; Tropsch, H. Process for the Catalytic Production of Multi-Membered Paraffinic Hydrocarbons From Carbon Oxides and Hydrogen. DE 484337 1925. Friedel, R. A.; Anderson, R. B. J. Am. Chem. Soc. 1950, 72 (5), 2307. Herington, E. F. G. Chem. Ind. 1946, (38), 346–347. Flory, P. J.; Flory, B. P. J.; Flory, P. J.; Flory, B. P. J. J. Am. Chem. Soc. 1936, 58 (10), 1877–1885. Schulz, G. V. Zeitschrift Fur Phys. Chemie-Abteilung B-Chemie Der Elem. Aufbau Der Mater. 1935, 30 (5/6), 379–398. Filot, I. A. W.; Van Santen, R. A.; Hensen, E. J. M. Catal. Sci. Technol. 2014, 4 (9), 3129–3140.

Promoted Fischer-Tropsch catalysts 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74.

75. 76. 77. 78. 79. 80.

377

Pichler, H.; Schulz, H.; Elstner, M. Brennstoff-Chemie 1967, 48 (3), 78. Dictor, R. A.; Bell, A. T. J. Catal. 1986, 97 (1), 121–136. Iglesia, E.; Reyes, S. C.; Madon, R. J. J. Catal. 1991, 129 (1), 238–256. Iglesia, E.; Reyes, S. C.; Madon, R. J.; Soled, S. L. Adv. Catal. 1993, 39, 221–302. Schulz, H.; Claeys, M. Appl. Catal. A Gen. 1999, 186 (1–2), 91–107. Steynberg, A. P. In Studies in Surface Science and Catalysis. Fischer-Tropsch Technology; Steynberg, A., Dry, M., Eds.; vol. 152; Elsevier: Amsterdam, 2004; pp 1–63. Chapter 1. Eilers, J.; Posthuma, S. A.; Sie, S. T. Catal. Lett. 1991, 7 (1–4), 253–269. Sie, S. T.; Senden, M. M. G.; Van Wechem, H. M. H. Catal. Today 1991, 8 (3), 371–394. van Steen, E.; Claeys, M. Chem. Eng. Technol. 2008, 31 (5), 655–666. Davis, B. H. Catal. Today 2003, 84 (1–2), 83–98. Dry, M. E. In Catalysis Science and Technology; Anderson, J. R., Boudart, M., Eds., Springer: Berlin, 1981; pp 159–256. Dry, M. E. In Studies in Surface Science and Catalysis. Fischer-Tropsch Technology; Steynberg, A., Dry, M., Eds.; vol. 152; Elsevier: Amsterdam, 2004; pp 1–63. Chapter 7. Jager, B.; Espinoza, R. Catal. Today 1995, 23 (1), 17–28. Raupp, G. B.; Delgass, W. N. J. Catal. 1979, 58 (3), 361–369. Niemantsverdriet, J. W.; Vanderkraan, A. M.; Vandijk, W. L.; Vanderbaan, H. S. J. Phys. Chem. 1980, 84 (25), 3363–3370. Butt, J. B. Catal. Lett. 1991, 7 (1–4), 61–81. Shroff, M. D.; Kalakkad, D. S.; Coulter, K. E.; Kohler, S. D.; Harrington, M. S.; Jackson, N. B.; Sault, A. G.; Datye, A. K. J. Catal. 1995, 156 (2), 185–207. Li, S.; Meitzner, G. D.; Iglesia, E. J. Phys. Chem. B 2001, 105 (24), 5743–5750. Paalanen, P. P.; Weckhuysen, B. M. ChemCatChem 2020, 12 (17), 4202–4223. Van Dijk, W. L.; Niemantsverdriet, J. W.; Van Dar Kraan, A. M.; Van Der Baan, H. S. Appl. Catal. 1982, 2 (4–5), 273–288. Mansker, L. D.; Jin, Y. M.; Bukur, D. B.; Datye, A. K. Appl. Catal. A Gen. 1999, 186 (1–2), 277–296. De Smit, E.; Cinquini, F.; Beale, A. M.; Safonova, O. V.; Van Beek, W.; Sautet, P.; Weckhuysen, B. M. J. Am. Chem. Soc. 2010, 132 (42), 14928–14941. Newsome, D. S. Catal. Rev. Eng. 1980, 21 (2), 275–318. Zhang, H. B.; Schrader, G. L. J. Catal. 1985, 95 (1), 325–332. Rethwisch, D. G.; Dumesic, J. A. J. Catal. 1986, 101 (1), 35–42. Lox, E. S.; Froment, G. F. Ind. Eng. Chem. Res. 1993, 32 (1), 61–70. Li, S.; Krishnamoorthy, S.; Li, A.; Meitzner, G. D.; Iglesia, E. J. Catal. 2002, 206 (2), 202–217. Krishnamoorthy, S.; Li, A.; Iglesia, E. Catal. Lett. 2002, 80 (1/2), 77–86. Khodakov, A. Y.; Chu, W.; Fongarland, P. Chem. Rev. (Washington, DC) 2007, 107 (5), 1692–1744. Keyser, M. J.; Everson, R. C.; Espinoza, R. L. Appl. Catal. A Gen. 1998, 171 (1), 99–107. Khassin, A. A.; Anufrienko, V. F.; Ikorskii, V. N.; Plyasova, L. M.; Kustova, G. N.; Larina, T. V.; Molina, I. Y.; Parmon, V. N. Phys. Chem. Chem. Phys. 2002, 4 (17), 4243. Yu, F.; Lin, T.; Wang, X.; Li, S.; Lu, Y.; Wang, H.; Zhong, L.; Sun, Y. Appl. Catal. A Gen. 2018, 563, 146–153. Borg, O.; Eri, S.; Blekkan, E.; Storsater, S.; Wigum, H.; Rytter, E.; Holmen, A. J. Catal. 2007, 248 (1), 89–100. Khodakov, A. Y.; Griboval-Constant, A.; Bechara, R.; Zholobenko, V. L. J. Catal. 2002, 206 (2), 230–241. Khodakov, A. Y.; Griboval-Constant, A.; Bechara, R.; Villain, F. J. Phys. Chem. B 2001, 105 (40), 9805–9811. Saib, A. M.; Claeys, M.; van Steen, E. Catal. Today 2002, 71 (3–4), 395–402. Parks, G. A. Chem. Rev. 1965, 65 (2), 177–198. Morales, F.; Weckhuysen, B. M. Catalysis; vol. 19; The Royal Society of Chemistry: Cambridge, 2006; pp 1–40. Girardon, J.; Quinet, E.; Gribovalconstant, A.; Chernavskii, P.; Gengembre, L.; Khodakov, A. J. Catal. 2007, 248 (2), 143–157. Vitos, L.; Ruban, A. V.; Skriver, H. L.; Kollar, J. Surf. Sci. 1998, 411 (1–2), 186–202. Tsakoumis, N. E.; Ronning, M.; Borg, O.; Rytter, E.; Holmen, A. Catal. Today 2010, 154 (3–4), 162–182. Baker, J. E.; Burch, R.; Niu, Y. Q. Appl. Catal. 1991, 73 (1), 135–152. Hilmen, A. M.; Schanke, D.; Holmen, A. Catal. Lett. 1996, 38 (3–4), 143–147. Ma, W.; Jacobs, G.; Keogh, R. A.; Bukur, D. B.; Davis, B. H. Appl. Catal. A Gen. 2012, 437–438, 1–9. Cook, K. M.; Perez, H. D.; Bartholomew, C. H.; Hecker, W. C. Appl. Catal. A Gen. 2014, 482, 275–286. Conner, W. C.; Falconer, J. L. Chem. Rev. 1995, 95 (3), 759–788. Shimura, K.; Miyazawa, T.; Hanaoka, T.; Hirata, S. Appl. Catal. A Gen. 2015, 494, 1–11. Karim, W.; Spreafico, C.; Kleibert, A.; Gobrecht, J.; VandeVondele, J.; Ekinci, Y.; van Bokhoven, J. A. Nature 2017, 541 (7635), 68–71. Lee, S. Y.; Aris, R. Catal. Rev. Eng. 1985, 27 (2), 207–340. Krishna, R.; Sie, S. T. Fuel Process. Technol. 2000, 64 (1–3), 73–105. Gamlin, T. D. (Johnson Matthey Davy Technologies Ltd). Fischer–Tropsch Process in a Radial Reactor. US Pat. Appl. 8906970 2012. Pearson, R.; Coe, A.; Paterson, J. Johnson Matthey Technol. Rev. 2021, 65 (3), 395–403. Deshmukh, S. R.; Tonkovich, A. L. Y.; McDaniel, J. S.; Schrader, L. D.; Burton, C. D.; Jarosch, K. T.; Simpson, A. M.; Kilanowski, D. R.; Leviness, S. Biofuels 2011, 2 (3), 315–324. Davis, B. H. Top. Catal. 2005, 32 (3–4), 143–168. Jacobs, G.; Davis, B. H. In Multiphase Catalytic Reactors: Theory, Design, Manufacturing, and Applications; Önsan, Z. I., Avci, A. K., Eds., John Wiley & Sons, Inc.: Hoboken, New Jersey, 2016; pp 269–294. Chapter 12. Weng, L.; Men, Z. In Multiphase Reactor Engineering for Clean and Low-Carbon Energy Applications; Cheng, Y., Wei, F., Jin, Y., Eds., John Wiley & Sons, Inc.: Hoboken, New Jersey, 2017; pp 219–270. Chapter 7. Duvenhage, D. J.; Shingles, T. Catal. Today 2002, 71 (3–4), 301–305. Peacock, M.; Paterson, J.; Reed, L.; Davies, S.; Carter, S.; Coe, A.; Clarkson, J. Top. Catal. 2020, 63 (3–4), 328–339. McCollum, D. L.; Zhou, W.; Bertram, C.; de Boer, H.-S.; Bosetti, V.; Busch, S.; Després, J.; Drouet, L.; Emmerling, J.; Fay, M.; Fricko, O.; Fujimori, S.; Gidden, M.; Harmsen, M.; Huppmann, D.; Iyer, G.; Krey, V.; Kriegler, E.; Nicolas, C.; Pachauri, S.; Parkinson, S.; Poblete-Cazenave, M.; Rafaj, P.; Rao, N.; Rozenberg, J.; Schmitz, A.; Schoepp, W.; van Vuuren, D.; Riahi, K. Nat. Energy 2018, 3 (7), 589–599. Clarke, L.; et al. In Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change; Edenhofer, O. R.; et al., Eds., Cambridge University Press: Cambridge, 2014; pp 413–510. Chapter 6. Rogelj, J.; et al. In An IPCC Special Report on the Impacts of Global Warming of 1.5 C Above Pre-Industrial Levels; Masson-Delmotte, V.; et al., Eds.; 2018; pp 93–174. Chapter 2. In Press. Schmidt, P.; Batteiger, V.; Roth, A.; Weindorf, W.; Raksha, T. Chemie Ing. Tech. 2018, 90 (1–2), 127–140. Panzone, C.; Philippe, R.; Chappaz, A.; Fongarland, P.; Bengaouer, A. J. CO2 Util. 2020, 38, 314–347. Hansen, J. B. Faraday Discuss. 2015, 182, 9–48. Torrell, M.; García-Rodríguez, S.; Morata, A.; Penelas, G.; Tarancón, A. Faraday Discuss. 2015, 182, 241–255.

378 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151.

Promoted Fischer-Tropsch catalysts Chen, X.; Guan, C.; Xiao, G.; Du, X.; Wang, J.-Q. Faraday Discuss. 2015, 182, 341–351. Shin, T. H.; Myung, J.-H.; Verbraeken, M.; Kim, G.; Irvine, J. T. S. Faraday Discuss. 2015, 182, 227–239. Zheng, Y.; Wang, J. C.; Yu, B.; Zhang, W. Q.; Chen, J.; Qiao, J. L.; Zhang, J. J. Chem. Soc. Rev. 2017, 46 (5), 1427–1463. Hauch, A.; Kungas, R.; Blennow, P.; Hansen, A. B.; Hansen, J. B.; Mathiesen, B. V.; Mogensen, M. B. Science 2020, 370 (6513), 186. Zhou, W.; Cheng, K.; Kang, J.; Zhou, C.; Subramanian, V.; Zhang, Q.; Wang, Y. Chem. Soc. Rev. 2019, 48 (12), 3193–3228. Ye, R.-P.; Ding, J.; Gong, W.; Argyle, M. D.; Zhong, Q.; Wang, Y.; Russell, C. K.; Xu, Z.; Russell, A. G.; Li, Q.; Fan, M.; Yao, Y.-G. Nat. Commun. 2019, 10 (1), 5698. Marlin, D. S.; Sarron, E.; Sigurbjornsson, O. Front. Chem. 2018, 6, 446. Carbon Recycling International. https://www.carbonrecycling.is/projects. Gibson, P. Faraday Discuss. 2017, 197, 547–555. Adesina, A. A. Appl. Catal. A Gen. 1996, 138 (2), 345–367. Vannice, M. A. J. Catal. 1975, 134, 129–134. Moodley, D. J.; van de Loosdrecht, J.; Saib, A. M.; Overett, M. J.; Datye, A. K.; Niemantsverdriet, J. W. Appl. Catal. A Gen. 2009, 354 (1), 102–110. Pichler, H.; Schulz, H. Chem. Ing. Tech. 1970, 42 (18), 1162–1174. Bukur, D. B.; Mukesh, D.; Patel, S. A. Ind. Eng. Chem. Res. 1990, 29 (2), 194–204. Madon, R. J.; Reyes, S. C.; Iglesia, E. J. Phys. Chem. 1991, 95 (20), 7795–7804. Biloen, P.; Helle, J. N.; Sachtler, W. M. H. J. Catal. 1979, 58 (1), 95–107. Van Santen, R. A.; Markvoort, A. J.; Ghouri, M. M.; Hilbers, P. A. J.; Hensen, E. J. M. J. Phys. Chem. C 2013, 117 (9), 4488–4504. van Santen, R. A.; Markvoort, A. J. ChemCatChem 2013, 5 (11), 3384–3397. Filot, I. A. W. A.; van Santen, R. A. A.; Hensen, E. J. M. J. Angew Chem. Int. Ed. Engl. 2014, 53 (47), 12746–12750. Pestman, R.; Chen, W.; Hensen, E. ACS Catal. 2019, 9 (5), 4189–4195. Chen, W.; Filot, I. A. W.; Pestman, R.; Hensen, E. J. M. ACS Catal. 2017, 7 (12), 8061–8071. Craxford, S. R.; Rideal, E. K. J. Chem. Soc. 1939, 1604–1614. Casavola, M.; Xie, J.; Meeldijk, J. D.; Krans, N. A.; Goryachev, A.; Hofmann, J. P.; Dugulan, A. I.; De Jong, K. P. ACS Catal. 2017, 7 (8), 5121–5128. Taylor, H. S. Proc. R. Soc. A 1925, 108 (745), 105–111. Zambelli, T.; Wintterlin, J.; Trost, J.; Ertl, G. Science 1996, 273 (5282), 1688–1690. Bezemer, G. L.; Bitter, J. H.; Kuipers, H. P. C. E.; Oosterbeek, H.; Holewijn, J. E.; Xu, X.; Kapteijn, F.; Van Diilen, A. J.; De Jong, K. P. J. Am. Chem. Soc. 2006, 128 (12), 3956–3964. Breejen, J.; Sietsma, J.; Friedrich, H.; Bitter, J. H.; De Jong, K. J. Catal. 2010, 270, 146–152. Eschemann, T. O.; Lamme, W. S.; Manchester, R. L.; Parmentier, T. E.; Cognigni, A.; Rønning, M.; De Jong, K. P. J. Catal. 2015, 328, 130–138. Prieto, G.; Martínez, A.; Concepción, P.; Moreno-Tost, R. J. Catal. 2009, 266 (1), 129–144. Fischer, N.; Van Steen, E.; Claeys, M. J. Catal. 2013, 299, 67–80. Ho, S. W.; Cruz, J. M.; Houalla, M.; Hercules, D. M. J. Catal. 1992, 135 (1), 173–185. Herranz, T.; Deng, X.; Cabot, A.; Guo, J.; Salmeron, M. J. Phys. Chem. B 2009, 113 (31), 10721–10727. Xiong, H.; Motchelaho, M. A. M. M.; Moyo, M.; Jewell, L. L.; Coville, N. J. J. Catal. 2011, 278 (1), 26–40. Cheng, Q.; Tian, Y.; Lyu, S.; Zhao, N.; Ma, K.; Ding, T.; Jiang, Z.; Wang, L.; Zhang, J.; Zheng, L.; Gao, F.; Dong, L.; Tsubaki, N.; Li, X. Nat. Commun. 2018, 9 (1). Martínez, A.; Prieto, G. J. Catal. 2007, 245 (2), 470–476. Yang, J.; Tveten, E. Z.; Chen, D.; Holmen, A. Langmuir 2010, 26 (21), 16558–16567. Den Breejen, J. P.; Radstake, P. B.; Bezemer, G. L.; Bitter, J. H.; Frøseth, V.; Holmen, A.; De Jong, K. P. J. Am. Chem. Soc. 2009, 131 (20), 7197–7203. Bridge, M. E.; Comrie, C. M.; Lambert, R. M. Surf. Sci. 1977, 67 (2), 393–404. Papp, H. Surf. Sci. 1983, 129 (1), 205–218. Papp, H. Surf. Sci. 1985, 149 (2–3), 460–470. Geerlings, J. J. C. C.; Zonnevylle, M. C.; de Groot, C. P. M. M. Surf. Sci. 1991, 241 (3), 315–324. Hardeveld, R. V.; Montfoort, A. V. Surf. Sci. 1966, 4, 396–430. Eischens, R. P.; Jacknow, J. Proc. 3rd Intern. Congress on Catalysis, 1965; p 627. Dahl, S.; Logadottir, A.; Egeberg, R. C.; Larsen, J. H.; Chorkendorff, I.; Törnqvist, E.; Nørskov, J. K. Phys. Rev. Lett. 1999, 83 (9), 1814–1817. Shetty, S.; Van Santen, R. A. Catal. Today 2011, 171 (1), 168–173. Zijlstra, B.; Broos, R. J. P.; Chen, W.; Oosterbeek, H.; Filot, I. A. W.; Hensen, E. J. M. ACS Catal. 2019, 9 (8), 7365–7372. Shetty, S.; Jansen, A. P. J.; Van Santen, R. A. J. Phys. Chem. C 2008, 112 (36), 14027–14033. Petersen, M. A.; van den Berg, J.-A. A.; Ciobîcǎ, I. M.; van Helden, P.; Ciobîca, I. M.; van Helden, P. ACS Catal. 2017, 7 (3), 1984–1992. Strebel, C.; Murphy, S.; Nielsen, R. M.; Nielsen, J. H.; Chorkendorff, I. Phys. Chem. Chem. Phys. 2012, 14 (22), 8005–8012. van Helden, P.; Ciobîca, I. M.; Coetzer, R. L. J. J.; Ciobîcə, I. M.; Coetzer, R. L. J. J.; Ciobîca, I. M.; Coetzer, R. L. J. J. Catal. Today 2016, 261, 48–59. Agrawal, R.; Phatak, P.; Spanu, L. Catal. Today 2018, 312 (December 2017), 174–180. van Etten, M. P. C.; Zijlstra, B.; Hensen, E. J. M.; Filot, I. A. W. ACS Catal. 2021, 8484–8492. Wilson, J.; de Groot, C. J. Phys. Chem. 1995, 99, 7860–7866. Klink, C.; Stensgaard, I.; Besenbacher, F.; Lægsgaard, E. Surf. Sci. 1995, 342 (1–3), 250–260. Klink, C.; Stensgaard, I.; Besenbacher, F.; Lægsgaard, E. Surf. Sci. 1996, 360 (1–3), 171–179. Ciobîcǎ, I. M.; van Santen, R. A.; van Berge, P. J.; van de Loosdrecht, J. Surf. Sci. 2008, 602 (1), 17–27. Kirsch, J. E.; Harris, S. Surf. Sci. 2003, 522 (1–3), 125–142. Kirsch, J. E.; Harris, S. Surf. Sci. 2004, 553 (1–3), 82–94. Tuxen, A.; Carenco, S.; Chintapalli, M.; Chuang, C. H.; Escudero, C.; Pach, E.; Jiang, P.; Borondics, F.; Beberwyck, B.; Alivisatos, A. P.; Thornton, G.; Pong, W. F.; Guo, J.; Perez, R.; Besenbacher, F.; Salmeron, M. J. Am. Chem. Soc. 2013, 135 (6), 2273–2278. Inderwildi, O. R.; Jenkins, S. J.; King, D. A. J. Phys. Chem. C 2008, 112 (5), 1305–1307. Hibbitts, D.; Iglesia, E. Acc. Chem. Res. 2015, 48 (5), 1254–1262. Zijlstra, B.; Broos, R. J. P.; Chen, W.; Bezemer, G. L.; Filot, I. A. W.; Hensen, E. J. M. ACS Catal. 2020, 10 (16), 9376–9400. Maitlis, P. M.; Quyoum, R.; Long, H. C.; Turner, M. L. Appl. Catal. A Gen. 1999, 186 (1–2), 363–374. Overett, M. J.; Hill, R. O.; Moss, J. R. Coord. Chem. Rev. 2000, 206–207, 581–605. Bub, G.; Baerns, M.; Büssemeier, B.; Frohning, C. Chem. Eng. Sci. 1980, 35 (1–2), 348–355. Golodets, G. I. Theor. Exp. Chem. 1985, 21 (5), 525–529. Yates, I. C.; Satterfield, C. N. Energy Fuels 1991, 5 (1), 168–173. Niemantsverdriet, J. W.; Chorkendorff, I. Concepts of Modern Catalysis and Kinetics, John Wiley & Sons, Ltd, 2003; pp 23–78. Filot, I. A. W. Introduction to Microkinetic Modeling, Eindhoven Univeristy Press: Eindhoven, 2018. Eyring, H. J. Chem. Phys. 1935, 37, 107–115. Byrne, G. D.; Hindmarsh, A. C. ACM Trans. Math. Softw. 1975, 1 (1), 71–96.

Promoted Fischer-Tropsch catalysts 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218.

379

Campbell, C. T. ACS Catal. 2017, 7 (4), 2770–2779. Santen, R. A. V.; Neurock, M.; Shetty, S. G. Chem. Rev. 2010, 110 (4), 2005–2048. Abild-Pedersen, F.; Greeley, J.; Studt, F.; Rossmeisl, J.; Munter, T. R.; Moses, P. G.; Skúlason, E.; Bligaard, T.; Nørskov, J. K. Phys. Rev. Lett. 2007, 99 (1), 4–7. Qi, Y.; Yang, J.; Chen, D.; Holmen, A. Catal. Lett. 2015, 145 (1), 145–161. Su, H. Y.; Zhao, Y.; Liu, J. X.; Sun, K.; Li, W. X. Catal. Sci. Technol. 2017, 7 (14), 2967–2977. Chen, P. P.; Liu, J. X.; Li, W. X. ACS Catal. 2019, 8093–8103. Arakawa, H.; Bell, A. T. Ind. Eng. Chem. Process Des. Dev. 1983, 22 (1), 97–103. Anderson, R. B.; Seligman, B.; Shultz, J. F.; Kelly, R.; Elliott, M. A. Ind. Eng. Chem. Process Des. Dev. 1952, 44, 391–397. van Santen, R. A.; Markvoort, A. J.; Filot, I. A. W. W.; Ghouri, M. M.; Hensen, E. J. M. M. Phys. Chem. Chem. Phys. 2013, 15 (40), 17038–17063. Andersen, M.; Plaisance, C. P.; Reuter, K. J. Chem. Phys. 2017, 147 (15). Bruix, A.; Margraf, J. T.; Andersen, M.; Reuter, K. Nat. Catal. 2019, 2 (8), 659–670. Van Heerden, T.; Van Steen, E. Faraday Discuss. 2017, 197, 87–99. Leckel, D. Energy Fuels 2009, 23 (5), 2342–2358. Diercks, R.; Arndt, J. D.; Freyer, S.; Geier, R.; Machhammer, O.; Schwartze, J.; Volland, M. Chem. Eng. Technol. 2008, 31 (5), 631–637. Amghizar, I.; Vandewalle, L. A.; Van Geem, K. M.; Marin, G. B. Engineering (Beijing, China) 2017, 3 (2), 171–178. Gary, J. H.; Handwerk, G. E. Petroleum Refining: Technology and Economics, 4th ed.; CRC Press, 2001. Schulz, H.; Claeys, M. Appl. Catal. A Gen. 1999, 186 (1–2), 71–90. Mross, W. D. Catal. Rev. Eng. 1983, 25 (4), 591–637. Anderson, R. B. In Catalysis, Vol. 4, Hydrocarbon Synthesis, Hydrogenation and Cyclization; Emmett, P. H., Ed., Reinhold: New York, 1956; pp 29–255. Chapter 2. Storch, H.; Golumbic, N.; Anderson, R. B. The Fischer-Tropsch and Related Synthesis, Wiley: New York, 1951. Shafer, W. D.; Gnanamani, M. K.; Graham, U. M.; Yang, J.; Masuku, C. M.; Jacobs, G.; Davis, B. H. Catalysts 2019, 9 (3), 259. Niemantsverdriet, J. W.; Vanderkraan, A. M. J. Catal. 1981, 72 (2), 385–388. Botes, G. F.; Bromfield, T. C.; Coetzer, R. L. J.; Crous, R.; Gibson, P.; Ferreira, A. C. Catal. Today 2016, 275, 40–48. McCue, A. J.; Anderson, J. A. Catal. Sci. Technol. 2014, 4 (2), 272–294. Ma, W.; Jacobs, G.; Sparks, D. E.; Shafer, W. D.; Hamdeh, H. H.; Hopps, S. D.; Pendyala, V. R. R.; Hu, Y.; Xiao, Q.; Davis, B. H. Appl. Catal. A Gen. 2016, 513, 127–137. Xu, J.-D.; Chang, Z.-Y.; Zhu, K.-T.; Weng, X.-F.; Weng, W.-Z.; Zheng, Y.-P.; Huang, C.-J.; Wan, H.-L. Appl. Catal. A Gen. 2016, 514, 103–113. Jiang, F.; Zhang, M.; Liu, B.; Xu, Y.; Liu, X. Catal. Sci. Technol. 2017, 7 (5), 1245–1265. Davidson, A. L.; Gibson, E. K.; Cibin, G.; van Rensburg, H.; Parker, S. F.; Webb, P. B.; Lennon, D. J. Catal. 2020, 392, 197–208. Warringham, R.; Davidson, A. L.; Webb, P. B.; Tooze, R. P.; Parker, S. F.; Lennon, D. Catal. Today 2020, 339, 32–39. Paalanen, P. P.; Vreeswijk, S. H.; Dugulan, A. I.; Weckhuysen, B. M. ChemCatChem 2020, 12 (20), 5121–5139. Paalanen, P. P.; Van Vreeswijk, S. H.; Weckhuysen, B. M. ACS Catal. 2020, 10 (17), 9837–9855. Madon, R. J.; Shaw, H. Catal. Rev. 1977, 15 (1), 69–106. Janardanarao, M. Ind. Eng. Chem. Res. 1990, 29 (9), 1735–1753. Bromfield, T. C.; Coville, N. J. Appl. Catal. A Gen. 1999, 186 (1–2), 297–307. Kritzinger, J. A. Catal. Today 2002, 71 (3–4), 307–318. Torres Galvis, H. M.; Bitter, J. H.; Khare, C. B.; Ruitenbeek, M.; Dugulan, A. I.; De Jong, K. P. Science 2012, 335 (6070), 835–838. Torres Galvis, H. M.; de Jong, K. P. ACS Catal. 2013, 3 (9), 2130–2149. Torres Galvis, H. M.; Koeken, A. C. J.; Bitter, J. H.; Davidian, T.; Ruitenbeek, M.; Dugulan, A. I.; de Jong, K. P. Catal. Today 2013, 303, 95–102. Shultz, J. F.; Hofer, L. J. E.; Karn, F. S.; Anderson, R. B. J. Phys. Chem. 1962, 66 (3), 501–506. Bartholomew, C. H.; Bowman, R. M. Appl. Catal. 1985, 15 (1), 59–67. Johnson, J. Y. (filed on behalf of I.G. Farbenindustrie Aktiengesselschaft). Improvements in the Manufacture and Production of Unsaturated Hydrocarbons of Low Boiling Point. British Patent 322 1929, 284. Karn, F. S.; Shultz, J. F.; Kelly, R. E.; Anderson, R. B. Ind. Eng. Chem. Prod. Res. Dev. 1964, 3 (1), 33–38. Stenger, H. G.; Satterfield, C. N. Ind. Eng. Chem. Process Des. Dev. 1985, 24 (2), 415–420. Anderson, R. B.; Karn, F. S.; Shultz, J. F. J. Catal. 1965, 4 (1), 56–63. Xie, J.; Yang, J.; Dugulan, A. I.; Holmen, A.; Chen, D.; De Jong, K. P.; Louwerse, M. J. ACS Catal. 2016, 6 (5), 3147–3157. Warringham, R.; Davidson, A. L.; Webb, P. B.; Tooze, R. P.; Ewings, R. A.; Parker, S. F.; Lennon, D. RSC Adv. 2019, 9 (5), 2608–2617. Benziger, J.; Madix, R. J. Surf. Sci. 1980, 94 (1), 119–153. Bartholomew, C. H.; Agrawal, P. K.; Katzer, J. R. In Advances in Catalysis; Eley, D. D., Pines, H., Weisz, P. B., Eds.; vol. 31; Academic Press: Cambridge, Massachusetts, 1982; pp 135–242. Barbier, J.; Lamy-Pitara, E.; Marecot, P.; Boitiaux, J. P.; Cosyns, J.; Verna, F. In Advances in Catalysis; Eley, D. D., Pines, H., Weisz, P. B., Eds.; vol. 37; Academic Press: Cambridge, Massachusetts, 1990; pp 279–318. Zhong, L.; Yu, F.; An, Y.; Zhao, Y.; Sun, Y.; Li, Z.; Lin, T.; Lin, Y.; Qi, X.; Dai, Y.; Gu, L.; Hu, J.; Jin, S.; Shen, Q.; Wang, H. Nature 2016, 538 (7623), 84–87. Xie, J.; Paalanen, P. P.; van Deelen, T. W.; Weckhuysen, B. M.; Louwerse, M. J.; de Jong, K. P. Nat. Commun. 2019, 10 (1), 167. Sadeqzadeh, M.; Karaca, H.; Safonova, O. V.; Fongarland, P.; Chambrey, S.; Roussel, P.; Griboval-Constant, A.; Lacroix, M.; Curulla-Ferré, D.; Luck, F.; Khodakov, A. Y. Catal. Today 2011, 164 (1), 62–67. Jiao, F.; Li, J.; Pan, X.; Xiao, J.; Li, H.; Ma, H.; Wei, M.; Pan, Y.; Zhou, Z.; Li, M.; Miao, S.; Li, J.; Zhu, Y.; Xiao, D.; He, T.; Yang, J.; Qi, F.; Fu, Q.; Bao, X. Science 2016, 351 (6277), 1065–1068. Zhu, Y.; Pan, X.; Jiao, F.; Li, J.; Yang, J.; Ding, M.; Han, Y.; Liu, Z.; Bao, X. ACS Catal. 2017, 7 (4), 2800–2804. Sun, J.; Yang, G.; Peng, X.; Kang, J.; Wu, J.; Liu, G.; Tsubaki, N. ChemCatChem 2019, 11 (5), 1412–1424. Fischer, F.; Tropsch, H. DRP 411216, 1922. Schulz, H.; Zein El Deen, A. Fuel Process. Technol. 1977, 1 (1), 31–56. Partington, R.; Clarkson, J.; Paterson, J.; Sullivan, K.; Wilson, J. J. Anal. Sci. Technol. 2020, 11 (1). Johnston, P.; Joyner, R. W.; Tetenyi, P.; Solymosi, F.; Bhasin, M. M.; Ponec, V.; Tamaru, K.; Zanderighi, L.; Duprez, D.; Ruckowski, J.; Wang, D.; Kuipers, E. W. Stud. Surf. Sci. Catal. 1993, 75, 165–180. Schulz, H.; Erich, E.; Gorre, H.; Van Steen, E. Catal. Lett. 1990, 7 (1–4), 157–167. Roelen, O. Angew. Chem. 1951, 63 (20), 482–483. Kummer, J. T.; Podgurski, H. H.; Spencer, W. B.; Emmett, P. H. J. Am. Chem. Soc. 1951, 73 (2), 564–569. Kummer, J. T.; Emmett, P. H. J. Am. Chem. Soc. 1953, 75 (21), 5177–5183. Kokes, R. J.; Hall, W. K.; Emmett, P. H. J. Am. Chem. Soc. 1957, 79 (12), 2989–2996. Hall, W. K.; Kokes, R. J.; Emmett, P. H. J. Am. Chem. Soc. 1960, 82 (5), 1027–1037. Frohning, C. D.; Kohlpaintner, C. W. In Applied Homogeneous Catalysis With Organometallic Compounds; Cornils, B., Hermann, W. A., Eds., VCH: Weinheim, 1996; p 27. Takeuchi, K.; Matsuzaki, T.; Hanaoka, T. A.; Arakawa, H.; Sugi, Y.; Wei, K. J. Mol. Catal. 1989, 55 (1–3), 361–370.

380 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. 229. 230. 231. 232. 233. 234. 235. 236. 237. 238. 239. 240. 241. 242. 243. 244. 245. 246.

Promoted Fischer-Tropsch catalysts Takeuchi, K.; Matsuzaki, T.; Arakawa, H.; Hanaoka, T.; Sugi, Y. Appl. Catal. 1989, 48 (1), 149–157. Davies, P.; Snowdon, F. F. (Imperial Chemical Industries Ltd). Production of Oxygenated Hydrocarbons. In US3326956; 1963. Klier, K. Adv. Catal. 1982, 31, 243–313. Martin, O.; Perez-Ramirez, J. Catal. Sci. Technol. 2013, 3 (12), 3343–3352. Luk, H. T.; Mondelli, C.; Ferre, D. C.; Stewart, J. A.; Perez-Ramirez, J.; Ferré, D. C.; Stewart, J. A.; Pérez-Ramírez, J.; Ferre, D. C.; Stewart, J. A.; Perez-Ramirez, J. Chem. Soc. Rev. 2017, 46 (5), 1358–1426. Filot, I. A. W.; Broos, R. J. P.; Van Rijn, J. P. M.; Van Heugten, G. J. H. A.; Van Santen, R. A.; Hensen, E. J. M. ACS Catal. 2015, 5 (9), 5453–5467. Spivey, J. J.; Egbebi, A. Chem. Soc. Rev. 2007, 36 (9), 1514–1528. Subramani, V.; Gangwal, S. K. Energy Fuels 2008, 22 (2), 814–839. He, J.; Zhang, W. J. Zhejiang Univ. A 2008, 9 (5), 714–719. Quek, X.-Y.; Guan, Y.; van Santen, R. A.; Hensen, E. J. M. ChemCatChem 2011, 3 (11), 1735–1738. Peregudova, A. S.; Barrios, A. J.; Ordomsky, V. V.; Borisova, N. E.; Khodakov, A. Y. Chem. Commun. 2020, 56 (2), 277–280. van Santen, R. A.; Neurock, M.; Vansanten, R. A.; Neurock, M. Catal. Rev. Eng. 1995, 37 (4), 557–698. Dry, M. E.; Shingles, T.; Boshoff, L. J.; Oosthuiz, G. J. Catal. 1969, 15 (2), 190. Xu, X. D.; Scholten, J. J. F.; Mausbeck, D. Appl. Catal. A Gen. 1992, 82 (1), 91–109. Dong, W.; Liu, J. J. J.; Zhu, H.; Ding, Y.; Pei, Y.; Liu, J. J. J.; Du, H.; Jiang, M.; Liu, T.; Su, H.; Li, W. J. Phys. Chem. C 2014, 118 (33), 19114–19122. Ao, M.; Pham, G. H.; Sunarso, J.; Tade, M. O.; Liu, S. ACS Catal. 2018, 8 (8), 7025–7050. Ichikawa, M. Bull. Chem. Soc. Jpn. 1978, 51 (8), 2268–2272. Ichikawa, M. Bull. Chem. Soc. Jpn. 1978, 51 (8), 2273–2277. Ichikawa, M. J. Chem. Soc. Chem. Commun. 1978, (13), 566. Yu, J.; Mao, D. S.; Han, L. P.; Guo, Q. S.; Lu, G. Z. Fuel Process. Technol. 2013, 112, 100–105. Umicore precious metals management. https://pmm.umicore.com/en/prices/rhodium. Ribeiro, M. C.; Gnanamani, M. K.; Azevedo, I. R.; Rabelo-Neto, R. C.; Jacobs, G.; Davis, B. H.; Noronha, F. B. Top. Catal. 2014, 57 (6–9), 550–560. Xiaoding, X.; Doesburg, E. B. M.; Scholten, J. J. F. Catal. Today 1987, 2 (1), 125–170. Herman, R. G. Catal. Today 2000, 55 (3), 233–245. Zaman, S.; Smith, K. J. Catal. Rev. Sci. Eng. 2012, 54 (1), 41–132. Schäfer, A. W. In Biofuels for Aviation; Chuck, C. J., Ed., Academic Press: Cambridge, Massachusetts, 2016; pp 3–16. Chapter 1. Claeys, M.; Cowan, R.; Schulz, H. Top. Catal. 2003, 26 (1), 139–143. Schulz, H. Catal. Today 2014, 228, 113–122.

6.17

Selective oxidation by mixed metal nanoparticles

Hannah Rogers and Simon J. Freakley, Department of Chemistry, University of Bath, Bath, United Kingdom © 2023 Elsevier Ltd. All rights reserved.

6.17.1 6.17.2 6.17.3 6.17.4 6.17.4.1 6.17.4.2 6.17.5 6.17.6 6.17.7 6.17.7.1 6.17.7.2 6.17.7.3 6.17.8 6.17.9 6.17.10 References

Introduction Importance of selective catalytic oxidation Multimetallic nanoparticle catalysts Classes of well-defined bimetallic nanoparticle catalysts Single atom alloys (SAAs) Near surface alloys Cluster beam deposition catalyst synthesis Bimetallic nanoalloys Selective oxidation reactions Catalytic oxidation of alkanes Biomass oxidation Glucose oxidation Glycerol Oxidation of bio-derived furanics Conclusions

381 382 383 384 384 386 387 389 390 390 391 392 394 395 398 398

Abstract The growth of modern society mirrors is dependent on the development of more efficient catalytic processes. Catalysis is a key enabling technology in the production of new materials, improving health of populations and securing food production. Oxidation reactions using molecular oxygen are a potential route to many commodities and driving a more sustainable chemical industry, however the large-scale adoption of selective oxidation reactions using precious metal catalysts is still limited. The development of new methods to produce well-defined bimetallic catalysts has accelerated the understanding of structure-activity relationships in many oxidation processes. This has led to significantly increased performance compared to monometallic analogues in terms of activity, selectivity and stability. This chapter concerns the use of well-defined bimetallic nanoparticles catalysts containing precious metals and reviews the classes of materials reported to date and the development of these materials as catalysts for selective oxidation of alkanes and biomass derived polyols.

6.17.1

Introduction

Over the last two centuries the development of catalytic science has been closely linked to the development of modern society. The world we live in today depends on the chemical industry, most notably the Haber Process, which supports a significant proportion of the Earth’s population via nitrogen fixation to produce fertilizers.1 Over 90% of chemical products utilize at least one catalytic step in their manufacture, making the development of effective catalysts essential not only to drive economic advancement but also reduce the environmental burden of many manufacturing processes.2 Many of the first examples of industrial heterogeneous catalysis build on the observations of Sir Humphrey Davy and the development of the Davy lamp, which demonstrated that a heated platinum gauze could facilitate the combustion of methane, hydrogen, and carbon monoxide. This is widely considered the first example of catalytic oxidation.3 The industrial applications of catalytic oxidation using finely dispersed metals or gauzes were quickly developed. In 1831, the first patent (P. Phillips, UK patent 6096) was filed concerning the manufacture of sulfuric acid via the oxidation of sulfurous acid with air in the presence of platinum. This process was operated to produce fuming sulfuric acid in 1875.4 The observation was quickly followed by demonstrations of ammonia oxidation to produce nitric acid using similar platinum based materials.5 Catalytic oxidation has played a key role in the subsequent development of the chemical industry based on petrochemical feedstocks as a source of carbon over the last century. Over the next 100 years, society must begin to shift from a linear to a circular economy with utilization of renewable alternative feedstocks such as biomass. Some of these challenges were already envisaged by Taylor in 1919 who suggested that Alcohol [ethanol] of the future should be obtained by the catalytic degradation of cellulose content of wood waste

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00001-7

381

382

Selective oxidation by mixed metal nanoparticles

in view of the increasing consumption and prospective exhaustion of fossil fuels

Taylor suggested this bio-derived ethanol should act as a starting point for further oxidative processes to produce acetic acid or acetone as solvents, in addition to proposing methane oxidation to yield formaldehyde as a precursor to polymeric electrical insulators and other “fancy articles.”6 Taylor also made significant contributions to the understanding of catalysis by proposing that a catalyzed chemical reaction does not occur over the entire surface of the solid material but instead at certain active sites.7 Almost a century later, the methods to synthesize these catalytic sites and characterize them beyond establishing comparative reaction kinetics is well developed. Modern catalytic chemists can synthesize catalytic materials containing well defined active sites in order to understand and develop detailed structure activity relationships. Beginning from the first examples of catalytic oxidation using precious metal gauzes, advances have been made in achieving greater utilization of precious metals by developing materials based on supported metal nanoparticles. The nanostructuring of these catalyst components increases the available metal surface containing coordinatively unsaturated metal centers available for reaction. These materials have significant potential in facilitating the selective oxidation of a range of functional groups in both the commodity and fine chemical industry and can contribute to the development of a more sustainable chemical industry.

6.17.2

Importance of selective catalytic oxidation

The modern chemical industry is built on fossil fuel derived hydrocarbons as a concentrated energy source. The total oxidation of these hydrocarbons via combustion processes releases the energy that the transport sector is currently reliant on. However, to allow these feedstocks to enter the chemical supply chain the partial or selective oxidation of these hydrocarbons is required, while avoiding the deep oxidation to carbon dioxide. Catalytic processes that can facilitate functional group conversions at mild conditions with low energy input are highly desirable. In addition, high reaction selectivity is key in designing processes that produce less waste and therefore consume less energy in subsequent purification steps. While many large scale oxidation processes are in operation, most notably for the production of high volumes of monomers and intermediates for the polymer industry (Table 1), there is still significant room for improvement.8 A common challenge is that the desired compounds are often kinetically favorable products rather than thermodynamically favorable. The oxidation of hydrocarbons with oxygen or air produces water and carbon dioxide as thermodynamic products when the desired products are typically alcohols, aldehydes, ketones, acids or alkenes. For example, the commercial oxidation of cyclohexane to cyclohexanol and cyclohexanone for adipic acid production must be operated at low conversions to prevent the accumulation of considerable amounts of over oxidation products due to the radical nature of the reaction.9 Other gas phase oxidation processes such as ethylene epoxidation using silver catalysts requires the use of promoters (in this case chlorine) to suppress over-oxidation. These additives must then be removed downstream at the cost of extra energy input in addition to the formation of undesired chlorinated by-products.10,11 A key driver in the development or improvement of new oxidation processes is the ability to avoid using stoichiometric oxidizing agents such as organic peroxides or oxidizing acids such as HNO3 and halogens. With sustainability in mind, the use of molecular Table 1

Selection of chemical building blocks synthesized by oxidation processes.

Organic building block

Capacity/Mta
1

Oxidant

Terephthalic acid Formaldehyde Ethylene Oxide 1,2-Dichloromethane Propylene Oxide Cyclohexanone Vinyl acetate Acrylonitrile Styrene Phenol/acetone Acrylic acid Adipic Acid Maleic anhydride Hydrogen Cyanide

44 19 18 18 8 6 6 6 5 5 5 3 2 2

Air Air O2 Air/O2/Cl2 Cl2/ROOH/H2O2 Air/O2 Air/O2 Air Air/O2 Air Air HNO3 Air Air

Adapted from reference Cavani, F.; Teles J.H. ChemSusChem 2009, 2, 508–534.

Selective oxidation by mixed metal nanoparticles

383

oxygen or air as an oxidant is highly desirable. The use of H2O2, which has a high active oxygen content, is only sustainable when the reduction in energy consumption or increase in product yield outweighs the environmental impact of H2O2 production via the anthraquinone process.12 To this end, supported metal nanoparticles which contain under coordinated transition metals represent a promising class of catalyst material considering the rich chemistry that can occur on interaction of O2 with a metal center or surface due to the bi-radical nature of molecular O2.13–16

6.17.3

Multimetallic nanoparticle catalysts

Despite the wide range of oxidation reactivity that can be achieved with monometallic nanoparticle catalysts, bimetallic nanoparticles have gained significant interest due to the often observed synergistic effects between the two metals.17 An obvious advantage is the increase in material composition space that can be achieved by combining multiple chemical elements into catalytic nanoparticles, enabling fine tuning of catalytic properties. Elements that are not favorably miscible at the bulk scale such as Au/Ru, Rh/Cu have been shown to have a decreased enthalpy of formation on the nanoscale. This is due to significant reductions in melting points as the particle size is decreased, meaning that the bimetallic alloys can be accessed by conventional chemical and thermal reduction methods.18,19 Bimetallic nanoparticles can also include the possible incorporation of late transition metals such as Ni, Fe, Cu as diluents for precious metals, significantly reducing the cost of the catalyst.20,21 In addition the possibility to include early transition metal elements such as Ti, Sc and Y in stable nanoalloys allows access to unconventional surfaces which, in some cases, can be stabilized against oxidation, to reduce the critical deactivation mechanism in bimetallic nanoparticle catalyzed oxidation reactions.22 Multimetallic nanoparticles offer high degrees of flexibility compared to monometallic analogues and effects inherent to the mixing of two metallic species in the nanometer size regime can be exploited to give enhanced catalytic properties by mechanisms such as electronic, lattice strain and ensemble effects depicted in Fig. 1. Electronic EffectsdThe binding strength of the reactive intermediates to the metal surface is often determined by the d-band center of the surface atoms in transition metal catalysis. By mixing metallic species with differing electronegativity, d-state hybridization and charge redistribution occurs between the metal species, resulting in changes in the density of states and positions of the d-band centers. These effects are often exploited in oxidation reactions by the combination of metals which bind oxygen strongly with those which bind oxygen weakly to give optimum properties according to the Sabatier Principle.23 This effect has been demonstrated recently in the study of AuPd bimetallic particles for the base free oxidation of alcohols such as benzylalcohol, 4-methylbenzyl alcohol and 4-phenethyl alcohol as shown in Fig. 2. By variation of the Au: Pd in the catalyst structure a correlation was possible between changes in charge at the Pd sites as a function of alloy composition and can be used as a predictive descriptor of catalyst performance. Using X-ray photoemission spectroscopy (XPS) and X-ray absorption spectroscopy (XAS) with XANES analysis a Au rich alloy was identified as the optimum catalyst with nine times the activity of the monometallic Pd catalyst as a result of optimum absorption strength of the primary alcohol to allow reaction and desorption of the products without surface poisoning by the

Electronic Effects

Lattice Strain Effects

d

eCompressive Strain Ensemble Effects

d

O=O Matched Lattice

O=O

O=O

d

Expansive Strain Fig. 1

Schematic representation of possible effects that can be exploited to design high activity bimetallic catalysts for selective oxidation reaction.

384

Selective oxidation by mixed metal nanoparticles

Fig. 2 (A) Relationship between change in Pd electronic state and TOFPd (red line) of the AuPd alloys and the entropy of activation for the oxidation of benzyl alcohol to benzaldehyde (navy line). (B) Comparison of the TOFPd values of the AuPd catalysts in the oxidations of 4-methylbenzyl alcohol (black line), 4-phenethyl alcohol (red line) and 4-methoxybenzyl alcohol (blue line). Figure reproduced without editing from reference Zhu, X.; Guo, Q.; Sun, Y.; Chen, S.; Wang, J.Q.; Wu, M.; Fu, W.; Tang, Y.; Duan, X.; Chen, D.; Wan, Y. Nat. Commun. 2019, 10, 1428 under a Creative Commons Attribution 4.0 International License.

substrate.24 Variables such as this can therefore be modelled in bimetallic catalyst systems by determining the optimum binding strength of surface intermediates over model surfaces as a descriptor for activity with the aim to identify the optimum surface coverage and binding strengths of reactants and intermediates.25–27 Strain EffectsdThe difference in atomic radii between two metals in a nanoparticle can lead to a compression or expansion of the bimetallic metal lattice. This compression or expansion of the surface lattice results in changes in position of the d-band center and hence can affect the binding strengths of intermediates at the particle surface which control catalytic activity. An example of this effect in controlling catalytic activity was recently reported by Tsung et al. when using PdxPty octahedral particles which contained a Au core in formic acid electrooxidation. They observed that while changing the Pd: Pt had an effect on activity due to the changing surface composition of the catalyst, the incorporation of a Au core led to a significant increase in observed current density due to perturbation of the PdxPty lattice, increasing the activity of the catalyst by over 4 times with the introduction of a 0.5% strain to the PdxPty lattice by the underlying Au particle.28 Ensemble EffectsdIn addition to electronic and lattice strain effects that can control catalytic activity, defined structural motifs within the bimetallic alloy can also determine catalytic activity and selectivity. These include breaking of extended monometallic metal surfaces, the presence of isolated metal atoms or dimers/small islands of a second metal embedded into an extended metal surface, or active sites comprised of adjacent metal atoms of different elements. These effects are prevalent in many oxidation reactions such as the incorporation of base metals into Au nanoparticles to enhance activation of molecular oxygen at the interface between the two components.29 This effect is relevant in selective oxidation reactions as the recent report by Henkelman and coworkers demonstrates that single atoms of Ir, Pd, Pt and Rh in extended surfaces of Au, Ag, and Cu led to high activity and selectivity in the partial oxidation of ethanol to acetaldehyde.30 These ensembles containing metals which bind oxygen strongly are extremely important in controlling the nature of surface oxygen species that lead to either total oxidation (from high surface coverage and cleavage of molecular oxygen) or selective oxidation (resulting from low surface coverage of a defined oxygen species) in the upgrading of hydrocarbons such as ethylene31 or propene32 to high volume commodity epoxides. In addition, a bimetallic catalyst may also operate as a bifunctional catalyst in a number of ways: (i) surface separation of nanoparticles can allow subsequent reactions to cascade as if two monometallic catalysts were present29 (ii) additional metals can be incorporated that bind and convert catalyst poisons and prevent accumulation on the catalyst surface without directly contributing to the reaction of interest33 (iii) additional intrinsic properties such light absorption by localized surface plasmon resonance using one metal component can be used to promote reactivity on the second metal.34 To fully elucidate the structure-activity relationships of these catalytic materials the complex nature of their surface structure needs to be understood. This includes the chemical nature or the oxidation state of the constituents, particles size and uniformity of the composition, and nanoscale structure of the particles. To this end we will focus on classes of bimetallic catalyst materials that have well defined structures and potential applications in selective oxidation reactions.

6.17.4

Classes of well-defined bimetallic nanoparticle catalysts

6.17.4.1

Single atom alloys (SAAs)

The use of supported single metal ions or atoms as single atom catalysts (SACs) represents the greatest possible utilization of catalytic metals.35 In these systems uniformity is theoretical possible by controlling the nature of the eO, eN, eC or Cl environment

Selective oxidation by mixed metal nanoparticles

385

typically surrounding the metal ion, however in reality multiple metal speciations are likely to be present depending on the anchoring point of the metal to the support material. By extension, single atom alloys consist of well-defined bimetallic nanostructures where one, typically catalytically active metal, is atomically dispersed into another host metal which is usually of lower activity but higher selectivity for the desired reaction. This motif allows facile activation of the reactants at the atomic site and weak adsorption of the reaction intermediates on the host surface which is crucial in many selective catalytic processes. The potential of a wide range of metal alloys such as PdCu,36 PtCu37,38 and PdAg,39 has been established in the field of selective hydrogenation by a mechanism which involves H2 activation at the atomic dopant site such as Pd followed by spill over to the weakly adsorbed reactants on the host metal surface. In contrast, many oxidation reactions require the cleavage of molecular oxygen with numerous studies having shown that this does not occur on isolated Pd sites and requires at least dimers of Pd in the metal surface.40 Despite this a number of oxidative process are being explored using these well-defined catalyst materials. Recent studies have focused on the modification of Au nanoparticles with atomic dispersions of Ni as an approach to improve the yields of oxidative cross coupling of aldehydes and alcohols.41 The production of methyl methacrylate through the oxidative cross coupling of methanol and methacrolein has been carried out with oxygen and Pd catalysts, but typically results in low yields due to over oxidation.42 The use of Au results in higher yields, but typically requires the addition of a base or basic support material to achieve this. The addition of Ni in a Ni: Au ratio of 0.005: 1 resulted in a twofold rate enhancement while maintaining 100% selectivity without the addition of base according the proposed mechanism shown in Fig. 3. The addition of Ni results in stronger binding sites for MeOH stabilization of the hemiacetal intermediates, and increases the number of active Au-Ox sites. However, the role of Ni in O2 activation remains unclear. A similar class of materials based on Pd0.01Cu/SiO2 nanoparticles have been investigated for the oxidation of methanol to methylformate. While Cu particles alone are not active at 160  C, Pd nanoparticles achieved high levels of conversion however show no selectivity towards methylformate, instead producing only CO2.43 The Pd0.01Cu/SiO2 catalyst was able to achieve 65% conversion with selectivity of 92% towards with methylformate with only 10% CO2 formation. Characterization of the fresh material by XANES demonstrates that initially the Pd is in the metallic form with the EXAFS analysis showing only PdeCu interactions at 2.2 Å with no PdePd or PdeO interactions. The catalyst shows stable activity and after reaction XANES and EXAFS analysis show the formation of cationic Pd centers with oxygen neighbors (and possible oxidation of neighboring Cu sites) with no evidence of formation of bulk PdO which returns to PdeCu interaction on reduction of the used catalyst. This result indicates that in SAA catalysts the reversible redox behavior of Pd(0)/Pd(II)eO species without agglomeration of the cationic Pd species and may represent the active phase or ensemble in this oxidation reaction. Despite the relative maturity of SAA catalysts in hydrogenation processes to date relatively little has been reported about their oxidation chemistry with molecular oxygen. Applications have been reported in non-oxidative dehydrogenation of ethanol and other alcohols due to the ease of CeH activation which could lead to applications in acceptorless dehydrogenation reactions to produce H2 and realize hydrogen storage in organic molecules.30,44,45 Other high temperature processes such as oxidative methane reforming also have potential by embedding active sites into surfaces which do not support coke formation if extended thermal stability at harsh conditions can be shown.46

Fig. 3 Proposed reaction scheme for the oxidative coupling mechanism of methacrolein with methanol over Ni-AuOx sites. Reprinted from reference Trimpalis, A.; Giannakakis, G.; Cao, S.; Flytzani-Stephanopoulos, M. Catal. Today, 355, 2020, 804–814, without modification with permission from Elsevier.

386 6.17.4.2

Selective oxidation by mixed metal nanoparticles Near surface alloys

Multimetallic metal nanoparticles rather than isolated metal atoms in surfaces will continue to play a key role in heterogeneous catalytic processes due to the possibility of achieving improved performance compared to monometallic particles. The vast compositional variations that can be achieved with multimetallic particles opens the possibility of designing optimized active sites, which show synergistic effects, compared to the constituent components. The chemical composition of a metal nanoparticle surface is often significantly different to nominal composition defined in the preparation and relative to the bulk and is largely driven by surface free energy in addition to other parameters such as strong metal support interactions, surface adsorbates, particle size effects and surface coverage of adsorbates.47 This can lead to significant variations and evolution of the catalyst surface via surface segregation driven by variations in the reactive atmosphere the nanoparticle is exposed to as the reaction proceeds. Under vacuum, the surface of a nanoparticle is often enriched by the element that has the intrinsically lower surface energy, however in reactive environments the strength of interaction with the adsorbates can drive surface segregation. This is practically relevant in high temperature oxidation processes using metallic nanoparticles where the balance between surface oxidation/passivation and alloy segregation is a delicate interplay and can lead to significant induction periods prior to reaching steady state activity. This can often be clearly observed in the calcination of bimetallic particles where one metallic component is significantly more oxophillic than the other. An example observed during catalyst preparation is the calcination of AuPd bimetallic particles, which have been shown to be active catalysts for reactions such as the benzyl alcohol and toluene oxidation. A 2.5 wt% Au/2.5 wt% Pd/Al2O3 catalyst reported by Hutchings and co-workers prepared by wet impregnation methods, shown in Fig. 4, exhibits a homogeneous distribution of Au and Pd after drying at 120  C.48 On calcination at high temperatures up to 400  C under static air, a clear evolution of the nanostructure to form a core shell structure containing a PdO shell with a Au rich core driven by the stronger interaction of Pd with the oxidizing atmosphere. Catalysts containing AuPd alloys have been extensively studied in a wide variety of selective oxidation reactions including the selective oxidation of glycerol49 and the oxidation of carbon monoxide.50 Mullins and co-workers systematically studied the effect of the atomic composition of the alloy surface on the activity towards both O2 dissociation and oxidation reactions.51,52 By preparing a range of model surface alloys by evaporation of Pd onto Au (111) surfaces it was possible to prepare materials with both AuePd interfaces in materials that have low Pd coverage and Pd(111) islands in materials with high Pd coverage. It was shown that the existence of the Pd(111)-like islands is directly linked to the dissociative adsorption of O2 at temperatures above 300 K and the activation energy of O2 dissociation decreased with increasing Pd coverage as shown in Fig. 5. Using CO oxidation as a test reaction it was observed that the rate of CO2 generation at low temperatures (below 200 K) is strongly related to the quantity of PdeAu interface sites. Their findings suggest that oxygen molecularly adsorbs on the PdeAu bimetallic surface as a precursor to dissociative O2 chemisorption, just as with pure Pd surfaces, and additionally, the enhanced reactivity of adsorbed atomic oxygen originates at the interfaces between Pd and Au domains. These model studies are key in

Fig. 4 HAADF image showing development of the core–shell structure in 2.5 wt%Au–2.5 wt%Pd/Al2O3 catalyst, Au green, Pd blue, Al2O3 red. Top row, dried at 120  C; middle row, calcined at 200  C; bottom row, calcined at 400  C. Adapted with permission from Edwards, J.K.; Freakley, S.J.; Carley, A.F.; Kiely, C.J.; Hutchings, G.J. Acc. Chem. Res. 2014, 47(3), 845–854. Copyright 2014 American Chemical Society.

Selective oxidation by mixed metal nanoparticles

387

Fig. 5 Activation barriers for CO oxidation (blue) and O2 dissociation (red) on various amounts of Pd-deposited PdeAu(111) surfaces. Adapted with permission from Han, S.; Mullins, C.B. ACS Catal. 2018, 8(4), 3641–3649. Copyright 2018 American Chemical Society.

developing further understanding of nanoparticle systems and could clearly demonstrate how the activity of a catalyst could change if the surface composition evolves over time from isolated AuePd sites to Pd(111) like islands on the Au surface resulting in a change in the nature of the oxygen species present. A similar approach was taken by Zhu et al. by studying a range of AuPd nanoalloy catalysts with various compositions supported on mesoporous carbons for base free oxidation of primary alcohols.24 Through characterization by advanced electron microscopy and X-ray absorption spectroscopy they identified that Pd single atoms and dimers present in the Au surface were the active sites for the base free oxidation of primary alcohols where AuPd nanoparticles catalysts have been shown to have significantly higher activity than either monometallic Au or Pd catalysts. The changes in Pd nuclearity in the surface with varying AuPd ratios results in changes in the d-band charge which is shown to be optimum at Pd 33–50 at.% where the population of single Pd and Pd dimers is high. This resulted in a pronounced effect on the absorption strength of primary alcohols and therefore higher aerobic oxidation activity and selectivity, however in this case the nature of the oxygen species was not evaluated as in the previous study. As the surface of a catalytic nanoparticle often evolves on exposure to reactants and as catalysis is a surface phenomenon, the understanding of these structures under reactive conditions is critical. Studies by Bukhitiyarov clearly demonstrate this effect and its reversibility using PdeAu bimetallic structures supported on highly oriented pyrolytic graphite (HPOG) using near ambient pressure XPS.53 The study reveals the flexibility of the structures on exposure to CO at various reaction temperatures. At low temperatures of about 150  C the homogenous surface is segregated to become Pd rich due to the strong interaction of Pd with CO. This structure reverts to an alloy above the temperature of CO desorption and catalytic activity towards CO oxidation is observed to correspond with a return to the homogenous alloy structure. This process is reversed on cooling under reaction conditions leaving an ex situ post reaction sample appearing to have a Pd enrichment at the surface which can lead to a misunderstanding of the structure-activity relationships. As the catalyst surface is often under non-equilibrium conditions and materials which do not typically form bulk alloys can readily mix at the nanoscale it is crucial to further develop methods to not only characterize the bulk but the surface under reaction conditions. Techniques such as X-ray absorption spectroscopy and ambient pressure X ray photoemission spectroscopy can be used to reveal any difference in surface compositions compared to the bulk under reaction conditions.54–56 While exciting progress is being made using model surfaces to identify these structure activity relationships, further development is needed with studies being performed on real catalytic systems containing bimetallic nanoparticles.

6.17.5

Cluster beam deposition catalyst synthesis

Methods for the synthesis of bimetallic nanoparticles are mostly reported using wet chemical methods which often involve the generation of large amounts of waste containing residual chemical agents which can be hazardous. An emerging route to the production of well-defined nanoparticles by a “green” solvent free process are cluster beam deposition (CBD) methods.57 The gas-phase synthesis of metal nanoparticles for catalysis has evolved from the development of cluster beam technologies. These rely on the homogeneous nucleation of metal particles from a super-saturated vapor (metal atoms in a inert gas) followed by controlled condensation and coalescence. A popular and versatile approach is magnetron-sputtering inert gas condensation which has allowed a wide range of nanoparticle structures to be formed in the gas phase by varying parameters during the synthesis such as: magnetron power, inert gas pressure, and condensation time with simultaneous analysis of the size distribution of the nanoparticle beam by TOF-MS, a schematic of which is shown in Fig. 6A. More recently new metal targets have been developed to allow the

388

Selective oxidation by mixed metal nanoparticles

Fig. 6 (A) Dual-magnetron sputtering, gas condensation cluster source, in which the positively charged clusters or deposited onto agitated powders with mass monitoring. (B) A MACS, in which a solid matrix is impacted with an Ar ion beam to sputter the metal, causing clusters to nucleate, ripen and eventually deposit onto surfaces. Reproduced with permission from Palmer, R.E.; Cai, R.; Vernieres, J. Acc. Chem. Res. 2018, 51, 2296–2304, https://pubs.acs.org/doi/full/10.1021/acs.accounts.8b00287, further permissions related to the material excerpted should be directed to the ACS.

Selective oxidation by mixed metal nanoparticles

389

synthesis of multiple component nanoparticles using two independent sources which can create bimetallic nanoparticles with composition, and in some cases morphology, controlled by the evaporation parameters. Recent developments such as the matrix assembly cluster source (MACS) which relies on the impact of an ion beam into a condensed matrix of rare gas and metal atoms as shown in Fig. 6B can results in much larger fluxes of metal clusters. This has recently been demonstrated to achieve a metal cluster flux high enough to produce 1 g of catalyst (1 wt%) in 1 h by landing these clusters onto a commercial catalyst support and are being validated in heterogeneous catalysis, electrocatalysis and sensing applications.58 Monometallic catalysts containing Au nanoparticles prepared by depositing 2–5 nm clusters onto TiO2 powder have been shown to be comparable in activity to equivalent catalysts prepared by colloidal methods for base catalyzed glycerol oxidation, showing high selectivity to glyceric acid in both cases with similar TOFs. Bimetallic AuCu nanoparticles have also been prepared using this methodology with controlled compositions and particle size and have been shown to be active for the hydrogenation of nitrophenol, however the application of bimetallic particles prepared by CBD method in selective oxidation is limited.59 One example reported by Vajda and co-workers demonstrated the synthesis of ultrasmall size-controlled clusters of Pt10, Ag9Pt2 and Ag9Pt3 and deposited them onto a planer amorphous Al2O3 surface prepared by ALD methods.60,61 The planar materials were tested under realistic CO oxidation conditions and showed some of the highest TOFs reported to date for this class of materials at 250–300 molecules per second per metal atom at 300  C with minimal deactivation even at the elevated temperatures used. High activity and reaction stability were the result of a synergistic role of Ag and Pt in the deposited nanoclusters, in which Pt anchors the clusters to the alumina and strongly binds CO molecules, while Ag activates O2, and Ag/Pt surface proximity prevents deactivation by either CO poisoning or over oxidation by promoting fast turn over. Cluster beam deposition methods provide a promising route for the production of well-defined bimetallic nanoparticles, especially in the sub nm cluster size range which is often difficult to access by wet chemical methods without significant ligand stabilization of dilute metal clusters in solution. Significant benefits could be realized due to the potential to develop solvent free catalyst synthesis with controlled size and composition of bimetallic nanoparticles if larger scale continuous applications are realized. This could provide extensive catalyst material libraries for rapid catalyst screening and has the potential to produce nanoparticles containing uncommon elements such as PteTi or AueTi core–shell nanoparticles (NP) of specified single particle mass for applications in catalysis.62

6.17.6

Bimetallic nanoalloys

Building from single atom alloys, bimetallic surface alloys and bimetallic sub nanometer cluster synthesis, a more prevalent class of well-defined bimetallic structures are core shell bimetallic nanoparticles in the size range 2–10 nm. Ferrando et al. defined four possible mixing patterns for bimetallic particles (Fig. 7), (A) core shell (B) segregated nano alloys (C) homogeneously mixed nano alloys (D) multi shell nano alloys where the core consists of one metallic components surrounded be a shell of the second metallic component.63 Bimetallic nanoalloys as a research field has been extensively reviewed by Paria and co-workers in the context of bimetallic metallic/inorganic materials64 and Hutchings and co-workers in the context of sustainable catalysis.65 Many possible combinations of bimetallic species have been reported (too extensive to mention in this chapter) however the type and stability of structures that can be obtained depends on a number of factors including: (i) the relative miscibility of the two metals (although bulk miscibility is not a strict prediction of mixing on the nanoscale) (ii) the surface energy of the two metals with the lower surface energy components tending to favor forming the shell (iii) relative atomic sizes and ease of charge or electron transfer – electron transfer between two components favors mixing (iv) adsorbed species – if an surface species such as a stabilizer ligand strongly favors binding to one component this may be drawn to the surface as described earlier in the effect of CO on AuPd alloys. An additional parameter which is crucial in determining the nanostructure of bimetallic catalysts is the preparation method. Selective oxidation catalysts consisting of bimetallic nanoparticles can be prepared in a variety of ways such as gas phase deposition (as described above), electrochemical reduction of metal species onto an electrode surface, radiolysis and sonochemical methods, thermal decomposition of metal precursor mixtures, and the most commonly used solution based chemical reduction methods. The chemical reduction method commonly applied to monometallic nanoparticles can be translated to bimetallic catalyst systems. The method involves a dilute metal precursor solution in a suitable solvent which is then mixed with a reducing agent to initiate nucleation and growth of metal nanoparticles. Parameters such as the metal concentration, strength of the reducing agent (such as H2, NaBH4, N2H4, ascorbic acid) and the presence of stabilizer agents (such as polyvinyl alcohol (PVA), thiols, polyvinylpyrolidone (PVP)) to cap the growing nanoparticles and stabilize a nano dispersion can all contribute to the final catalyst material properties. While the translation of this method from monometallic samples to bimetallic samples seems simple on first inspection, control of the reduction process is crucial. The conditions must allow the simultaneous reduction and nucleation of both metal ions in bimetallic precursors solutions to efficiently form bimetallic nanoparticles. If the rates of reduction are mismatched each metal will nucleate and grow in isolation resulting in two distributions of monometallic particles. This was reported in the case of AueCu particles when using a strong reducing agent such as NaBH4 by Sra et al., resulting in isolated populations of Au and Cu particles which required further thermal treatments to generate bimetallic particles.66 This mismatch can also result in the formation of inhomogeneous distributions of nanoparticles as in the case reported by Li et al., in the preparation of Ni containing bimetallic particles using hydrazine.67 One approach to overcome this is to separate the metal reduction steps by forming seeds of one metal after reduction and reducing the second metal onto the surface of the prepared monometallic nanoparticles – the final morphology

390

Selective oxidation by mixed metal nanoparticles

Fig. 7 Cross sectional representation of possible bimetallic mixing patterns: (A) core shell (B) segregated nano alloys (C) homogeneously mixed nano alloys (D) multi shell nano alloy. Adapted with permission from Ferrando, R.; Jellinek, J.; Johnston, R.L. Chem. Rev. 2008, 108(3), 845–910. Copyright 2008 American Chemical Society.

can then be controlled by appropriate heat treatments or by selecting stabilizing ligands of appropriate strength to slow particle growth allowing the use of a stronger reducing agents.68 The breadth and flexibility of this colloidal synthesis of bimetallic nanoparticles makes it far too big a subject to review in this chapter however several excellent reviews on the topic are already available. The following section will focus on the range of selective oxidation reactions that can be catalyzed by bimetallic nanoparticles of the classes described above and highlight their potentials in achieving sustainable chemical industry.

6.17.7

Selective oxidation reactions

6.17.7.1

Catalytic oxidation of alkanes

While society is making efforts to move towards renewable and sustainable chemical feedstocks, there is still a strong reliance on crude oil. The majority of hydrocarbons which make up crude oil are alkanes of various molecular weights. While alkanes themselves can be used for fuels, the transformation by selective oxidation to high value added petrochemical products is essential. The activation of CeH bonds to produce alcohols or carboxylic acids is often characterized by low yields and the formation of unwanted side products.69 These limitations are amplified when longer hydrocarbon chains are used, increasing the difficulty of selective oxidation while minimizing cracking processes.70,71The intrinsic stability of the CeH bond results in a typically high

Selective oxidation by mixed metal nanoparticles

391

activation barrier with a high bond dissociation energies and a HOMO/LUMO poorly orientated to interact with catalytic transition metal centres.72 Other than for the petrochemical industry, an alternate need for alkane oxidation can be seen in the reduction of greenhouse gas emissions. Methane can be released as a pollutant from the oil industry, decaying biomass and agriculture. According to the World Meteorological Organization, the atmospheric methane concentration has seen a larger percentage growth than both CO2 and N2O since the industrial revolution. As of 2018, the methane concentration in the atmosphere sat at 1856 ppb, and is rising steadily.73 Consequentially there has been much research in the field of methane oxidation, both as a syngas supply for industrial processes and to reduce the environmental impact of the greenhouse gas. In accordance with both experimental and computational studies it could be accepted that Pd/Al2O3 is one of the most active catalysts for the activation of methane.74,75 However, it was found that the stability of Pd catalysts was limiting activity as water formation was found to cause hydroxyl accumulation on the catalyst, preventing the binding of O2 and therefore leading to deactivation.76 There is also evidence of deactivation by coking if the partial pressure of oxygen is too low, and sintering effects at high temperatures.77,78 In order to increase the activity of methane oxidation catalysts various bimetallic catalyst compositions have been reported with a view to both increase reaction rates of elementary steps and prevent deactivation. Zou et al. designed a catalyst with a NiO core surrounded by a shell of PdO (PdO being the active form of Pd) immobilized on an alumina support.79 The addition of the NiO core allowed the active PdO to be formed more quickly, stabilizing the tetragonal lattice. It was also found that there was suppression of water adsorption on the catalyst surface, leaving active sites free, this allowed the bimetallic catalyst to operate at between 90% and 100% conversion of methane for the 50 h, while the conversion using the Ni free analogue dropped to 60% over the course of the time on stream. The distribution of Pd on the surface of the NiO also allowed for more effective utilization of the Pd active sites. NiO plays an important role in the re-oxidation of Pd after a catalytic cycle, therefore the adjacent positioning of the two metals is highly beneficial. Similar finding were observed by Feng et al., when they synthesized a catalyst consisting of a Pd core and ceria shell.78 A 10 nm core was encapsulated with 5 nm ceria particles to achieve a total radius of approximately 45 nm. Feng’s objective was also to promote the formation and stability of PdO, this was achieved by utilizing the good oxygen storage properties of the ceria shell. The adsorption of a large amount of oxygen would also benefit the stability of the catalyst by preventing detrimental amounts of water adsorbing on the surface. In the work of both Feng and Zou it seems to be the interaction between Pd and oxide that was key to maximizing the rate of methane oxidation, along with allowing sufficient surface area (free from coke deposits or adsorbed water) of PdO to which substrate can bind.78,79 There are examples in the literature of non-Pd catalysts being used for catalytic methane oxidation. Neal et al. synthesized a catalyst with Fe2O3 core and LaSrFeO perovskite shell, designed to oxidize methane.80 Using this catalyst, lattice oxygen could be used as the oxidant. It was shown that there is preferential depletion of the oxygen in the oxygen-dense Fe2O3 core over the perovskite lattice, this was attributed to the greater lattice energy of LaSrFeO. The transport of oxygen between the core and shell allowed methane to be oxidized on the catalyst surface without the need for simultaneous O2 binding. The fully spent catalyst consisted of an iron core surrounded by the intact perovskite. Re-generating the catalyst involved the reduction of CO and CO2, allowing lattice oxygen to migrate back into the core, re-oxidizing Fe.80 After methane, the most abundant compound in natural gas is ethane.81 Ethane has many uses in industry, an example being dry reforming, where it is oxidized with CO2 to produce syngas. An incomplete oxidation of ethane produces ethylene, a monomer widely used in petrochemicals. Myint et al. have devised a bimetallic catalyst which is capable of selectively oxidizing ethane to ethylene in the presence of CO2.82 A CeO2 support, with immobilized nanoparticles of a Co and Pt alloy was chosen as a starting point due to its previous appearance in the literature.83 As it had been evidenced that the alloy had better activity and selectivity than either monometallic counterpart, a number of different bimetallic alloys were employed to try and improved the oxidation reaction. It was found that CoPt, CoMo, and NiMo favored the production of syngas, while FeNi was effective in performing oxidative dehydrogenation via CeH bond cleavage to form ethylene.82 For each of these catalysts a ceria support was used. Common in alkane oxidation, ceria again displays excellent ability to act as a source of oxygen, easily regenerated via the C]O scission of CO2.84,85 A ceria support also decreases catalyst inactivity due to coking, a common problem in the oil industry. The abundance of active lattice oxygen means that any carbon deposits are easily oxidized and removed.86,87

6.17.7.2

Biomass oxidation

Carbohydrates and molecules that can be derived from further processing of natural sugars have recently attracted significant attention as renewable alternatives to petroleum based feedstocks for a wide range of industrial chemicals. The transformation of these molecules will form a key component of the transition from a linear to a circular chemical industry, therefore there is a continued drive for clean catalytic reactions to replace existing processes that rely on stoichiometric reagents. Despite biomass having typically high oxygen contents there is still a requirement for the development of a number of selective oxidation reactions that utilize molecular oxygen and suppress over oxidation and the production of CO2.

392

Selective oxidation by mixed metal nanoparticles

6.17.7.3

Glucose oxidation

The selective oxidation of glucose to mono (gluconic) or di (glucaric) acids represents a potential green route to commodity chemicals such as adipic acid. Glucaric acid is classified as one of the 12 most important building block chemicals by the US DoE and can be transformed further into a wide variety of commodity chemicals in the food and paper industry in addition to being applied in cancer prevention and control.88,89 Bimetallic oxidation catalysts developed for the selective oxidation of glucose to gluconic and glucaric acid using molecular oxygen (Scheme 1), as opposed to the processes using nitric acid or TEMPO/hypochlorite oxidizing agents, could contribute greatly to the sustainability of this process and also have applications in the upgrading of other polyols such a glycerol and furanic alcohols (discussed below). Some of the earliest reports concerning glucose oxidation to gluconic acid utilized Pt/C as a catalyst a mild reaction conditions (35–80  C, ambient pressure in aqueous media). While these catalysts were active they suffered from loss in activity over time at neutral pH suggesting that the formation of acids could lead to catalyst deactivation by promoting metal leaching or product inhibition of further reactivity by surface poisoning.90 Maintaining the pH above pH 7 by constant addition of NaOH minimized catalyst deactivation but complicated product isolation due to the production of an alkali metal salt instead of free acids in addition to requiring neutralization of the reaction mixture post reaction.91 Studies have also shown that the oxidation of gluconic acid over these Pt catalysts does not occur until all of the glucose has been consumed, indicating a stronger absorption of glucose than gluconic acid to the catalyst surface, suppressing over oxidation.92,93 Alkali promoters are known to accelerate glucose/polyol oxidation, however the presence of base can also accelerate reactions leading to the cleavage of CeC and CeO bonds and reduced selectivity. This lead to the development of the first bimetallic catalysts reported for glucose oxidation containing Pd nanoparticles promoted by bismuth which have been shown to be able to achieve close to 100% selectivity towards gluconic acid. The addition of Bi to Pd/C catalysts has been studied extensively in terms of the Pd particle size and bismuth precursor but in many cases leaching of bismuth into the reaction solution was detected meaning the promotional effects could only be considered quasi heterogeneous.94,95 This has driven the need to develop new catalysts that can enable glucose oxidation to gluconic/glucaric acids under base free conditions or without pH control. Gold and gold containing bimetallic catalysts have been shown to be selective towards the formation of gluconic acid in the absence of additional alkali promoters. Biella et al., proposed the use of Au particles prepared by a colloidal method and demonstrated a higher activity than Pt, PdeBi and PtePdeBi trimetallic catalysts.96 Crucially the Au catalyst system was able to operate at much wider pH ranges than the Pt-group catalysts with deactivation mainly resulting from metal leaching under the acidic conditions generated by the production of gluconic acid. Sayago et al., reported a range of Au nanoparticle based catalysts supported on CeO2 and Al2O3 and ZrO2 and demonstrated a relationship between the Lewis acidity of the support and CeC cleavage to produce lactic acid in the reaction.97 The effect of bimetallic catalysts was demonstrated by Comotti who synthesized colloidal sols containing AuePt, AuePd and AueRh particles and tested them under mild reaction conditions both with and without pH control.98 Whereas the activity of single metals was low (Au and Pt – 51-60 h
1) the activity of the bimetallic particles resulted in significant improvements in TOF (AuePt - 924 h
1). In the presence of base the activity of all monometallic and bimetallic samples increased and the synergistic effects were less pronounced. The fine-tuning of the ratio of gold to platinum led to an optimized ratio of Au: Pt of 2: 1 (molar) and with this catalyst at pH 9.5 a TOF of 17,600 h
1was observed. More recently Hutchings and co-workers reported the activity of a range of AuPd nanoparticles prepared by similar colloidal methods for the aerobic oxidation of glucose at mild conditions without the addition of base.99 The use of a basic support material, in this case MgO/Mg(OH)2, in place of an alkali metal promoter, such as Bi, allowed glucose oxidation to gluconic acid to be carried out under base free conditions and ambient air pressure, significantly milder reaction conditions that previously reported. Despite this it was apparent that some of the Mg(OH)2 was solubilized during the reaction presenting a challenge for continuous operation, however control experiments determined the levels did not significantly promote the reaction.100 Despite the progress that has been made in using monometallic Au based catalysts for this oxidation reaction at mild conditions without the addition of base, generally Pt is the only known metal that can promote the secondary oxidation of gluconic acid to glucaric acid which occurs at typically much harsher conditions. A significant result was reported by Derrien et al. in which they demonstrated that a Au-Pt/ZrO2 catalyst was able to show a 50% yield of glucaric acid at complete conversion of glucose and gluconic acid at elevated temperatures and pressure (100  C, 40 bar O2). Recent patent literature has also focused on Pt containing catalysts, however the role of heterogeneous and homogeneous Pt species has not been full elucidated and could significantly complicate product analysis.101 Despite this there is significant scope to explore Pt based alloy catalysts for this process. Shi et al. recently reported a comprehensive study into bimetallic Pt catalysts and the effect of support, synthesis, and choice of second metal on glucose oxidation to glucaric acid at 80  C and 15 bar O2.102 They observed promising activity for a Pt-Cu/TiO2 catalyst representing a first move away from precious metal catalysts. The catalyst contained small PteCu particles (2.8 nm in diameter) consisting of PteCu spherical alloy particles (Fig. 8A). The catalyst achieved 60% glucaric acid selectivity in one step glucose OH

OH

OH O

HO

O OH

HO OH

OH

D-glucose Scheme 1

OH

[O]

Oxidation of glucose to valuable C6 acids.

OH

OH

D-gluconic acid

O

OH

OH

[O]

OH

HO OH

OH

O

D-glucaric acid

Selective oxidation by mixed metal nanoparticles

393

Fig. 8 (A) HRTEM and EDX images of 5% Pt–Cu/TiO2 reported by Shi et al. as an effective catalyst for glucaric acid formation. (B) Time on line of glucose oxidation showing formation of glucaric acid at glucose conversion 300 nm) from a Xe lamp. c Determined by inductively-coupled plasma optical emission spectroscopy. d Not analyzed. b

doped with Al3þ in the presence of a SrCl2 flux using an alumina crucible, and the resulting photocatalyst provided an AQY of 56% at 365 nm during photocatalytic overall water splitting.18 It is probably possible that Y3þ ions were substituted at both Ti4þ and Sr2þ sites, which likely decreased the activity. It has also been reported that the use of especially pure precursor materials provides more active doped SrTiO3, possibly by minimizing unintentional doping with undesired impurities.19 The effect of doping on the dynamics of charge carriers generated in SrTiO3 has been investigated using transient absorption spectroscopy,19,20 and significant spectral changes were found after doping. Specifically, doping with Al3þ strengthened the signal attributed to shallowly trapped electrons and extended the lifetime of such electrons.20 This is important because the water splitting activity of SrTiO3 is correlated with the concentration of shallowly trapped electrons, and so these electrons are thought to react with water to produce hydrogen. Conversely, there was no clear correlation between the concentration of deeply trapped electrons and the water splitting activity, because deeply trapped electrons are less reactive with water.19,20 However, when a hydrogen evolution cocatalyst was loaded, electrons that would otherwise be deeply trapped seemed to be utilized in hydrogen evolution, suggesting that electrons could be trapped by the cocatalyst, or that trapped electrons were transferred to the cocatalyst.19

6.18.2.2

Visible light absorption via excitation of mid-gap states

In the early days of the development of particulate photocatalysts for solar energy conversion, TiO2 and SrTiO3 were both reported to promote the overall water splitting reaction.35,36 Even so, both these materials have a bandgap wider than 3 eV and can only absorb the ultraviolet component of sunlight. In order to realize efficient solar energy conversion, it is necessary to modify wide bandgap semiconductor photocatalysts such as these so that they are active under visible light. As early as the 1980s, research showed that doping wide bandgap semiconductors with transition metals induced coloration of these materials, indicating the absorption of visible light.11,12 The associated theory holds that visible light absorption by these materials is related to electronic transitions involving impurity levels formed by the d-orbitals of the dopants in the bandgap of the host compound.37 Subsequent studies demonstrated that anionic doping can also have a similar effect.38 However, impurity levels derived from adding a dopant will have a relatively low density of states, so that the corresponding visible light absorption is often weak. The minimal interaction between these dopants also results in discrete orbitals and poor conductivity of photoexcited carriers. Moreover, dopants induce the formation of defects. As a result of defect formation, the photocatalytic activity of the host material under ultraviolet light is decreased significantly by doping, and the activity under visible light also tends to be low.39,40 At the beginning of the 21st century, doped oxides were applied to photocatalytic water splitting but, until very recently, only in the form of Z-scheme systems. It is therefore necessary to take a closer look at the role of dopants in order to effectively utilize visible light absorption induced by dopants in the photocatalytic water splitting reaction. This section discusses the doping of TiO2 and SrTiO3 as representative and well-examined systems, although it should be noted that other oxides,41–46 carbon nitrides47–49 and sulfides50 have also shown visible light absorption after doping.

6.18.2.2.1

TiO2

TiO2 doped with transition metals or anions exhibits visible light absorption. According to theoretical calculations, when rutile-type TiO2 is doped with a 3d transition metal, an impurity level consisting of the dopant t2g states is formed in the bandgap.37 The potential of this impurity level shifts positively as the atomic number of the dopant increases. Therefore, when added at high concentrations, the impurity levels of some dopants hybridize with the conduction or valence band of the host material and delocalize, leading to band edge tailing and extending the absorption edge toward longer wavelengths. In contrast, the doping of anatasetype TiO2 with nitrogen generates impurity levels originating as a result of the positioning of the N 2p orbital near the valence band, which narrows the energy gap associated with the electron transition.38 In such cases, oxygen defects are inevitably formed so as to maintain charge neutrality, deteriorating the photocatalytic activity. Doping with sulfide ions will also form impurity levels near the valence band and these ions have the same valence as oxide ions. However, it is difficult to dope sulfide ions into TiO2 because they are larger than the oxide ions, and so S-doped TiO2 is structurally unstable.

Doped semiconductor photocatalysts

405

Fig. 3 Diffuse reflectance spectra of TiO2:Sb(X%)/Cr(2.3%) having Sb/Cr ratios of (i) 0, (ii) 0.5, (iii) 1, (iv) 1.5, (v) 2, (vi) 2.5 and (vii) 3.5, together with the spectra of (viii) hydrogen-reduced (i). Reprinted with permission from Kato, H.; Kudo, A. Visible-Light-Response and Photocatalytic Activities of TiO2 and SrTiO3 Photocatalysts Codoped with Antimony and Chromium. J. Phys. Chem. B 2002, 106, 5029–5034. Copyright © 2002 American Chemical Society.

As shown in Fig. 3, Cr-doped TiO2 exhibits an orange to black color depending on the valence of the Cr ions,51 and will contain varying concentrations of Cr3þ, Cr6þ and oxygen vacancies to maintain charge neutrality. Therefore, various types of mid-gap states are formed, resulting in continuous visible light absorption. In prior work, black Cr-doped TiO2 did not exhibit photocatalytic activity because the Cr6þ and oxygen defects in this material promoted non-radiative charge recombination. However, following codoping with Cr and an appropriate amount of Sb, an orange powder with an absorption onset near 570 nm was obtained. This material was found to promote oxygen generation from an aqueous solution containing a sacrificial reagent. The addition of Sb5þ to this material stabilized Cr3þ ions while suppressing the formation of Cr6þ and oxygen defects as a result of charge compensation. Similar effects have been observed in work with TiO2 photocatalysts codoped with Ni and Sb,51 Ni and Ta,52 Ni and Nb,52 and Rh and Sb,53 and so codoping is evidently a very effective means of controlling the dopant valence. An orange Cr-doped TiO2 powder can also be obtained by converting Cr6þ ions to Cr3þ using a hydrogen reduction treatment.51 However, Cr-doped TiO2 does not exhibit photocatalytic activity because it inevitably contains oxygen defects that promote non-radiative deactivation. The ability of TiO2 codoped with Cr and Sb to promote oxygen evolution from an aqueous silver nitrate solution was investigated, employing a fixed Cr level of 2.3 mol% while varying the extent of Sb doping.51 Samples doped with insufficient amounts of Sb exhibited a black color and did not show photocatalytic activity, while the addition of equivalent amounts of Sb and Cr produced an orange material that was active during the oxygen evolution reaction. The incorporation of higher Sb levels further improved the photocatalytic activity up to an eventual plateau. Theoretically, the addition of equimolar amounts of Sb and Cr would be expected to transition all Cr6þ ions in the material to the Cr3þ state, but volatilization during the calcination step slightly reduced the actual amount of Sb in the product. Consequently, an excess of Sb was required to obtain the maximum photocatalytic activity. Even if Sb is added in excess, some will be lost during the subsequent calcination in air, and both Sb3þ and Sb5þ can be present in the finished material. Prior studies have demonstrated that Sb is an important codopant in terms of controlling the valence of the other codopant without generating oxygen defects. Effective charge compensation by codoping has also been observed in work using anions. As discussed above, TiO2 doped with nitrogen absorbs visible light because the filled orbitals derived from the N 2p orbitals form an energy level near the valence band. However, there is a limit to the extent of nitrogen doping because oxygen defects are formed to maintain charge neutrality, and so the absorption of visible light remains weak. Therefore, the codoping of TiO2 with nitrogen and fluorine has been examined as a means of maintaining charge neutrality and increasing the quantity of N3
and F
ions in the product.54 In the case that the O2
ions in the material are fully replaced by pairs of N3
and F
ions, TiNF would presumably be obtained, but this compound has not yet been reported. Thus, N- and F-codoped TiO2 is typically expressed as TiNxOyFz in the literature. Anatase-type TiNxOyFz can be obtained by heating a compound containing N3
and F
such as (NH4)2TiF6 or NH4TiOF3 together with an oxide (as an oxygen source) in an NH3 flow.54,55 N-doped TiO2 prepared by heating anatase-type TiO2 in an NH3 stream was found to only weakly absorb visible light up to about 500 nm, whereas anatase-type TiNxOyFz showed a light absorption onset near 570 nm and strongly absorbed visible light. TiNxOyFz can have a nitrogen concentration more than 10 times higher than that of Ndoped TiO2,54 which is considered to be responsible for the enhanced visible light absorption of this material. Theoretical analyses

406

Doped semiconductor photocatalysts

Fig. 4 The energy band dispersions of TiNxOyFz (Ti64N2O124F2) and TiO2 (Ti32O64). Reprinted with permission from Maeda, K.; Shimodaira, Y.; Lee, B.; et al. Studies on TiNxOyFz as a Visible-Light-Responsive Photocatalyst. J. Phys. Chem. C 2007, 111, 49, 18264–18270. Copyright © 2007 American Chemical Society.

have demonstrated that the visible light absorption of TiNxOyFz results from the transition from the impurity level consisting of N 2p orbitals near the top of the valence band of the host TiO2 to the conduction band consisting of Ti 3d orbitals (Fig. 4). In addition, the impurity level consisting of F 2p orbitals was shown to be isolated at a more positive potential than the valence band, and so could not contribute to light absorption. Interestingly, the effects of co-doping with N3
and F
are essentially the same as are obtained from doping with these ions individually.55 TiNxOyFz was found to generate oxygen from an aqueous silver nitrate solution under visible light, with an AQY of about 1% at 420 nm, but was almost inactive with regard to hydrogen evolution from an aqueous methanol solution. Rutile-type TiNxOyFz can also be prepared by heating a mixture of rutile-type TiO2 and (NH4)2TiF6 in an NH3 stream.56 This material exhibited light absorption beginning at approximately 550–580 nm, and its visible light absorption increased with increasing nitrogen and fluorine concentrations, as shown in Fig. 5. This behavior is similar to that of anatase-type TiNxOyFz. Rutile-type TiNxOyFz has been found to evolve oxygen with an AQY of 6.4% at 420 nm from an aqueous silver nitrate solution. Notably, this material exhibited an AQY equal to or greater than that of undoped TiO2 even in response to ultraviolet light at 365 nm. This result indicated that the introduction of the impurity levels did not disturb the charge transport properties of the material, which is unique among doped photocatalysts. Rutile-type TiNxOyFz was also active during oxygen generation reactions from aqueous NaIO3 and FeCl3 solutions when loaded with a RuO2 cocatalyst that provided reduction and oxidation sites. Both IO3
/I
and Fe3þ/Fe2þ are reversible redox couples that are used as redox mediators in Z-scheme systems, and so allowed rutiletype TiNxOyFz to act as an oxygen evolution photocatalyst (OEP) in a Z-scheme process. Specifically, a Z-scheme system consisting of Rh-doped SrTiO3 as the hydrogen evolution photocatalyst (HEP), rutile-type TiNxOyFz as the OEP and tris(2,20-bipyridyl)cobalt(II) sulfate ([Co(bpy)3]2 þ) as the redox mediator showed water splitting activity under visible light, with a solar-to-hydrogen energy conversion efficiency (STH) of 0.02% under simulated sunlight.56 The effect of co-doping on the dynamics of photoexcited carriers in this material was investigated by transient absorption spectroscopy, and TiO2 doped with only nitrogen was found to generate a strong signal derived from deeply trapped electrons in trap levels derived from oxygen defects. This signal became less intense as the extent of fluorine codoping was increased, because charge compensation by lower valence F
ions suppressed the formation of such defects. At the same time, the concentration of trapped holes capable of participating in water oxidation

Doped semiconductor photocatalysts

407

Fig. 5 Diffuse reflectance spectra of rutile-type TiNxOyFz obtained by nitriding mixtures of rutile-type TiO2 and (NH4)2TiF6 using various (NH4)2TiF6 concentrations (C, in units of mol%) at 773 K for 1 h. Note that the plot labeled with “a” was instead nitrided at 873 K. Reprinted with permission from Miyoshi, A.; Vequizo, J. J. M.; Nishioka, S.; et al. Nitrogen/Fluorine-Codoped Rutile Titania as a Stable Oxygen-Evolution Photocatalyst for SolarDriven Z-Scheme Water Splitting. Sustain. Energy Fuels 2018, 2, 2025–2035. © 2018 The Royal Society of Chemistry.

was increased. However, above an optimal fluorine amount, the concentration of shallowly trapped and free electrons decreased and deep trap levels were generated. As a result, the oxygen evolution activity of rutile-type TiNxOyFz was maximized when nitrogen and fluorine were co-doped at high concentrations, within the range over which the structure of the rutile phase was maintained and no impurities were produced. The resulting TiNxOyFz exhibited sufficient absorption of visible light together with a low density of mid-gap states that can act as deep traps for charge carriers. Charge compensation by codoping is not limited to anion or cation only codoping, but may also be accomplished via the addition of both cations and anions. As an example, the codoping of Ta5þ ions into N-doped rutile-type TiO2 was shown to reduce the concentration of oxygen defects and prolong the lifetime of reactive excited electrons.57 Is should be noted that co-doped TiO2 have been used exclusively as an OEP. TiO2 does not have a sufficiently negative conduction band minimum and is an n-type semiconductor, and so the driving force provided by this material for the hydrogen generation reaction using excited electrons would be insufficient, as shown in Fig. 6B.51 Dopants in this oxide are thought to form a donor band so that holes present at the impurity level have enough mobility to drive the water oxidation reaction (Fig. 6D).

6.18.2.2.2

SrTiO3

SrTiO3 is a ternary oxide and can undergo a variety of elemental substitutions, and so the doping and codoping of this material with cations have been studied extensively. Cr- and Ta-codoped SrTiO3 was reported to exhibit hydrogen evolution activity in response to wavelengths longer than 460 nm.58 In 2001, a Z-scheme system employing SrTiO3 codoped with Cr and Ta (acting as the HEP) was shown to exhibit water splitting under visible light.59 Based on ionic radii values, Cr and Ta are believed to substituted at Ti sites in this oxide. Prior work has indicated that doping with Cr alone produces Cr3þ, which provides visible light activity, together with Cr6þ and oxygen defects that promote charge recombination, resulting in broad visible light absorption peaking at around 520 and tailing beyond 700 nm (Fig. 7).60 However, after codoping with Ta, the lack of positive charges due to the Cr3þ in the material was compensated for by the presence of Ta5þ rather than the formation of Cr6þ ions. As a result, the doped Cr was stabilized in the Cr3þ state and the formation of Cr6þ and oxygen defects was suppressed. SrTiO3 co-doped with Cr and Ta showed a light absorption onset at 530 nm together with weak absorption derived from the d-d transition of Cr ions over the range of 570–700 nm. Cr6þ species can also be removed by hydrogen reduction, but this process tends to produce oxygen defects. Cr6þ ions can also be gradually regenerated on the basis of oxidation in ambient air. The highest degree of hydrogen evolution from an aqueous methanol solution was obtained from codoped SrTiO3 that had also been reduced under hydrogen, reflecting the effective suppression of Cr6þ and oxygen defects. The modification of dopant valences by codoping was also observed in the case of SrTiO3 codoped with Ni and Ta.52 Adding Ta5þ to Ni-doped SrTiO3 was found to stabilize Ni2þ and suppress the formation of Ni3þ, likely leading to the trapping of electrons. Analyses by transient absorption spectroscopy indicated that the lifetime of excited electrons was shortened when only Ni or Ta was incorporated, but was extended when Ni and Ta were codoped.61 However, it was also noted that the photocatalytic activity of the material remained low because the excited electrons were not able to reduce oxygen efficiently. It was reported in 2002 that SrTiO3 codoped with Cr and Sb exhibited hydrogen production activity in the presence of sacrificial electron donors, but almost no activity with regard to the oxygen production reaction.51 Cr3þ ions doped into SrTiO3 form discrete donor levels in the bandgap of the host oxide and readily capture holes generated in the valence band of the oxide, such that a sharp

408

Doped semiconductor photocatalysts

TiO2:Sb/Cr After band gap excitation

SrTiO3:Sb/Cr

After visible light band excitation

After band gap excitation

After visible light band excitation

Photoluminescence property CB

e- e-

e-

Luminescence from Cr3+

Non radiative transition

Continuous DB Cr3+(+O2-)

VB

CB

e-

h+ h+ (a)

Luminescence from Cr3+ Discrete DL (Cr3+)

h+

VB

(c)

e-

h+

h+

h+ (e)

(g)

Photocatalytic property Ag H2

Potential / eV vs. NHE

0

e-

Ag H2

H2

e-

e-

H+ Ag+ H+

H+/H2

H+

Ag+ H+

1

O2/H2O CH3OH H2O

H 2O

2

h+ 3

H2

e-

O2

h+

Ox

H 2O

h+

Ox

O2

H 2O

h+ (b)

O2

CH3OH H2O

h+ O2

(d)

Favorable path

(f)

O2

(h)

Unfavorable path

Fig. 6 Diagrams summarizing the photoluminescence and photocatalytic mechanisms on TiO2:Sb/Cr and SrTiO3:Sb/Cr. Adapted with permission from Kato, H.; Kudo, A. Visible-Light-Response and Photocatalytic Activities of TiO2 and SrTiO3 Photocatalysts Codoped with Antimony and Chromium. J. Phys. Chem. B 2002, 106, 5029–5034. Copyright © 2002 American Chemical Society.

fluorescence emission peak is generated by this photocatalyst (Fig. 6E and G). Therefore, during excitation by either ultraviolet or visible light, holes are able to relax to the impurity level and so show lower mobility. As such, it is difficult to drive a multi-electron reaction such as water oxidation (Fig. 6H and I), although it should be noted that these holes can still drive simple reactions such as the oxidation of methanol. In addition, the potential of the bottom of the conduction band in the photocatalyst is sufficiently more negative than the hydrogen generation potential. Therefore, Cr- and Sb-codoped SrTiO3 will promote the hydrogen evolution reaction under visible light. The characteristics of this material are in contrast to those of Cr- and Sb-codoped TiO2, likely due to the difference in the local crystal structures of the host photocatalyst materials.51 In 2004, the photocatalytic activity of SrTiO3 doped with precious metal cations such as Ru, Rh and Ir was reported (Fig. 8).62 Rh-doped SrTiO3 is an especially promising photocatalyst, as it has p-type semiconductor characteristics and has played a major role in the development of the Z-scheme-type water splitting reaction systems. Rh-doped SrTiO3 can use visible light up to a wavelength of approximately 540 nm for photocatalytic reactions.62 This photocatalyst exhibits a suitable level of activity under visible light during the hydrogen production reaction from an aqueous solution containing a sacrificial reagent when loaded with precious metal cocatalysts. Notably, Rh-doped SrTiO3 also promotes hydrogen evolution from an aqueous solution containing Fe2þ, which serves as a reversible redox mediator via the Fe3þ/Fe2þ pair.63,64 The hydrogen evolution reaction and the reduction reaction of Fe3þ

Doped semiconductor photocatalysts

409

Fig. 7 Diffuse reflectance spectra obtained from doped SrTiO3 with and without a H2 reduction step, containing (a) 4 mol% Cr, (b) 4 mol% Cr (reduced), (c) 4 mol% Cr and 4 mol% Ta, and (d) 4 mol% Cr and 4 mol% Ta (reduced). Reprinted with permission from Ishii, T.; Kato, H.; Kudo, A. H2 Evolution from an Aqueous Methanol Solution on SrTiO3 Photocatalysts Codoped with Chromium and Tantalum Ions Under Visible Light Irradiation. J. Photochem. Photobiol. A Chem. 2004, 163, 181–186. © 2004 Elsevier B.V.

Fig. 8 Diffuse reflectance spectra of SrTiO3:M(0.5%) in which M is (a) Mn, (b) Ru, (c) Rh, (d) Pd, (e) Ir, and (f) Pt. The dashed line indicates the spectrum of undoped SrTiO3. Reprinted with permission from Konta, R.; Ishii, T.; Kato, H.; Kudo, A. Photocatalytic Activities of Noble Metal Ion Doped SrTiO3 under Visible Light Irradiation. J. Phys. Chem. B 2004, 108, 8992–8995. Copyright © 2004 American Chemical Society.

will compete on a Pt cocatalyst loaded on Rh-doped SrTiO3. However, because Fe2þ species are adsorbed on the Pt and subsequently inhibit the uptake of Fe3þ, the hydrogen production reaction proceeds with significant selectivity even in the presence of Fe3þ.64 Rh-doped SrTiO3 can be combined with various OEPs (including WO3, Bi2MoO6, BiVO4 and others) to drive Z-scheme water splitting reactions, employing a reversible redox pair to transfer electrons from the OEP to the HEP.65 A solid-state junction can also be used instead of a reversible redox couple because Rh-doped SrTiO3 is a p-type semiconductor.66,67 The Ru functions as an effective cocatalyst to promote Z-scheme water splitting.65 It has been shown that Fe2þ ions are not as readily oxidized on a Ru cocatalyst as on Pt in the presence of oxygen, and this low activity for the reverse reaction is thought to be the origin of the high Z-scheme water splitting activity observed when using a Ru-loaded HEP. Ir-doped SrTiO3 is active during hydrogen production from an aqueous methanol solution under visible light up to nearly 800 nm.68 It has been reported that the hydrogen evolution activity of this material can be improved by loading metallic Ir nanoparticles having intimate contact with the photocatalyst as a cocatalyst, via impregnation and hydrogen reduction. The resulting AQY obtained for the hydrogen evolution reaction was 3% at a wavelength of 420 nm. Ru-doped SrTiO3 can also produce hydrogen from aqueous methanol with the aid of a Pt cocatalyst, and generate oxygen from an aqueous silver nitrate solution under visible

410

Doped semiconductor photocatalysts

light up to nearly 750 nm.69 The oxygen evolution activity of this photocatalyst can be enhanced by converting the Ru4þ ions to Ru3þ via hydrogen reduction, with a resulting AQY of 0.1% at 420 nm. Ru4þ can also be converted to Ru3þ by codoping an appropriate amount of Sb5þ, although codoping Nb5þ or Ta5þ into Ru-doped SrTiO3 does not produce a charge compensation effect. The Ru3þ and Nb5þ or Ta5þ ions must be in close proximity on the oxide for charge compensation to occur, but this is unlikely to be the case after incorporation into SrTiO3 because these ions are all larger than Ti4þ. In contrast, the ionic radius of Sb5þ is almost equal to that of Ti4þ, and so Sb5þ ions can be situated near Ru3þ ions to effectively stabilize this Rh valence state. The valence state of Rh doped into SrTiO3 has been investigated in detail because of the unique semiconductor properties and photocatalytic activity of Rh-doped SrTiO3.62 The Rh species are evidently substituted at Ti4þ sites as either Rh3þ or Rh4þ, and diffuse reflectance spectroscopy of the resulting materials has shown variations in light absorption properties. As shown in Fig. 9, the presence of Rh3þ leads to an absorption onset wavelength of approximately 420 nm (due to electronic transitions to the conduction band), while Rh4þ results in an absorption peak around 580 nm (due to electronic transitions from the valence band).62,70 Weak light absorption derived from the d-d transition of Rh is additionally observed near 1000 nm. X-ray absorption spectroscopy, X-ray emission spectroscopy and theoretical calculations have also demonstrated that Rh4þ ions form a donor level near the top of the valence band, whose main component is the O 2p orbitals, and an acceptor level near the middle of the bandgap of the host SrTiO3, and that Rh3þ forms a donor level near the top of the valence band. In addition, the donor level of Rh3þ is located at a slightly more negative potential than that of Rh4þ. These impurity levels are composed of Rh 4d t2g orbitals. Notably, Rh4þ forms a mid-gap level due to states destabilized by an unpaired electron, but Rh3þ does not have an unpaired electron and so orbital destabilization leading to the formation of a mid-gap state does not occur. The impurity levels derived from Rh4þ are largely eliminated during the hydrogen evolution reaction under light, such that only those resulting from Rh3þ ions remain.62 Therefore, it is believed that Rh3þ contributes to visible light activity. Transient absorption spectroscopy has also confirmed that Rh4þ is primarily responsible for promoting recombination,71 and theoretical analyses have established that SrTiO3 is converted to a p-type semiconductor by doping with Rh.70 Finally, earlier work has indicated that Rh-doped SrTiO3 produces a photocathode current when processed into a photoelectrode.72 Attempts have been made to stabilize Rh3þ ions in Rh-doped SrTiO3 by codoping with another cation having higher valence, such as Sb5þ,73 and analyses by Raman spectroscopy have indicated that the Sb ions are substituted at Ti sites. These ions are present as Sb5þ, such that the Rh3þ state becomes more stable than the Rh4þ state and the concentration of the latter ions is decreased. Additional assessments using electron spin resonance spectroscopy have indicated that a small amount of Rh4þ remains when equimolar amounts of Sb and Cr are added, but disappears when the amount of Sb added is 1.5 times that of Cr. Simultaneously, the broad absorption peak around 580 nm derived from Rh4þ is weakened. The study also suggested that the Rh3þ species consisted of unstable Rh3þ that was reversibly converted to Rh4þ, together with Rh3þ stabilized by the hydrogen reduction treatment or codoping with Sb5þ, each forming a donor level near the valence band. The unstable Rh3þ species had stronger electron-donating properties and formed donor levels at a slightly more negative potential than the stabilized Rh3þ species. The photocatalytic activity of Rh- and Sb-codoped SrTiO3 was also studied,73 and the results demonstrated that the hydrogen evolution activity of Rh-doped SrTiO3 was reduced following the addition of Sb. This likely occurred because the conduction band composed of Ti 3d orbitals was disturbed by the presence of Sb. In addition, Rh4þ ions promoting recombination, which would be expected to exist in the absence of Sb, can be reduced to Rh3þ during the hydrogen evolution reaction. Conversely, during the oxygen evolution reaction, Rh-doped SrTiO3 did not show photocatalytic activity, presumably because electrons excited to the Rh4þ acceptor level did not have sufficient potential to reduce Ag. In addition, Rh4þ ions were unlikely to be reduced to Rh3þ during

(A)

(B)

Rh3+ Rh4+

CB B A VB

400

600 800 1000 Wavelength (nm)

1200

Rh4+:SrTiO3

Energy

Energy

Absorbance (a.u.)

Eg=3.2eV CB

C VB

Rh3+:SrTiO3

Fig. 9 (A) Ultraviolet–visible–near infrared absorption spectra obtained from Rh4þ:SrTiO3 and Rh3þ:SrTiO3 with 1 atom% Rh doping and (B) diagrams of the estimated electronic structures of these materials. In (B), states A and C are Rh donor levels while state B is an acceptor level. Reprinted with permission from Kawasaki, S.; Akagi, K.; Nakasugi, K.; et al. Elucidation of Rh-Induced In-Gap States of Rh:SrTiO3 Visible-LightDriven Photocatalyst by Soft X-ray Spectroscopy and First-Principles Calculations. J. Phys. Chem. C 2012, 116, 24445–24448. © 2012 American Chemical Society.

Doped semiconductor photocatalysts

CB

CB

e−

411

e− Ag+/Ag

Rh4+ O2/H2O Rh4+ stabilized by oxygen defects

h+

VB (a) SrTiO3:Rh CB

h+

Poor ability for O2 evolution

Ag+ e−

IrOx

VB

Effective O2 evolution site

CB

Ag+ e−

Ag

Ag Ag+/Ag

4+

No Rh

Positive factor

O2/H2O

H 2O Rh3+ stabilized by antimoney codoping

Negative factor

(b) IrOx/SrTiO3:Rh

H2O

h+

h+ O2

O2 IrOx

VB (c) SrTiO3:Rh/Sb

Poor ability for O2 evolution

VB

Effective O2 evolution site

(d) IrOx/SrTiO3:Rh/Sb

Fig. 10 Diagrams showing the proposed mechanisms for photocatalytic O2 evolution over various SrTiO3-based photocatalysts under visible light. Reprinted with permission from Niishiro, R.; Tanaka, S.; Kudo, A. Hydrothermal-Synthesized SrTiO3 Photocatalyst Codoped with Rhodium and Antimony with Visible-Light Response for Sacrificial H2 and O2 Evolution and Application to Overall Water Splitting. Appl. Catal. B Environ. 2014, 150– 151, 187–196. © 2013 Elsevier B.V. All rights reserved.

the oxygen reduction reaction (Fig. 10A and B). Accordingly, it is necessary to intentionally convert Rh4þ to Rh3þ in order to activate Rh-doped SrTiO3 for this reaction, and codoping with Sb is more effective than hydrogen reduction because the former does not generate oxygen defects. Even so, multi-electron oxidation with this photocatalyst remains difficult (Fig. 10C), and so it is essential to modify the photocatalyst with IrO2 as an oxygen evolution cocatalyst (Fig. 10D). In addition, excessive codoping of Sb should be avoided, because Sb3þ has a larger ionic radius than Ti4þ and so hinders the electron conduction in the conduction band. The AQY values for Rh- and Sb-codoped SrTiO3 with regard to the hydrogen and oxygen evolution reactions were reported to be 0.8% and 4.5% at 420 nm, respectively. A lack of codoping or codoping with Sb activates Rh-doped SrTiO3 for the hydrogen or oxygen evolution reaction, respectively. Therefore, Z-scheme-type water splitting is possible using SrTiO3 doped with Rh and SrTiO3 codoped with Rh and Sb as the HEP and OEP, respectively. In 2014, one-step excitation overall water splitting was achieved under visible light using Rh- and Sb-codoped SrTiO3 modified with IrO2, RuO2 or Ru as cocatalysts. A sample loaded with IrO2 decomposed water most efficiently, with an STH value of 0.01%.74 IrO2 is well known as an oxygen evolution cocatalyst, but can also promote hydrogen evolution reactions, most likely because partially reduced IrO2 provides reduction sites. Moreover, unlike metallic Ir, IrO2 is not as efficient at promoting the reverse reaction that produces water from hydrogen and oxygen. In the case that the dopant concentration is low, the density of states derived from impurity levels will also be low. Therefore, it is difficult for photoexcited charge carriers generated in the impurity level to move over long distances. In order to enhance visible light absorption and increase the mobility of these carriers, it is preferable to add the dopant at a high concentration so as to form a continuous impurity band. However, heavy doping also induces defects and distortion of the crystal structure. It may be possible to mitigate these issues using a core/shell-type spatial distribution in terms of the dopant concentrations, such that a shell in which impurities are heavily doped is generated over a core of highly crystalline SrTiO3. This core may function as a template to suppress degradation of the crystallinity of the doped layer. At the same time, the region in which visible light is absorbed is limited to the thin doped shell, so that the distance that the photoexcited charge carriers must travel can be shortened. The synthesis of La- and Rh-codoped SrTiO3 having such a core/shell-type dopant distribution was reported in 2014,75 via the heating of SrTiO3, La2O3 and Rh2O3 for an optimal period of time. During the heating process, La3þ and Rh3þ ions diffused from the SrTiO3 surface and substituted for Sr2þ and Ti4þ ions, respectively. The appropriate heating durations produced Laand Rh-codoped SrTiO3 with high La and Rh concentrations near the surfaces of the SrTiO3 particles, as shown in Fig. 11. However, when the heating was extended, the dopants were also observed in the centers of the particles, likely as a result of

412

Doped semiconductor photocatalysts

Fig. 11 Scanning transmission electron microscopy images and energy-dispersive X-ray spectroscopy maps of La- and Rh-codoped SrTiO3 particles prepared using a two-step solid-state reaction followed by calcination for (A) 6 and (B) 10 h. Reprinted with permission from Wang, Q.; Hisatomi, T.; Ma, S. S. K.; Li, Y.; Domen, K. Core/Shell Structured La- and Rh-Codoped SrTiO3 as a Hydrogen Evolution Photocatalyst in Z-Scheme Overall Water Splitting under Visible Light Irradiation. Chem. Mater. 2014, 26, 4144–4150. Copyright © 2014 American Chemical Society.

thermally-promoted diffusion. This tendency was also observed by elemental analysis in the depth direction of the particles using X-ray photoelectron spectroscopy. These data confirm that core/shell-type particles could be obtained throughout the entire sample simply by controlling the heating duration. The presence of a highly crystalline SrTiO3 core was thought to maintain the crystallinity of the doped shell while accommodating the dopant at high concentrations, because the formation of impurities was suppressed compared with the case where SrCO3, TiO2, La2O3 and Rh2O3 were directly heated. La- and Rh-codoped SrTiO3 showed activity in a Z-scheme water splitting system when combined with Ta3N5, where the former functioned as the HEP and the latter as the OEP. Moreover, efficient Z-scheme water splitting with STH values exceeding 1.0% was realized using La- and Rh-codoped SrTiO3 as the HEP and Mo-doped BiVO4 as the OEP in a particulate photocatalyst sheet configuration.76 The electronic structures of Rh-doped SrTiO3 and La- and Rh-codoped SrTiO3 have been analyzed in detail by spectroelectrochemistry, transient absorption spectroscopy, computational chemistry and photoelectron spectroscopy.77 It was repeatedly reported that, in the case of Rh-doped SrTiO3, Rh4þ species formed energy levels in the bandgap and trapped charge carriers. However, when the potential was shifted to a sufficiently negative potential, Rh4þ was reduced to Rh3þ and the trap level

Doped semiconductor photocatalysts

413

disappeared dynamically. Under these conditions, the Rh-doped SrTiO3 could function as a photocatalyst by efficiently using excited electrons for hydrogen generation. Rh4þ species were found not to exist in La-codoped samples even when a negative potential was not applied, and La- and Rh-codoped SrTiO3 functioned effectively as a photocatalyst for hydrogen generation under various reaction conditions. This result indicated that La codoping had the same effect as the application of a negative potential with regard to controlling the electronic structure of Rh-doped SrTiO3. This is an important aspect of adjusting the electronic structure of particulate photocatalysts, the potentials of which cannot be controlled externally, and will be applicable to the future development of particulate photocatalysts with effectively controlled electronic structures.

6.18.2.3 6.18.2.3.1

Modification of the particle morphology NaTaO3

Sodium tantalate (NaTaO3) is a perovskite-type ternary oxide with a band gap energy of 4.0 eV.78 This material was reported in 1998 to be active during water splitting reactions under ultraviolet radiation.79 Since that time, studies have shown that the water splitting activity of this oxide can be dramatically improved by suppressing Na defects via the addition of an excess of Na and by loading a NiO cocatalyst.80 Furthermore, it was demonstrated that doping with various lanthanoids and alkaline earth metals changed the particle morphology and improved the water splitting activity.81,82 As an example, La doping decreased the particle size of NaTaO3 from the micrometer scale to the submicrometer scale, and NaTaO3 doped with La exhibited a high AQY value of 56% during overall water splitting at 270 nm.83 La3þ ions have an ionic radius of 150 pm in the 12-coordinated state, which is closer to the radius of Naþ (153 pm in the 12-coordinated state) than that of Ta5þ (78 pm in the six-coordinated state), and thus are considered to preferentially replace Naþ in NaTaO3. Evidence for the substitution of La3þ for Naþ ions was provided by elemental analyses using energy-dispersive X-ray spectroscopy and X-ray photoelectron spectroscopy as well as structural analyses with La-K edge X-ray absorption fine structure and Raman spectroscopy. Na defects are believed to be simultaneously generated to maintain charge neutrality in the La-doped NaTaO3 material. In such materials, the majority of the La3þ species were found near the NaTaO3 surface, while the remainder were situated in the bulk of the material,83 such that a La-rich layer about 2 nm in thickness could be observed on the surface.84 The lattice mismatch between the La-rich shell and the La-lean core was relaxed by the formation of a characteristic nanostep surface structure (Fig. 12).83 Notably, NiO cocatalyst particles could be selectively loaded on the edge parts of the nanosteps on La-doped NaTaO3, and hydrogen evolution (that is, reduction) and oxygen evolution (oxidation) proceeded on the edges and the grooves of the nanosteps, respectively, such that they were spatially separated. This effect is thought to have suppressed charge recombination on the photocatalyst surface and thereby enabled highly efficient photocatalytic water splitting over the NiO-loaded La-doped NaTaO3 photocatalyst.83 As noted, the spatial distribution of the La3þ species could be altered by varying the heating duration.84 Extending the heating step distributed La3þ ions evenly throughout the particles, while simultaneously causing the nanostep structure to disappear. The water splitting activity of this material was improved by La doping, but this improvement effect was diminished with longer heating durations, presumably because of the loss of surface nanosteps. The effect of La3þ ion doping on the dynamics of photoexcited carriers in NaTaO3 under ultraviolet light was also investigated using infrared absorption spectroscopy.84 When exposed to ultraviolet radiation, the absorbance of the NaTaO3 at near-infrared wavelengths was found to increase, and this absorbance increased monotonically with the wavelength. This effect can be attributed to light absorption by electrons in the conduction band resulting from ultraviolet excitation, meaning that this technique can be used to estimate the relative concentration of excited electrons under steady light irradiation. It was determined that the steadystate concentration of photoexcited electrons was influenced by the distribution of La3þ species in the photocatalyst and was also positively correlated with the photocatalytic water splitting activity. The effect of the distribution of La3þ ions on the photocatalytic activity was found to be associated with the extent of band bending and charge separation (Fig. 13). The incorporation

Fig. 12 A scanning electron microscopy image of NiO-loaded NaTaO3 doped with La and a diagram showing the particle structure and surface processes. Reprinted with permission from Kato, H.; Asakura, K.; Kudo, A. Highly Efficient Water Splitting into H2 and O2 over Lanthanum-Doped NaTaO3 Photocatalysts with High Crystallinity and Surface Nanostructure. J. Am. Chem. Soc. 2003, 125, 3082–3089. Copyright © 2003 American Chemical Society.

414

Doped semiconductor photocatalysts

CB

energy gradient e

excitation

Postential / eV V vs NHE at pH 0

−1.1

−

driven away

slo

hv (4.0 eV)

e−

w

re

ce

om

bi

na

tio

n

h+

2.9

VB

high

low BULK

La concentration

SURFACE

Fig. 13 An energy band diagram showing the processes occurring on La-doped NaTaO3 in response to ultraviolet irradiation. Reprinted with permission from Sudrajat, H.; Kitta, M.; Ito, R.; et al. Water-Splitting Activity of La-Doped NaTaO3 Photocatalysts Sensitive to Spatial Distribution of Dopants. J. Phys. Chem. C 2020, 124, 15285–15294. © 2020 American Chemical Society.

of La in the NaTaO3 distorted the Ta-O-Ta chains in the crystal lattice and weakened the interaction between Ta 5d orbitals. As a consequence, the conduction band consisting of Ta 5d orbitals became thinner and the conduction band minimum was shifted negatively. Following a shorter heating period, a concentration gradient of La3þ species was observed and the surface was enriched with La3þ. Along with this, an energy gradient was also formed in the conduction band minimum, inducing upward band bending. This band bending was thought to transfer electrons to the bulk of the material and so facilitate charge separation. However, because the height of the energy barrier was small enough for electrons to overcome at room temperature, the electrons in the bulk of the material could still reach the surface. It was also possible for shallow trap levels to form near the conduction band minimum in conjunction with La doping, but electrons could escape from these shallow traps and thus remain reactive. In the case that sufficient heating was applied to La-doped NaTaO3, the La3þ concentration gradient was found to be not as pronounced but still present. Therefore, charge separation was promoted and water splitting activity was improved compared to undoped NaTaO3 without an energy gradient in the conduction band minimum. Data acquired using X-ray diffraction and Raman spectroscopy indicated that the sites occupied by Sr cations in Sr-doped NaTaO3 changed depending on the synthesis method and the extent of doping.85 As shown in Fig. 14, the use of a smaller quantity of Sr in the solid state reaction caused the Sr to localize near the surface and to substitute at Naþ sites. However, a large amount of Sr resulted in the Sr2þ ions being situated at both Ta5þ and Naþ sites to form core/shell-structured NaTaO3–SrSr1/3Ta2/3O3 solid solutions in which a Sr-rich shell covered a Sr-poor core. Employing a hydrothermal synthesis method caused the Sr to occupy only A sites regardless of the doping amount. The effect of Sr2þ ion doping on the dynamics of photoexcited carriers in NaTaO3 under ultraviolet light was also investigated, using infrared absorption spectroscopy, which provided information concerning the steady state concentration of photoexcited electrons. The results established that Sr doping above a specific level increased the absorbance of the oxide in the near-infrared region of the spectrum by as much as two orders of magnitude. Excessive doping was also found to be deleterious, since mid-gap states capable of trapping conduction band electrons were formed and minimal increases in near infrared absorbance were observed. Different results were obtained from low level doping in conjunction with the solid state synthesis of Sr-doped NaTaO3 or when the material was prepared using a hydrothermal technique. In these scenarios, the

Fig. 14 Diagrams showing the (A) cross-sectional composition of NaTaO3 particles with various Sr concentrations synthesized via a solid-state reaction and (B) the mismatch of the perovskite-structure lattices at the heteroepitaxial core–shell interface, generating regularly separated steps on the surface. White and gray squares represent unit cells in the core and shell, respectively. Reprinted with permission from An, L.; Onishi, H. Electron–Hole Recombination Controlled by Metal Doping Sites in NaTaO3 Photocatalysts. ACS Catal. 2015, 5, 3196–3206. © 2015 American Chemical Society. The direct link to the original article is https://pubs.acs.org/doi/abs/10.1021/acscatal.5b00484. Further permissions related to the material excerpted should be directed to the American Chemical Society.

Doped semiconductor photocatalysts

415

absorbance in the near infrared region was not increased, suggesting that the steady-state electron concentration was elevated only when Sr was doped into Ta5þ sites. During the synthesis of Sr-doped NaTaO3 using a flux method, it is possible to control the distribution of Sr by varying the duration of heating without changing the total amount of Sr that is added.86 Specifically, a heating duration as short as 1 h was found to form a Sr-rich layer near the oxide surface, while a 60 h heating step caused the Sr to be evenly distributed throughout the NaTaO3. The degree of increase in near infrared light absorption under steady ultraviolet irradiation was evidently sensitive to the duration of heating applied to the flux, and thus to the spatial distribution of Sr. The most significant enhancement as compared with undoped NaTaO3 was obtained at a relatively short heating time of 1 h. This outcome suggests that the electron recombination was slower when the Sr-rich layer was localized on the surface and a concentration gradient was formed with respect to the bulk. Depending on the concentration of Sr, the bottom of the conduction band could be shifted to the negative potential side by a maximum of 0.1 eV, such that a potential gradient was induced by the concentration gradient. This effect appears to have promoted the separation of electrons and holes, although the water splitting activity was not necessarily correlated with the concentration of excited electrons. Evidently, it was difficult for electrons to reach the surface even with a high electron concentration when the energy gradient of the conduction band was too large.

6.18.2.3.2

SrTiO3

As noted, Al-doped SrTiO3 showed improved water splitting activity in conjunction with a reduced Ti3þ concentration.14 Al can be effectively doped into this oxide using a SrCl2 flux17 and modifies both the physical characteristics of the semiconductor and the particle morphology. Fig. 15 presents SEM images of the pristine SrTiO3 powder and of specimens heated together with Al2O3 as an Al source and/or SrCl2 as the flux in an yttria crucible.17 The sample subjected to a flux treatment showed well-grown cubic particles, while that heated with Al2O3 had slightly larger particle sizes than the pristine material. When both the flux and Al2O3 were employed, the particle growth was suppressed. In addition, the use of an alumina crucible instead of an yttria crucible produced even larger SrTiO3 particles, likely because Y contamination was avoided. These results demonstrated that the inclusion of impurities can prevent SrTiO3 particles from growing into large crystals. The water splitting activity of these SrTiO3 samples appeared to have no correlation with the particle morphology. The pristine SrTiO3 and samples heated with both Al2O3 and SrCl2 in an yttria crucible showed similar particle sizes, although the activities of these materials differed by more than two orders of magnitude (Table 1). In addition, well grown SrTiO3 particles heated with SrCl2 in an yttria crucible did not show high water splitting activity while those heated in an alumina crucible did. Thus, higher activity was clearly associated with Al doping, and so suppressing the reduction of Ti4þ species is evidently more important than controlling the particle size. Nevertheless, the exposure of well-defined facets can be essential for efficient charge separation based on anisotropic charge transport.21,87,88 The Al-doped SrTiO3 particles obtained by the flux method were determined to have a truncated cuboidal geometry, with exposed flat large {100} facets and cut out {110} facets.21 This morphology is shown in the SEM image in Fig. 16, which also indicates that nanoparticles were loaded on the particle surfaces. These nanoparticles comprised hydrogen evolution cocatalysts (Cr2O3/Rh) and an oxygen evolution cocatalyst (CoOOH) site-selectively loaded on the {100} facets by photoreduction and on the {110} facets by photooxidation, respectively. The {100} and {110} facets of SrTiO3 are thought to collect photoexcited electrons and holes, respectively, which allows site-selective cocatalyst loading by photodeposition. As a result

Fig. 15 Scanning electron microscopy images of (A) SrTiO3, (B) SrTiO3 heated with SrCl2, (C) SrTiO3 heated with Al2O3, (D) SrTiO3 heated with Al2O3 and SrCl2 in an yttria crucible, and (E) SrTiO3 heated with SrCl2 in an alumina crucible. Heating temperature: 1373 K.

416

Doped semiconductor photocatalysts

Fig. 16 (A) A diagram and (B) scanning electron microscopy image of Al-doped SrTiO3 site-selectively coloaded with a hydrogen evolution cocatalyst (Cr2O3/Rh) and an oxygen evolution cocatalyst (CoOOH).

of this effect, charge separation is promoted and spatially separated electrons and holes can be used efficiently, leading to overall water splitting into hydrogen and oxygen virtually without recombination loss.

6.18.3

Concluding remarks and prospects

This chapter examined the effects of doping on the physical characteristics, visible light absorption properties, particle morphology and water splitting activity of semiconductor photocatalysts. Research to date has shown that doping has a wide range of both positive and negative effects. In addition, the water splitting reaction itself is complicated, and involves many processes occurring in parallel over wide spatial and temporal ranges, and so it is difficult to generalize the effects of doping. However, due in part to the effects of doping, it has become possible to prepare oxide photocatalysts that can drive the water splitting reaction with essentially no recombination losses.21 In addition, more recent work has provided performance improvements based on the nonuniform distribution of dopants.75,83–86 It has also been reported that the valence of the dopant, and thus the electronic structure of the doped photocatalyst, changes dynamically during photocatalytic reactions, which can have a significant effect.62,77 Thus, understanding the spatial distributions and the working states of dopants has becoming increasingly important to the design of doped photocatalysts and the elucidation of the associated reaction mechanisms. This is not unexpected, because particulate photocatalysts are known to exhibit compositional and size variations within and among particles, meaning that they are essentially inhomogeneous. In addition, photocatalysis proceeds in conjunction with non-equilibrium steady states involving photoexcited electrons and holes. Considering these factors, the effective ongoing development of doped photocatalysts will require the support of analytical methods that allow in situ observations with suitable spatiotemporal resolution.3,89

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Fujishima, A.; Zhang, X.; Tryk, D. A. TiO2 Photocatalysis and Related Surface Phenomena. Surf. Sci. Rep. 2008, 63, 515–582. Chong, M. N.; Jin, B.; Chow, C. W. K.; Saint, C. Recent Developments in Photocatalytic Water Treatment Technology: A Review. Water Res. 2010, 44, 2997–3027. Hisatomi, T.; Domen, K. Reaction Systems for Solar Hydrogen Production Via Water Splitting with Particulate Semiconductor Photocatalysts. Nat. Catal. 2019, 2, 387–399. Fabian, D. M.; Hu, S.; Singh, N.; et al. Particle Suspension Reactors and Materials for Solar-Driven Water Splitting. Energ. Environ. Sci. 2015, 8, 2825–2850. Setoyama, T.; Takewaki, T.; Domen, K.; Tatsumi, T. The Challenges of Solar Hydrogen in Chemical Industry: How to Provide, and How to Apply? Faraday Discuss. 2017, 198, 509–527. Kim, J. H.; Hansora, D.; Sharma, P.; Jang, J.-W.; Lee, J. S. Toward Practical Solar Hydrogen ProductiondAn Artificial Photosynthetic Leaf-to-Farm Challenge. Chem. Soc. Rev. 2019, 48, 1908–1971. Wang, Q.; Domen, K. Particulate Photocatalysts for Light-Driven Water Splitting: Mechanisms, Challenges, and Design Strategies. Chem. Rev. 2020, 120, 919–985. Hisatomi, T.; Takanabe, K.; Domen, K. Photocatalytic Water-Splitting Reaction from Catalytic and Kinetic Perspectives. Catal. Lett. 2015, 145, 95–108. Maeda, K. Z-Scheme Water Splitting Using Two Different Semiconductor Photocatalysts. ACS Catal. 2013, 3, 1486–1503. Ng, B.-J.; Putri, L. K.; Kong, X. Y.; et al. Z-Scheme Photocatalytic Systems for Solar Water Splitting. Adv. Sci. 2020, 7, 1903171. Kudo, A.; Niishiro, R.; Iwase, A.; Kato, H. Effects of Doping of Metal Cations on Morphology, Activity, and Visible Light Response of Photocatalysts. Chem. Phys. 2007, 339, 104–110. Chen, X.; Shen, S.; Guo, L.; Mao, S. S. Semiconductor-Based Photocatalytic Hydrogen Generation. Chem. Rev. 2010, 110 (11), 6503–6570. Osterloh, F. E. Inorganic Nanostructures for Photoelectrochemical and Photocatalytic Water Splitting. Chem. Soc. Rev. 2013, 42, 2294–2320. Takata, T.; Domen, K. Defect Engineering of Photocatalysts by Doping of Aliovalent Metal Cations for Efficient Water Splitting. J. Phys. Chem. C 2009, 113, 19386–19388. Zhao, Z.; Goncalves, R. V.; Barman, S. K.; et al. Electronic Structure Basis for Enhanced Overall Water Splitting Photocatalysis with Aluminum Doped SrTiO3 in Natural Sunlight. Energ. Environ. Sci. 2019, 12, 1385–1395. Sakata, Y.; Miyoshi, Y.; Maeda, T.; et al. Photocatalytic Property of Metal Ion Added SrTiO3 to Overall H2O Splitting. Appl. Catal. A. Gen. 2016, 521, 227–232. Ham, Y.; Hisatomi, T.; Goto, Y.; et al. Flux-Mediated Doping of SrTiO3 Photocatalysts for Efficient Overall Water Splitting. J. Mater. Chem. A 2016, 4, 3027–3033. Goto, Y.; Hisatomi, T.; Wang, Q.; et al. A Particulate Photocatalyst Water-Splitting Panel for Large-Scale Solar Hydrogen Generation. Joule 2018, 2, 509–521. Jiang, J.; Kato, K.; Fujimori, H.; Yamakata, A.; Sakata, Y. Investigation on the Highly Active SrTiO3 Photocatalyst Toward Overall H2O Splitting by Doping Na Ion. J. Catal. 2020, 390, 81–89. Yamakata, A.; Ham, Y.; Kawaguchi, M.; et al. Morphology-Sensitive Trapping States of Photogenerated Charge Carriers on SrTiO3 Particles Studied by Time-Resolved Visible to Mid-IR Absorption Spectroscopy: The Effects of Molten Salt Flux Treatments. J. Photochem. Photobiol. A Chem. 2016, 313, 168–175.

Doped semiconductor photocatalysts

417

21. Takata, T.; Jiang, J.; Sakata, Y.; et al. Photocatalytic Water Splitting with a Quantum Efficiency of Almost Unity. Nature 2020, 581, 411–414. 22. Sakata, Y.; Matsuda, Y.; Nakagawa, T.; et al. Remarkable Improvement of the Photocatalytic Activity of Ga2O3 Towards the Overall Splitting of H2O. ChemSusChem 2011, 4, 181–184. 23. Sakata, Y.; Hayashi, T.; Yasunaga, R.; Yanaga, N.; Imamura, H. Remarkably High Apparent Quantum Yield of the Overall Photocatalytic H2O Splitting Achieved by Utilizing Zn Ion Added Ga2O3 Prepared Using Dilute CaCl2 Solution. Chem. Commun. 2015, 2015 (51), 12935. 24. Abdi, F. F.; Han, L.; Smets, A. H. M.; et al. Efficient Solar Water Splitting by Enhanced Charge Separation in a Bismuth Vanadate-Silicon Tandem Photoelectrode. Nat. Commun. 2013, 4, 2195. 25. Abdi, F. F.; Firet, N.; van de Krol, R. Efficient BiVO4 Thin Film Photoanodes Modified with Cobalt Phosphate Catalyst and W-Doping. ChemCatChem 2013, 5, 490–496. 26. Sivula, K.; Le Formal, F.; Grätzel, M. Solar Water Splitting: Progress Using Hematite (a-Fe2O3) Photoelectrodes. ChemSusChem 2011, 4, 432–449. 27. Liu, J.; Hisatomi, T.; Ma, G.; et al. Improving the Photoelectrochemical Activity of La5Ti2CuS5O7 for Hydrogen Evolution by Particle Transfer and Doping. Energ. Environ. Sci. 2014, 7, 2239–2242. 28. Sakai, E.; Nagamura, N.; Liu, J.; et al. Investigation of the Enhanced Photocathodic Activity of La5Ti2CuS5O7 Photocathodes in H2 Evolution by Synchrotron Radiation Nanospectroscopy. Nanoscale 2016, 8, 18893–18896. 29. Sun, S.; Hisatomi, T.; Wang, Q.; et al. Efficient Redox-Mediator-Free Z-Scheme Water Splitting Employing Oxysulfide Photocatalysts under Visible Light. ACS Catal. 2018, 8, 1690–1696. 30. Wang, X.; Huang, H.; Wang, J.; Li, Z.; Zou, Z. Suppression of Point Defects for Band Edge Engineering in a Semiconducting Photocatalyst. J. Phys. Chem. Lett. 2020, 11, 1708–1713. 31. Arai, N.; Saito, N.; Nishiyama, H.; et al. Effects of Divalent Metal Ion (Mg2þ, Zn2þ and Be2þ) Doping on Photocatalytic Activity of Ruthenium Oxide-Loaded Gallium Nitride for Water Splitting. Catal. Today 2007, 129, 407–413. 32. Izawa, C.; Kobayashi, T.; Kishida, K.; Watanabe, T. Ammonothermal Synthesis and Photocatalytic Activity of Lower Valence Cation-Doped LaNbON2. Adv. Mater. Sci. Eng. 2014, 465720. 33. Seo, J.; Takata, T.; Nakabayashi, M.; et al. Mg–Zr Cosubstituted Ta3N5 Photoanode for Lower-Onset-Potential Solar-Driven Photoelectrochemical Water Splitting. J. Am. Chem. Soc. 2015, 137, 12780–12783. 34. Pan, C.; Takata, T.; Kumamoto, K.; et al. Band Engineering of Perovskite-Type Transition Metal Oxynitrides for Photocatalytic Overall Water Splitting. J. Mater. Chem. A 2016, 4, 4544–4552. 35. Domen, K.; Naito, S.; Soma, M.; Onishi, T.; Tamaru, K. Photocatalytic Decomposition of Water Vapour on an NiO-SrTiO3 Catalyst. J. Chem. Soc. Chem. Commun. 1980, 12, 543–544. 36. Sato, S.; White, J. M. Photodecomposition of Water over Pt/TiO2 Catalysts. Chem. Phys. Lett. 1980, 72, 83–86. 37. Umebayashi, T.; Yamaki, T.; Itoh, H.; Asai, K. Analysis of Electronic Structures of 3d Transition Metal-Doped TiO2 Based on Band Calculations. J. Phys. Chem. Solids 2002, 63, 1909–1920. 38. Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. Visible-Light Photocatalysis in Nitrogen-Doped Titanium Oxides. Science 2001, 293, 269–271. 39. Choi, W.; Termin, A.; Hoffmann, M. R. The Role of Metal Ion Dopants in Quantum-Sized TiO2: Correlation between Photoreactivity and Charge Carrier Recombination Dynamics. J. Phys. Chem. 1994, 98, 13669–13679. 40. Tang, J.; Cowan, A.; Durrant, J. R.; Klug, D. R. Mechanism of O2 Production from Water Splitting: Nature of Charge Carriers in Nitrogen Doped Nanocrystalline TiO2 Films and Factors Limiting O2 Production. J. Phys. Chem. C 2011, 115 (7), 3143–3150. 41. Lv, M.; Sun, X.; Wei, S.; et al. Ultrathin Lanthanum Tantalate Perovskite Nanosheets Modified by Nitrogen Doping for Efficient Photocatalytic Water Splitting. ACS Nano 2017, 11, 11441–11448. 42. Maeda, K.; Tokunaga, Y.; Hibino, K.; et al. New Precursor Route Using a Compositionally Flexible Layered Oxide and Nanosheets for Improved Nitrogen Doping and Photocatalytic Activity. ACS Appl. Energy Mater. 2018, 1, 1734–1741. 43. Suzuki, H.; Tomita, O.; Higashi, M.; Nakada, A.; Abe, R. Improved Visible-Light Activity of Nitrogen-Doped Layered Niobate Photocatalysts by NH3-Nitridation with KCl Flux. Appl. Catal. Environ. 2018, 232, 49–54. 44. Kudo, A.; Yoshino, S.; Tsuchiya, T.; Udagawa, Y.; Takahashi, Y.; et al. Z-Scheme Photocatalyst Systems Employing Rh- and Ir-Doped Metal Oxide Materials for Water Splitting under Visible Light Irradiation. Faraday Discuss. 2019, 215, 313–328. 45. Morikawa, T.; Saeki, S.; Suzuki, T.; Kajino, T.; Motohiro, T. Dual Functional Modification by N Doping of Ta2O5: P-Type Conduction in Visible-Light-Activated N-Doped Ta2O5. Appl. Phys. Lett. 2010, 96, 142111. 46. Suzuki, T. M.; Saeki, S.; Sekizawa, K.; et al. Photoelectrochemical Hydrogen Production by Water Splitting over Dual-Functionally Modified Oxide: P-Type N-Doped Ta2O5 Photocathode Active Under Visible Light Irradiation. Appl. Catal. Environ. 2017, 202, 597–604. 47. Zhang, J.; Zhang, M.; Lin, S.; Fu, X.; Wang, X. Molecular Doping of Carbon Nitride Photocatalysts with Tunable Bandgap and Enhanced Activity. J. Catal. 2013, 310, 24–30. 48. Schwinghammer, K.; Tuffy, B.; Mesch, M. B.; et al. Triazine-Based Carbon Nitrides for Visible-Light-Driven Hydrogen Evolution. Angew. Chem. Int. Ed. 2013, 52, 2435–2439. 49. Bhunia, M. K.; Yamauchi, K.; Takanabe, K. Harvesting Solar Light with Crystalline Carbon Nitrides for Efficient Photocatalytic Hydrogen Evolution. Angew. Chem. Int. Ed. 2014, 53, 11001–11005. 50. Yamato, K.; Iwase, A.; Kudo, A. Photocatalysis Using a Wide Range of the Visible Light Spectrum: Hydrogen Evolution from Doped AgGaS2. ChemSusChem 2015, 8, 2902–2906. 51. Kato, H.; Kudo, A. Visible-Light-Response and Photocatalytic Activities of TiO2 and SrTiO3 Photocatalysts Codoped with Antimony and Chromium. J. Phys. Chem. B 2002, 106, 5029–5034. 52. Niishiro, R.; Kato, H.; Kudo, A. Nickel and either Tantalum or Niobium-Codoped TiO2 and SrTiO3 Photocatalysts with Visible-Light Response for H2 or O2 Evolution from Aqueous Solution. Phys. Chem. Chem. Phys. 2005, 7, 2241–2245. 53. Niishiro, R.; Konta, R.; Kato, H.; et al. Photocatalytic O2 Evolution of Rhodium and Antimony-Codoped Rutile-Type TiO2 under Visible Light Irradiation. J. Phys. Chem. C 2007, 111, 17420. 54. Nukumizu, K.; Nunoshige, J.; Takata, T.; et al. TiNxOyFz as a Stable Photocatalyst for Water Oxidation in Visible Light (< 570 nm). Chem. Lett. 2003, 32, 196–197. 55. Maeda, K.; Shimodaira, Y.; Lee, B.; et al. Studies on TiNxOyFz as a Visible-Light-Responsive Photocatalyst. J. Phys. Chem. C 2007, 111 (49), 18264–18270. 56. Miyoshi, A.; Vequizo, J. J. M.; Nishioka, S.; et al. Nitrogen/Fluorine-Codoped Rutile Titania as a Stable Oxygen-Evolution Photocatalyst for Solar-Driven Z-Scheme Water Splitting. Sustain. Energy Fuels 2018, 2, 2025–2035. 57. Nakada, A.; Nishioka, S.; Vequizo, J. J. M.; et al. Solar-Driven Z-Scheme Water Splitting Using Tantalum/Nitrogen Co-Doped Rutile Titania Nanorod as an Oxygen Evolution Photocatalyst. J. Mater. Chem. A 2017, 5, 11710–11719. 58. Sayama, K.; Mukasa, K.; Abe, R.; Abe, Y.; Arakawa, H. A New Photocatalytic Water Splitting System Under Visible Light Irradiation Mimicking a Z-Scheme Mechanism in Photosynthesis, 2002. 59. Sayama, K.; Mukasa, K.; Abe, R.; Abe, Y.; Arakawa, H. Stoichiometric Water Splitting into H2 and O2 Using a Mixture of Two Different Photocatalysts and an IO3
/I
Shuttle Redox Mediator Under Visible Light Irradiation. Chem. Commun. 2001, 2416–2417. 60. Ishii, T.; Kato, H.; Kudo, A. H2 Evolution from an Aqueous Methanol Solution on SrTiO3 Photocatalysts Codoped with Chromium and Tantalum Ions Under Visible Light Irradiation. J. Photochem. Photobiol. A Chem. 2004, 163, 181–186. 61. Yamakata, A.; Kawaguchi, M.; Murachi, R.; Okawa, M.; Kamiya, I. Dynamics of Photogenerated Charge Carriers on Ni- and Ta-Doped SrTiO3 Photocatalysts Studied by TimeResolved Absorption and Emission Spectroscopy. J. Phys. Chem. C 2016, 120 (15), 7997–8004.

418

Doped semiconductor photocatalysts

62. Konta, R.; Ishii, T.; Kato, H.; Kudo, A. Photocatalytic Activities of Noble Metal Ion Doped SrTiO3 under Visible Light Irradiation. J. Phys. Chem. B 2004, 108, 8992–8995. 63. Kato, H.; Hori, M.; Shimodaira, Y.; Kudo, A. Construction of Z-Scheme Type Heterogeneous Photocatalysis Systems for Water Splitting into H2 and O2 Under Visible Light Irradiation. Chem. Lett. 2004, 33, 1348–1349. 64. Kato, H.; Sasaki, Y.; Iwase, A.; Kudo, A. Role of Iron Ion Electron Mediator on Photocatalytic Overall Water Splitting Under Visible Light Irradiation Using Z-Scheme Systems. Bull. Chem. Soc. Jpn. 2007, 80, 2457–2464. 65. Sasaki, Y.; Iwase, A.; Kato, H.; Kudo, A. The Effect of Co-Catalyst for Z-Scheme Photocatalysis Systems with an Fe3þ/Fe2þ Electron Mediator on Overall Water Splitting Under Visible Light Irradiation. J. Catal. 2008, 259, 133–137. 66. Sasaki, Y.; Nemoto, H.; Saito, K.; Kudo, A. Solar Water Splitting Using Powdered Photocatalysts Driven by Z-Schematic Interparticle Electron Transfer without an Electron Mediator. J. Phys. Chem. C 2009, 113, 17536–17542. 67. Iwase, A.; Ng, Y. H.; Ishiguro, Y.; Kudo, A.; Amal, R. Reduced Graphene Oxide as a Solid-State Electron Mediator in Z-Scheme Photocatalytic Water Splitting under Visible Light. J. Am. Chem. Soc. 2011, 133, 11054–11057. 68. Suzuki, S.; Matsumoto, H.; Iwase, A.; Kudo, A. Enhanced H2 Evolution over an Ir-Doped SrTiO3 Photocatalyst by Loading of an Ir Cocatalyst Using Visible Light up to 800 nm. Chem. Commun. 2018, 54, 10606–10609. 69. Suzuki, S.; Iwase, A.; Kudo, A. Long Wavelength Visible Light-Responsive SrTiO3 Photocatalysts Doped with Valence-Controlled Ru for Sacrificial H2 and O2 Evolution. Catal. Sci. Technol. 2020, 10, 4912–4916. 70. Kawasaki, S.; Akagi, K.; Nakasugi, K.; et al. Elucidation of Rh-Induced In-Gap States of Rh:SrTiO3 Visible-Light-Driven Photocatalyst by Soft X-ray Spectroscopy and FirstPrinciples Calculations. J. Phys. Chem. C 2012, 116, 24445–24448. 71. Furuhashi, K.; Jia, Q.; Kudo, A.; Onishi, H. Time-Resolved Infrared Absorption Study of SrTiO3 Photocatalysts Codoped with Rhodium and Antimony. J. Phys. Chem. C 2013, 117, 19101–19106. 72. Iwashina, K.; Kudo, A. Rh-Doped SrTiO3 Photocatalyst Electrode Showing Cathodic Photocurrent for Water Splitting under Visible-Light Irradiation. J. Am. Chem. Soc. 2011, 133, 13272–13275. 73. Niishiro, R.; Tanaka, S.; Kudo, A. Hydrothermal-Synthesized SrTiO3 Photocatalyst Codoped with Rhodium and Antimony with Visible-Light Response for Sacrificial H2 and O2 Evolution and Application to Overall Water Splitting. Appl. Catal. Environ. 2014, 150–151, 187–196. 74. Asai, R.; Nemoto, H.; Jia, Q.; et al. A Visible Light Responsive Rhodium and Antimony-Codoped SrTiO3 Powdered Photocatalyst Loaded with an IrO2 Cocatalyst for Solar Water Splitting. Chem. Commun. 2014, 50, 2543–2546. 75. Wang, Q.; Hisatomi, T.; Ma, S. S. K.; Li, Y.; Domen, K. Core/Shell Structured La- and Rh-Codoped SrTiO3 as a Hydrogen Evolution Photocatalyst in Z-Scheme Overall Water Splitting under Visible Light Irradiation. Chem. Mater. 2014, 26, 4144–4150. 76. Wang, Q.; Hisatomi, T.; Jia, Q.; et al. Scalable Water Splitting on Particulate Photocatalyst Sheets with a Solar-to-Hydrogen Energy Conversion Efficiency Exceeding 1%. Nat. Mater. 2016, 15, 611–615. 77. Moss, B.; Wang, Q.; Butler, K. T.; et al. Linking In-situ Charge Accumulation to Electronic Structure in Doped SrTiO3 Reveals Design Principles for Hydrogen Evolving Photocatalysts. Nat. Mater. 2021, 20 (4), 511–517. 78. Onishi, H. Sodium Tantalate Photocatalysts Doped with Metal Cations: Why Are They Active for Water Splitting? ChemSusChem 2019, 12, 1825–1834. 79. Kato, H.; Kudo, A. New Tantalate Photocatalysts for Water Decomposition into H2 and O2. Chem. Phys. Lett. 1998, 295, 487–492. 80. Kato, H.; Kudo, A. Highly Efficient Decomposition of Pure Water into H2 and O2 over NaTaO3 Photocatalysts. Catal. Lett. 1999, 58, 153–155. 81. Kudo, A.; Kato, H. Effect of Lanthanide-Doping into NaTaO3 Photocatalysts for Efficient Water Splitting. Chem. Phys. Lett. 2000, 331, 373–377. 82. Iwase, A.; Kato, H.; Kudo, A. The Effect of Alkaline Earth Metal Ion Dopants on Photocatalytic Water Splitting by NaTaO3 Powder. ChemSusChem 2009, 2, 873–877. 83. Kato, H.; Asakura, K.; Kudo, A. Highly Efficient Water Splitting into H2 and O2 over Lanthanum-Doped NaTaO3 Photocatalysts with High Crystallinity and Surface Nanostructure. J. Am. Chem. Soc. 2003, 125, 3082–3089. 84. Sudrajat, H.; Kitta, M.; Ito, R.; et al. Water-Splitting Activity of La-Doped NaTaO3 Photocatalysts Sensitive to Spatial Distribution of Dopants. J. Phys. Chem. C 2020, 124, 15285–15294. 85. An, L.; Onishi, H. Electron–Hole Recombination Controlled by Metal Doping Sites in NaTaO3 Photocatalysts. ACS Catal. 2015, 5, 3196–3206. 86. An, L.; Kitta, M.; Iwase, A.; et al. Photoexcited Electrons Driven by Doping Concentration Gradient: Flux-Prepared NaTaO3 Photocatalysts Doped with Strontium Cations. ACS Catal. 2018, 8, 9334–9341. 87. Kato, H.; Kobayashi, M.; Hara, M.; Kakihana, M. Fabrication of SrTiO3 Exposing Characteristic Facets Using Molten Salt Flux and Improvement of Photocatalytic Activity for Water Splitting. Cat. Sci. Technol. 2013, 3, 1733–1738. 88. Mu, L.; Zhao, Y.; Li, A.; et al. Enhancing Charge Separation on High Symmetry SrTiO3 Exposed with Anisotropic Facets for Photocatalytic Water Splitting. Energ. Environ. Sci. 2016, 9, 2463–2469. 89. Hisatomi, T.; Domen, K. Progress in the Demonstration and Understanding of Water Splitting Using Particulate Photocatalysts. Curr. Opin. Electrochem. 2017, 2, 148–154.

6.19

Structure-reactivity relations in electrocatalysis

Sebastian A. Watzele, Batyr Garlyyev, Elena Gubanova, and Aliaksandr S. Bandarenka, Physics of Energy Conversion and Storage, Physik-Department, Technische Universität München, Garching bei München, Germany © 2023 Elsevier Ltd. All rights reserved.

6.19.1 6.19.1.1 6.19.1.2 6.19.2 6.19.2.1 6.19.2.2 6.19.3 6.19.4 References

Introduction Fundamentals of electrocatalysis Electrocatalytic reactions in electrolyzers and fuel cells Structure-reactivity relations The generalized coordination number Direct instrumental investigation of active sites Structural change of catalysts during the reaction Summary

419 419 421 422 423 429 431 434 435

Abstract This chapter concerns electrocatalysis as a sub-division of heterogeneous catalysis dealing with reactions at the electrified interfaces. The aim is to provide an understanding of the concept of structure-activity relations in electrocatalysis. The readers will be able to understand the concept of catalyst active sites and their identification. The chapter is structured in three main sections. Initially, the fundamental principles of electrocatalysis and the basics of electrolyzers and fuel cells are introduced. The concept of active sites and so-called volcano plots are elucidated. In the second section, the structure-activity relations in electrocatalysis are elaborated in detail. The electrocatalytic performance of various surface facets, motifs are described. Further, a method based on scanning tunneling microscopy for the direct experimental identification of active sites is demonstrated. In the third section, the restructuring of different electrocatalyst surfaces during reaction conditions is discussed. In addition, aspects of the stability of nanostructured alloyed metal electrocatalysts and approaches to increase it are elaborated.

6.19.1

Introduction

Electrocatalysis is a discipline in chemistry that deals with reactions taking place at the electrified interfaces, and it is considered as a sub-division of heterogeneous catalysis. It plays a paramount role in renewable energy storage and conversion systems. In this section, besides the structure-activity relations, the fundamentals of electrocatalysis will also be introduced. As an example, for an industrial application of electrocatalysis, the basic principles of renewable energy conversion systems such as water electrolyzers and fuel cells will be highlighted.

6.19.1.1

Fundamentals of electrocatalysis

In recent years, the need to produce an efficient and stable catalyst for various renewable energy conversion systems brought electrochemical and catalytic communities closer together. Various mechanistic principles and features are common in both disciplines, and the fundamental principles of Sabatier,1 Langmuir,2 Taylor3 hold in both fields. Now, let us take a closer look at the details of these principles. According to the Sabatier principle described in 1902,1,4 an effective catalyst needs to have optimal interaction with reactants and reaction intermediates. They should bind to the catalyst surface neither too strong nor too weak. A certain binding strength is required for sufficient activation. However, too strong binding leads to poisoning of surface sites. The quantitative description of the Sabatier principle was introduced by Balandin5 in the late 1960th in the form of a so-called “volcano” plot. The name originates from the shape of the plot. Such volcano plots depict a catalyst performance indicator, a kinetic property, as a function of a thermodynamic property describing the interaction of reaction intermediates with the catalyst surface. Later, in 1972, volcano plots were adapted in the field of electrocatalysis by Trasatti.6 Kinetic properties such as the activation energy or current density are the catalytic performance indicators. This indicator is plotted versus a descriptor, i.e., the binding energy or the adsorption Gibbs free energy for reactants or reaction intermediates. The usual shape of a volcano plot is triangular with a single peak, where the position indicates the optimal binding energy. As an example, consider the volcano plot for the hydrogen evolution reaction (HER). Fig. 1 shows a volcano plot presenting the experimental exchange current density (j0) as a function of the calculated Gibbs free energy for hydrogen adsorption (D GH*) over different transition metal catalysts.7 Varying material composition and structure allows identifying the D GH* value for optimal catalytic activity. This allows screening to identify promising

Comprehensive Inorganic Chemistry III, Volume 6

https://doi.org/10.1016/B978-0-12-823144-9.00021-2

419

420

Structure-reactivity relations in electrocatalysis

Fig. 1 Volcano plot of experimentally measured exchange current densities as a function of Gibbs free energy of hydrogen adsorption on pure metals calculated by density functional theory. Open symbols indicate single crystal data; solid symbols indicate polycrystalline materials. Adapted with permission of IOP Publishing, Ltd., from reference, Nørskov, J. K.; Bligaard, T.; Logadottir, A. et al. Trends in the Exchange Current for Hydrogen Evolution. J. Electrochem. Soc. 2005, 152, J23–J26. Copyright (2005). Permission conveyed through Copyright Clearance Center, Inc.

catalytic materials with the optimum binding energy. Furthermore, modern computational techniques can be used to calculate the binding energies and accelerate the catalyst screening to identify the promising catalyst materials. The interaction of reactants with the catalyst’s surface was first quantified in 1916 as the strength of adsorption by Langmuir.8 The related Langmuir model describes reactant and intermediate adsorption as a partial monolayer on sites with homogeneous binding strength and no interaction among the sites.9 This model provides the first surface science approach in heterogeneous catalysis. The mathematical interpretation of the model at constant temperature leads to a specific asymptotic increase of coverage with rising partial pressure, the so-called Langmuir isotherm. More complex isotherm shapes are observed in practice, implying that more than one type of site might be present at a catalyst surface. These observations lead to the idea that catalyst surface should be regarded as a checkboard with different sites. In 1925, H.S. Taylor3 suggested that the activity of the catalyst surface can be non-uniformly distributed, and reaction kinetics can depend on individual features. In other words, the overall catalytic activity of a material can be governed by certain highly active sites. Such non-uniform surface sites arise from the different compositions of catalysts and different coordination numbers of surface atoms. The efficiency of an active catalytic site at the molecular level can be expressed as the turnover frequency (TOF), which quantifies how many catalytic reaction cycles proceed per site and per unit of time. Taylor already predicted in the same work that “[t]here will be all extremes between the case in which all the atoms in the surface are active and that in which relatively few are so active” and “[.] the amount of surface which is catalytically active is determined by the reaction catalyzed.”3 This turned out largely correct, and therefore, a distinction was made between reactions where all surface sites are active and the case that only specific sites are active: i) Structure insensitive reactions: every single site may contribute to catalyst activity. The turnover frequency of each site, i.e., each surface atom, is independent of the structure and cluster size in which the atom is located. As a consequence, the only objective in catalyst design would be to obtain a material with the highest possible dispersion of active atoms. The maximum concentration of single active atoms on the electrode surface would be most desirable, and complex structural features cannot further enhance the activity. ii) Structure sensitive reactions: according to Boudart’s classification, those catalytic reactions where the rate changes as the particle size of the (supported) catalyst is changed or as the crystallographic face of a (single crystal) catalyst is altered.10 This implies that only a specific structural feature induces catalytic activity. Hence, typically a minimum ensemble size or coordination number is required to constitute a highly active site. Therefore, TOF expressed per active surface atom typically rises with increasing cluster size until reaching the ensemble size of the active site. Consequently, the objective in catalyst design is to obtain the highest concentration of clusters with specific structural features containing the maximum number of those active sites. Higher or lower dispersion of active atoms will lead to a lowered mass activity of the material, i.e., activity per quantity of the active metal. Various active sites could be relevant in structure-sensitive reactions. Possible structures of sites can be rationalized by features exposed by an active metal cluster on single crystalline surfaces. A related schematic in Fig. 2 shows prominent examples of possible

Structure-reactivity relations in electrocatalysis

421

Fig. 2 Schematic of different adsorption sites on single crystal surfaces. Four different types of adsorption sites are indicated: top, 2-fold bridge, 3fold and 4-fold sites.

surface sites. As indicated, even single crystalline surfaces have various adsorption sites. Distinguishable are top, 2-fold bridge, 3-fold and 4-fold sites, where the adsorption can take place on top of one atom, between two, three or even four atoms, respectively. Further details about the structure-activity relationship will be explained in the second section of this chapter.

6.19.1.2

Electrocatalytic reactions in electrolyzers and fuel cells

In the following, several electrocatalytic reactions, which are important for renewable energy conversion systems such as water electrolyzers and hydrogen fuel cells, are used as the model reactions. The carbon-free molecular structure makes hydrogen one of the most environmentally benign energy carriers. Fuel cells allow direct conversion of chemical energy to electrical energy without emissions of environmentally harmful gases.11 However, for a clean hydrogen-based economy, the production of hydrogen gas should be based on renewable energy sources. Water electrolysis allows generating hydrogen with only oxygen as a coupled product. Therefore, water electrolyzers and hydrogen fuel cells are key devices of energy conversion and storage in terms of the hydrogen economy.12 Water electrolysis is a process using a direct current to split water molecules into oxygen and hydrogen in a device called an electrolyzer. The water electrolyzer or electrolytic cell was invented in the 18th century by Deiman and van Troostwijk.13 The electrolyzer is equipped with two electrodes connected internally via the electrolyte and externally via the outer electric power supply. In water splitting, a hydrogen-containing species is reduced to form gaseous hydrogen in an electrode reaction called hydrogen evolution reaction (HER). At the anode, an oxygen-containing species is oxidized to gaseous oxygen in the so-called oxygen evolution reaction (OER). In acidic media, protons are the main charge carrier that diffuse from the anode to the cathode via a highly effective bond shift mechanism. Protons are reduced to hydrogen in the HER by recombination of two hydrogen atoms at the cathode. The OER occurs at the anode and consists of several reaction steps. Reactions at the electrodes and across the overall cell is written as follows: OER at the anode :

2 H2 O ðlÞ/4 e þ 4 Hþ ðaqÞ þ O2 ðgÞ

HER at the cathode : Overall reaction :

4 e þ 4 Hþ ðaqÞ/2 H2 ðgÞ 2 H2 OðlÞ/2 H2 ðgÞ þ O2 ðgÞ

In alkaline media, hydroxyl ions are the main charge carriers in the electrolyte. The HER at the cathode converts water to hydroxyl species and gaseous hydrogen. Hydroxyl species are oxidized in the OER at the anode under the formation of water and gaseous oxygen. The overall reaction across the cell is written as follows: OER at the anode : 4 OH ðaqÞ/4 e þ 2 H2 OðlÞ þ O2 ðgÞ HER at the cathode :

4 e þ 4 H2 OðlÞ/4 OH ðaqÞ þ 2 H2 ðgÞ

Overall reaction :

2 H2 OðlÞ/2 H2 ðgÞ þ O2 ðgÞ

422

Structure-reactivity relations in electrocatalysis

The detailed mechanism of the electrolysis is significantly more complex than the above consideration of reactions implies. Hence, limitations to the reaction kinetics result from specific reaction steps proceeding rather slowly. At standard conditions, the thermodynamic potential for the HER and the OER are 0 V and 1.23 V vs. reference reversible hydrogen electrode (RHE), respectively. Therefore, the thermodynamic cell equilibrium potential is 1.23 V (at standard conditions). However, to achieve a certain production rate, considerably higher potentials need to be applied during actual water electrolysis. The difference between the applied potential and the equilibrium potential is called overpotential. More active electrocatalysts can enhance the electrochemical process due to their morphological, topological, and electrical properties and can, in turn, reduce the overpotential in water electrolysis. The “counterpart” to an electrolyzer is the fuel cell (FC), which converts chemical energy back into electrical energy. Compared to combustion engines, hydrogen fuel cells have higher thermodynamic efficiency and operate at low temperatures. In the simplest case, hydrogen recombines with oxygen via a pair of spatially separated redox reactions. In 1839 William Robert Grove created the first fuel cell when he discovered that electricity could be generated through the reaction of hydrogen and oxygen dissolved in water at Pt electrodes.14 Modern fuel cells are distinguished by the type of electrolyte used: Solid oxide fuel cells (SOFC) and molten carbonate fuel cells (MCFC) operate at such high temperatures that pure hydrogen is seldom considered to be an appropriate fuel. The use of pure hydrogen is common for phosphoric acid fuel cells (PAFC), alkaline fuel cells (AFC) and polymer electrolyte membrane fuel cells (PEMFC). The main products of hydrogen conversion are water, heat, and electricity without almost any emission of hazardous pollutions and greenhouse gases. Fuel cell designs are comparable to electrolyzers consisting typically of a cathode, anode, and electrolyte in between. Fuel species (like hydrogen) react at catalysts (usually platinum-based), covering the anode to form electrons and protons. The ORR at the catalytic surface of the cathode is in PEMFCs, usually catalyzed by platinum-based alloys. The electrochemical reduction involves electrons supplied from the anode via the external circuit. The formal formation of oxide ions is followed by their recombination with protons passing the electrolyte from the anode to the cathode. As an example, the chemical reactions for the PEMFC system can be expressed as follows: HOR at the anode : ORR at the cathode : Overall reaction :

2 H2 ðgÞ/4 Hþ ðaqÞ þ 4 e

O2 ðgÞ þ 4 Hþ ðaqÞ þ 4e /2 H2 OðlÞ O2 ðgÞ þ 2 H2 ðgÞ/2 H2 OðlÞ

As shown, the combination of electrolysis and fuel cell units allows building an efficient and virtually emission-free system for electrical energy provision. Key advantages are the high energy density (by mass) of hydrogen enabled by chemical energy storage. Additionally, hydrogen is interesting for automotive applications as refueling time is considerably shorter than recharging of batteries. Nevertheless, the compactness of the system and attainable power levels crucially depend on the efficiency of the electrochemical conversion. The performance of the electrochemical device, such as fuel cells and electrolyzers, can be significantly enhanced by properly selected catalyst materials. The latter is directly dependent on features of the catalyst and reaction medium, which influence activity and the reaction mechanism. The reactions taking place in a PEM hydrogen fuel cell, namely the HOR and the ORR, are good examples of highly structure sensitive reactions,15–17 with great significance for state-of-the-art energy conversion and storage applications.

6.19.2

Structure-reactivity relations

The nature of the near-surface electrocatalyst significantly affects the catalytic activity and selectivity for structure structure-sensitive reactions. For example, even for the model surfaces such as Pt(111), the ORR activity can be clearly improved by introducing steps or concavities.18–20 Therefore, when the target is to reduce the amount of required catalyst, which is particularly desirable for scarce noble metals, rational design of catalyst structure becomes indispensable.21,22 This requires a quantitative understanding of the structure-activity relations. Hence, one should first clarify what is meant by “structure” and “arrangement of atoms with different coordination numbers.” To build up a fundamental understanding, model structures such as highly ordered single crystalline surfaces are typically used.19,23,24 The obvious advantage of such periodic surfaces is their limited number of different surface motifs. This and the periodic boundary conditions also simplify simulations of these surfaces. Apart from simulations, empirical electrochemical measurements are an important pillar and necessary to verify the theoretical predictions. For this purpose, the single crystalline surfaces are produced by cutting and abrading single crystals. Different cutting planes through the crystal lattice result in a variety of periodic surfaces.25 Cuts along the principal axes of the lattice result in simple surfaces such as the (111) plane or the (100) plane.25 Those surfaces are planar at the atomic level without steps and have a low number of structural motifs. Other cuts lead to more complex periodically stepped surfaces with various terrace lengths and step shapes, resulting in numerous surface motifs to be investigated. The stereographic triangle shown in Fig. 3 is a small excerpt of the full stereographic projection and summarizes the most common investigated crystal faces. Exemplary surfaces of face-centered cubic (fcc) crystals are visualized with their corresponding particles. As

Structure-reactivity relations in electrocatalysis

423

Fig. 3 Stereographic triangle of an fcc crystal with a (111) normal. Schematics of common crystal surfaces and corresponding nanoparticles are shown.

can be seen, the surface shell of nanoparticles is often composed of single-crystal surfaces, which implies the relevance of predictions made from model surfaces to an actual application. The particles shown in Fig. 3 represent the case of nanoparticles composed of only one type of crystal face. For instance, the surface of cubic particles solely consists of edges and (100)-planes. The electrochemical activity and selectivity are dependent on the adsorption properties of the structural motifs with respect to the reaction intermediates. These are determined by the local electron density, which results from the superposition of the wave functions for the atoms present. Therefore, binding energies can be estimated by density function theory, DFT calculations. However, although several simplifications are typically made, such estimations are resource-intensive, as accurate calculations of large structures are often too expensive in the sense of computational power.26 Hence, simplified rules for structure-activity relations are advantageous for the identification of electrocatalytically active structures and their rational design.

6.19.2.1

The generalized coordination number

To assess the relative binding energies of different surface motifs, coordination number (cn) and level of local strain are often important parameters. For the commonly known coordination number, the next nearest neighbor atoms j are counted.27,28 cn ¼

nj X

1 ¼ nj

(1)

k¼1

For illustration, we consider a top site of a Pt(111) surface as shown in Fig. 4A. Around the red marked Pt atom in the center, there are six neighboring surface atoms and three neighboring Pt atoms in the next lower layer. Hence, the coordination number can be calculated: cn ¼ nj ¼ 6 þ 3 ¼ 9. However, especially for nanostructured catalysts, a qualitative prediction of binding energies based on the number of nearest neighboring atoms is often imprecise. If one looks at a stepped surface, the binding energy changes drastically. For example, examining a (533) surface as shown in Fig. 4B, it is striking that the coordination number obviously changes only for the atoms immediately at a step edges. The coordination number for the red marked central atom would remain at 9, similar to the case of (111) surface, although the actual binding energy should change noticeably. A much better correlation between structural features and the actual binding energy can be achieved if also the second nearest neighboring atoms are considered.29,30 In the generalized coordination number CN 29,30 these are accounted by weighting the neighboring atoms with the number of their own neighbors. The

424

Structure-reactivity relations in electrocatalysis

(A)

(111) surface

(B)

(533) surface

Fig. 4 Single crystal surfaces as examples for the estimation of coordination and generalized coordination numbers. The coordination number of the central atom and for the next nearest neighbor atoms are shown. (A) Example of a (111) surface. The central (red marked) atom and all surface atoms have a coordination number of 9, while the atoms in the next layer have a coordination number of 12. (B) Example of a (533) surface. The central atom has a coordination number of 9 again. However, two of its neighboring surface atoms have a coordination number of cn¼10, which has an influence on the generalized coordination number of the central atom.

generalized coordination number is calculated as the sum of the coordination numbers cn(j) of each neighboring atom j, divided by the maximum coordination number cnmax.29,30 CN ðiÞ ¼

nj X cnð jÞ j¼1

cnmax

(2)

In the example of the (111) surface, the generalized coordination number is calculated as follows. The six surface atoms themselves each have a coordination number of cnsurface ¼ 9 and the three neighboring atoms in the second uppermost layer have the coordination number of bulk atoms in the fcc lattice, i.e., cnbulk ¼ cnmax ¼ 12. Therefore, the generalized coordination number is 7.5, as shown in Eq. (3): CN ðiÞ ¼

nj X cnð jÞ

cnmax j¼1

¼

6  9 þ 3  12 ¼ 7:5 12

(3)

The generalized coordination number of the labeled atom on the (533) surface, on the other hand, is higher (CN ¼ ¼ 7:67) as the two neighboring atoms on the step edge have an increased coordination number (10 instead of 9). In the following example, OH adsorption energies of different Pt surfaces atoms are compared with their conventional and generalized coordination numbers. It should be noted that Pt was selected as the model catalyst, and similar observations can be made on other catalysts. The OH-adsorption energy is of particular relevance for numerous electrochemical reactions such as the ORR, CO oxidation,31 and CO2 reduction32 reaction and frequently used as a descriptor in volcano plots. As examples, extended surfaces, as well as surfaces of nanoparticles of different sizes (the number stands for the number of Pt atoms), are considered. As can be seen in Fig. 5, the coordination number cn for extended surfaces correlates relatively well with the binding energy. However, if one examines smaller nanoparticles, e.g., Pt38, it is striking that the prediction of the binding energies becomes significantly less accurate. With the help of the generalized coordination number, a much better correlation to the OHbinding energy can be achieved. One can observe an almost perfect linear dependence of the OH-binding energy on the generalized coordination number, which enables the identification of active sites based on their surrounding structure. As discussed, the OH-binding energy of the electrode surface strongly influences the ORR activity. Fig. 6A shows the ORR activity of various Pt single-crystal surfaces measured in 0.1 M HClO4. The ORR kinetic currents were measured as a function of the applied voltage. As can be seen, the activity increases significantly when steps are introduced. The highest activity among pristine Pt surfaces is observed for the Pt(221) surface, followed by Pt(331), Pt(775), and the Pt(111) surfaces. When plotting the ORR-specific activity of different Pt surfaces versus the OH binding energies of their active centers, a volcano type plot is observed. As shown in Fig. 6B an ideal ORR catalyst should bind the OH groups  0.1 V weaker than the Pt(111) surface. Therefore, the active sites for the ORR are over-coordinated with respect to surface atoms of Pt(111). The active sites of the (221), (331), (533), and (111) have generalized coordination numbers of CN¼ 8.00, CN¼ 7.83, CN¼ 7.67, and CN¼ 7.50, respectively. This fits well with the observed activity trend. The optimum coordination for a Pt active site would be CN opt ¼ 8.3.33,34 Considering the HER activity of these single-crystal surfaces, a very similar trend can be observed as shown in Fig. 6C. For a meaningful comparison, the activities are normalized to the active surface and corrected by a factor that accounts for different surface density of the active centers. The trend of ORR and the HER activity is very similar because the Pt(111) surface is also binding 49þ210þ312 12

Structure-reactivity relations in electrocatalysis

425

 The statistical Fig. 5 OH-adsorption energies of different sites on Pt model surfaces, nanoparticles, and cavities as a function of (A) cn and (B) CN. data for the trend line is shown in the inset. Reprinted with permission from reference. Calle-Vallejo, F.; Tymoczko, J.; Colic, V. et al. Finding Optimal Surface Sites on Heterogeneous Catalysts by Counting Nearest Neighbors. Science 2015, 350, 185–189. Copyright (2015) AAAS.

the protons slightly stronger than optimum. Therefore, over-coordinated Pt surface sites are more active toward both reactions. As shown in Fig. 6D the activity can be well described by the volcano plot using the generalized coordination number as the descriptor. In the graph SE and SB designate the step edge and bottom, respectively, and AD stands for adatoms on the surface of Pt(111). Note that protons adsorb on the Pt 3-fold sites as shown in Fig. 6E and F. Therefore, the generalized coordination number is different compared to the case of top sites. As an example, for the (111) surface the 3-fold fcc site has a generalized coordination number nj P cnð jÞ 612þ99 ¼ 6:95. The different adsorption sites for the Pt(111), Pt(331) and Pt(533) are shown with correof CN ðiÞ ¼ cnmax ¼ 22 j¼1

sponding CN values in Fig. 6F. The Pt(311)-SB sites have CN ¼ 7:33; which is close to the optimum of CN opt  7:7; which agrees with experimentally measured activities.35 The generalized coordination number can be further extended to strained structures, as compressive and tensile strain can influence the actual binding energies of surfaces.36 In the case of strain, the interatomic distance between two atoms datom(j)  datom(k) differs from the usual distance of atoms in the bulk material dbulk atomatom . Accounting compressions and tensile strains of the lattice spacing, the coordination number can be calculated as36: cnð jÞ ¼

nj X

dbulk atomatom d  datomðkÞ k¼1 atomð jÞ

(4)

Again, the generalized coordination number sums up the coordination numbers of each neighboring atom. This time it also accounts for interatomic distance changes due to strained lattice. Hence, the generalized coordination number for strained surfaces CN ðiÞ estimates as36: CN ðiÞ ¼

nj ni X 1 X dbulk atomatom cnmax j¼1 k¼1 datomð jÞ  datomðkÞ

(5)

In case the distance of all surface near atoms is very similar, it can be described by a mean distance, datomð jÞ  datomðkÞ zdatomatom

(6)

36

Then, the generalized coordination number simplifies to : ! ! nj ni X dbulk 1 X dbulk CN ðiÞ z atomatom 1z atomatom CN ðiÞ datomatom cnmax j¼1 k¼1 datomatom

(7)

We can define the strain as the relative change of interatomic distance36: S¼

datomatom 1 dbulk atomatom

(8)

426

Structure-reactivity relations in electrocatalysis

(A)

(B)

ORR ac vity

(C)

HER ac vity

(D)

(E)

FCC 3-fold site (F)

Fig. 6 Structure-reactivity relation of Pt based catalysts. (A) Kinetic ORR current density of various stepped Pt surfaces in O2 saturated 0.1 M HClO4 electrolyte. (B) The ORR volcano plot: kinetic current density of stepped Pt surfaces as function of their OH binding energy. The surfaces are composed of (111) terraces with length n and (111)-steps (open triangles) or (100)-steps (open squares). Among those surfaces, Pt(221) shows

Structure-reactivity relations in electrocatalysis

427

Obviously, in the case of compressive strain, S takes a value smaller than zero. Conversely, the value of S is positive for tensile strain when the lattice is expanded. Substituting this definition of strain into Eq. (7), the generalized coordination number for the isotropic strain can be rewritten as follows36: X    nj 1 cnð jÞ 1 CN ðiÞ z ¼ CN ðiÞ (9) 1 þ S j¼1 cnmax 1þS Note that for the particular case of zero strain, the equation simplifies meaningfully to CN ðiÞ ¼ CN ðiÞ. In bulk, the strain can reach values up to 6%, however at the surface, typically only values of  3% are reached.37,38 In the following, we look again at the (111) and (533) surfaces shown in Fig. 4 and investigate the effect of the strain. As calcu  lated, the labeled atoms in the case of a relaxed lattice have a generalized coordination number of CN ¼ 7.5 and CN ¼ 7.67, respectively. If the compressive strain of S ¼  3% is applied, the generalized coordination number in both cases increases to    CN ¼ 7.73 and CN ¼ 7.91 according to Eq. (9). Conversely, for the tensile strain of S ¼ 3%, CN values decrease to 7.28 and 7.44. Fig. 7 also shows the cases of lower strain levels such as S  2% and S  1%. Again, the binding energies, estimated by density  functional theory, are affected in a similar way as the generalized coordination numbers. Binding energies and CN show almost 36 perfect linear correlation. In summary, the generalized coordination number enables the estimation of binding energies with significantly reduced computational power. The method allows an effective and relatively good prediction of binding energies and thus enables the forecast of activity of larger structures. Moreover, it gives an easy and descriptive concept for the rational design of catalyst surfaces. This can be used e.g., by machine learning approaches to identify promising nanostructured catalysts.39 In the end, of course, the activities still need to be investigated experimentally. Notably, even DFT calculations cannot perfectly predict electrocatalytic activities since important factors such as the influence of the electrolyte are typically neglected. Further, the focus will be on applications of the generalized coordination number for the structure-activity relations. The CO oxidation reaction is a good example: on the one hand, it has great relevance for energy conversion applications such as

Fig. 7 Trends in DGOH atop for Pt(111), Pt(533), for strain in the range of 3%. The mean/maximum absolute (percentage) errors (MAE/MAX/ MAPE) for the linear fits are provided. Reprinted with permission from reference, Calle-Vallejo, F.; Bandarenka, A. S. Enabling Generalized Coordination Numbers to Describe Strain Effects. ChemSusChem 2018, 11, 1824–1828. Copyright (2018) John Wiley and Sons.

=

highest activity. (C) The HER activity of Pt surfaces in Ar saturated 0.1 M HClO4. Values are normalized to electroactive surface area and corrected by a factor accounting for different surface density of active sites. (D) The HER volcano plot: experimentally measured HER activities of different surface sites as a function of their generalized coordination number. Sites on step edges (SE), at the bottom of step edges (SB) for Pt(331) and Pt(553) and adatoms (AD) on Pt(111) surface are compared. The inset shows that the ideal HER catalyst should bind protons 0.09 V weaker than the Pt(111) surface. The optimal generalized coordination number is CN opt  7:7. (E) The adsorption site on Pt(111) for HER: 3-fold hollow site. The neighboring atoms are labeled with their coordination number. (F) Different adsorption sites with their corresponding CN values. (A) is reprinted and (B) is adapted with permission from reference Calle-Vallejo, F.; Pohl, M. D.; Reinisch, D.; Loffreda, D.; Sautet, P.; Bandarenka, A. S. Why Conclusions from Platinum Model Surfaces Do Not Necessarily Lead to Enhanced Nanoparticle Catalysts for the Oxygen Reduction Reaction. Chem. Sci. 2017, 8, 2283–2289. Copyright (2017) Royal Society of Chemistry. (C, D, E, and F) are reprinted with permission from reference, Pohl, M. D.; Watzele, S.; Calle-Vallejo, F.; Bandarenka, A. S. Nature of Highly Active Electrocatalytic Sites for the Hydrogen Evolution Reaction at Pt Electrodes in Acidic Media. ACS Omega 2017, 2, 8141–8147. Copyright (2017) American Chemical Society.

428

Structure-reactivity relations in electrocatalysis

methanol/ethanol fuel cells. On the other hand, it is relatively simple with only two electrons transferred and is considered a model reaction. CO oxidation consists of three reaction steps40:



COð g Þ þ / CO

(10)

CO þ H2 OðlÞ þ / CO þ  OH þ Hþ þ e

(11)



CO þ  OH þ /CO2 ð g Þ þ Hþ þ e

(12)

The first step is the adsorption of CO, which is followed by the water dissociation/OH adsorption. The reaction is completed with the recombination step where the product, CO2 is released. Looking at the individual reaction enthalpies as shown for Pt surfaces in Fig. 8, it becomes clear that the reaction rate is mainly limited by reaction steps two and three.31 Obviously, for the second reaction step, the OH adsorption energy is a critical factor. As shown before, it is influenced by the surface structure and can be estimated by the generalized coordination number. With growing CN values, the binding energy also increases which results in an acceleration of the second reaction step. On the other hand, increasing CN also rises the binding energy of the reactants of the third reaction step (*CO and *OH), which decreases its reaction rate. According to Sabatier’s principle, there is an optimal configuration in which the speed of the overall reaction reaches its maximum. This occurs at a binding energy of DGCOopt ¼  1.1 eV and DGOHopt ¼ 0.4 eV for CO and OH, respectively. On a platinum surface, this requires a CN opt of 5.4 as shown in Fig. 8.31 These slightly undercoordinated motifs are found on step edges which naturally occur on less than perfect Pt(111) surfaces. However, a much higher number of optimal step edges can be found, for instance, on (331) and (211) surfaces.31 This theoretical prediction from generalized coordination number can be investigated experimentally e.g., by CO-stripping voltammetry. The investigated Pt surfaces (775), (221), (331) and (110) are very similar and consist of (111) terraces with the length of 7, 4, 3, and 2, respectively and (111) step edges. Additional comparison is made with the Pt(111) surface with and without additional cavities. CO adsorbed in an almost complete monolayer on the electrode surface is oxidized during an anodic potential scan.

(A)

(C)

(B)

Fig. 8 (A) Reaction enthalpy of the three steps of CO oxidation for different active sites i on Pt as a function of their generalized coordination  ði Þ. (B) Electrochemical overpotential h for the CO oxidation on Pt as a function of CN  ði Þ. Highest TOFs can be expected for sites with number CN  ði Þ ¼ 5:4; which shows optimal binding energies. Sites close to the step edges of Pt(331) and Pt(221) possess almost optimal CN  . (C) CO CN oxidation voltammetry for various single-crystalline Pt electrodes: Pt(111), Pt(775), Pt(331), Pt(221), Pt(110), Pt(111) with concave surface defects, and CuPt(111) surface alloy. During the first anodic scan in 0.1 M HClO4 the adsorbed CO layer is oxidized and stripped off from the electrode surfaces. dE/dt ¼ 50 mV/s. Reprinted with permission from reference, Calle-Vallejo, F.; Pohl, M. D.; Bandarenka, A. S. Quantitative Coordination– Activity Relations for the Design of Enhanced Pt Catalysts for CO Electro-Oxidation. ACS Catal. 2017, 7, 4355–4359. Copyright (2017) American Chemical Society.

Structure-reactivity relations in electrocatalysis

429

Comparing the oxidation peaks, the lowest potential is found for the (331) surface, followed by the (221), (110), (775), (111) and the (111) surface with cavities. As can be seen, the steps lower the peak potential and thus increase the activity of the surfaces. The activity increases with shorter terrace length up to a terrace length of three, as the density of step edges increases. This is in good agreement with the theoretical predictions since the generalized coordination number of active centers changes for terraces shorter than three atoms in length. In small cavities, the sites are all over coordinated CN  7.5, which results in unfavorable binding energies. Therefore, sluggish reaction kinetics and the relatively high peak potential which can be observed in Fig. 8C. In accord with previous considerations, trisoctahedron-shaped Pt particles as shown in the stereograph triangle in Fig. 3 would be a promising CO reduction catalyst as it exposes numerous sites with optimized CN value due to its (331) surfaces. The binding energies could be even further tailored by introducing just the right amount of strain e.g., by alloying/dealloying catalyst subsurface. Other prominent examples of structure sensitive reactions, where CN can help to rationally design catalysts is the ORR in acidic media on Pt surfaces with an optimal value of CN opt ¼ 8.3.33 This means, that optimal sites are slightly over coordinated compared to the (111) surface (CN ¼ 7.5), which can be most effectively achieved by surface concavities. Similarly, also the HER requires slightly over coordinated sites CN ¼ 7.7. In contrast, the CO2 reduction reaction on Cu surfaces is with CN¼ 3.1 an example of reactions catalyzed by strongly undercoordinated surface atoms.32 To reach such low CN values, step edges alone are insufficient and small adatom clusters of 2–3 atoms are required.32

6.19.2.2

Direct instrumental investigation of active sites

So far focus was on structure-activity relation to predict activities and catalytically active structural motives. Of course, such predictions must be verified experimentally. This is often done by examining reaction rates at different model structures. However, this is an indirect way of measurement, and typically only average TOFs of rather large surfaces are investigated. Therefore, one is often blind for single features such as imperfections of crystal surfaces which can influence the activity. In the following, a method for the direct instrumental investigation of active sites is presented. It is based on scanning tunnel microscopy (STM) that was invented in 1981.41 Tunneling microscopy is an imaging technique in which a tip is used to scan the sample surface without actual contact using the tunnel current between the tip and the sample. To be precise, the height topology of constant electron density is imaged, and the measurement provides information about the local electronic structure.41 Without the additional application of a conductive coating, this method is only possible with conductive or semiconductive samples. There are two general modes of application: first, the constant height mode, where the height is fixed and the tunneling current is monitored and, second, the constant current mode, where an internal feedback loop adjusts the height such that the current stays constant during the scan. In the electrochemical (EC)-STM, which was derived from STM, 7 years later, the sample to be examined is the working electrode of a 3-electrode cell.42 Thus, its electrochemical potential can be controlled. The scanning tip represents an additional electrode. It allows measurements in the electrochemical environment. Note that the current measured at the tip is now not only composed of tunneling current but also of faradaic current and capacitive currents. To suppress the latter contributions, the tip is largely insulated, and tip potentials are chosen such that faradaic reactions are minimized. A special application of the EC-STM for the direct experimental identification of active centers is the noise measurement developed in 2017.43 Here, as the name implies, the noise levels of the recorded signal are analyzed. If a reaction takes place on the surface of the working electrode, i.e., locally between the sample and the tip, this leads to a fluctuation of the measurable current at the STM tip. This is due to several factors such as changes of the tunneling barrier, alterations of the dielectric constant, and redox reactions at the tip e.g., by reaction products. Therefore, the local signal noise level gives information about the local reaction rate. In practical applications, an image is normally acquired once without electrochemical reaction, in order to image the surface morphology in the best possible way. Further images are recorded at applied overpotential with respect to the investigated reaction. This allows to spatially map the noise level of the surface which can be attributed to electrochemical activity. This measurement method can be illustrated by the example of HER on a Pt(111) surface in 0.1 M HClO4 electrolyte, as shown in Fig. 9. When a slightly negative potential (vs. RHE) is applied, the HER is enabled (HER “ON”). As can be seen in Fig. 9A, when the tip is above a Pt(111) terrace, considerably less noise is recorded compared to the case when the tip is above a step edge, shown in Fig. 9B. However, when the potential is increased, HER is disabled (HER “OFF), the overall noise level decreases, and the local noise level of step edges approaches the noise level of the terraces, as it is seen in Fig. 9C. This implies that the noise is in fact, the result of the HER, which is related to the diffusion of reactants and products toward and away from the active centers, respectively. Therefore, noise EC-STM discloses the position of electrocatalytically active sites for the HER at the vicinity of the step edges. This can be seen by looking at a single line scan over several step edges as presented in Fig. 9D. This finding agrees with other experimental works, which confirm that the stepped Pt surfaces show enhanced HER activity.35 The advantages of the noise EC-STM method include the up to atomic-scale spatial resolution and the direct experimental identification of active sites. This method has been successfully utilized in detecting the active sites on a variety of different materials for various reactions.43–47 It should be noted that, up to now, the activity assessments are rather qualitative. However, by applying different overpotentials, more exact statements about activity differences can be made.43 The latest efforts are to directly link the noise distribution itself with the actual TOF and thus to convert measured noise data into TOF maps.44

430

Structure-reactivity relations in electrocatalysis

Fig. 9 Schematic of Noise-EC-STM measurements to detect electrochemically active sites for the HER on a Pt single crystal surface. At the condition “off” where sample potentials are chosen such that the reaction rate becomes minimal, the surface morphology can be investigated. Changing the potential below the onset of the HER starts the reaction “on” which manifests in noise in the signal of the tunneling current. The level of the noise is an indicator of the local electrochemical reaction rate. In this example, certain sites at the steps have optimized catalytic properties, which results in an increased noise level at the step edges. (A) Minimum noise level of tip current at the Pt (111) terraces. (B) Increased noise level at the step edges caused by the HER. (C) EC-STM image of a Pt surface consisting of a (111) terrace and 10-atom step edge. At the HER “ON” a significantly increased noise level is detected at the vicinity of the step edge indicating active centers. (D) An EC-STM line scan over several terraces and step edges recorded at HER “ON” condition. Reprinted with permission from reference, Pfisterer, J. H.; Liang, Y.; Schneider, O.; Bandarenka, A. S. Direct Instrumental Identification of Catalytically Active Surface Sites. Nature 2017, 549, 74–77. Copyright (2017) Springer Nature.

(A)

(B)

Fig. 10 TEM images of Pt nanoparticles (black) on carbon support (grey). The images show how the nanoparticles lose their shape after long-term usage as the oxygen reduction reaction electrocatalyst. (A) The initial TEM image of Pt nanoparticles supported on carbon. (B) The same Pt nanoparticles after 10 k cycles between 0.6 and 1.0 V vs RHE. Adapted with permission from reference, Fichtner, J.; Watzele, S.; Garlyyev, B. et al. Tailoring the Oxygen Reduction Activity of Pt Nanoparticles through Surface Defects: A Simple Top-Down Approach. ACS Catal. 2020, 10, 3131–3142. Copyright (2020) American Chemical Society.

Structure-reactivity relations in electrocatalysis

6.19.3

431

Structural change of catalysts during the reaction

In real electrochemical systems, the catalyst structure can undergo drastic changes, which significantly affects their performance. This section focuses on electrocatalysts’ structural changes during their utilization. Bulk and nanostructured metal/metal oxide materials in contact with liquid electrolytes are considered. In electrocatalysis, in addition to typical parameters such as the operational temperature, pressure, and the environment, the operating potential also influences the stability of catalysts.48 Moreover, recently, several studies showed that even electrolyte species that don’t participate in the reaction itself could have a drastic influence on the overall catalyst performance and stability.49,50 As an example for surface restructuring, let us consider what happens to Pt nanoparticles supported on carbon supports used as an electrocatalyst toward ORR. As previously discussed, the ORR is of paramount importance for fuel cells but also for other applications such as metal-air batteries. In Fig. 10A an initial TEM image of Pt on carbon is shown. The TEM image given in Fig. 10B illustrates the same Pt nanoparticles after 10 k potential cycles between 0.6 and 1.0 V (potential range at which the typical hydrogen fuel cell operates) at 80  C.51 As it is seen in the images, the nanoparticles lose their initial structure, which was tailored for maximum activity, and form much larger aggregates. This, in turn, decreases their overall performance as surface area and the density of active sites are reduced. Therefore, it is important to understand how the catalyst materials change their structure over the course of their lifespan. Moreover, the question arises of how such structural changes can be prevented to preserve their high performance for a longer time. When metals are subjected to certain potentials, they undergo surface changes such as oxidation or reduction, and certain species in the electrolyte can adsorb to or desorb from the catalyst surface. For instance, when single-crystalline platinum, Pt(111), is

Fig. 11 STM image of Pt(111) surface after the surface was subjected to oxidation and reduction cycles. Initially, the surface is clean, and only a few Pt islands are seen. Once the electrode is exposed to potential cycling in 0.1 M HClO4 electrolyte, the number of islands starts to increase. After 31 cycles, the Pt clusters are larger in size and cover the complete surface. After exposing the surface to 170 cycles, the islands mainly grow in height, and no significant changes can be seen in the STM image. The potential cycling was done between 0.06 and 1.35 V vs reversible hydrogen electrode at the scan speed of 50 mV/s. The investigated area is highlighted by the red line. Reprinted with permission from reference, Jacobse, L.; Huang, Y. F.; Koper, M. T.; Rost, M. J. Correlation of Surface Site Formation to Nanoisland Growth in the Electrochemical Roughening of Pt (111). Nat. Mater. 2018, 17, 277–282. Copyright (2018) Springer Nature.

432

Structure-reactivity relations in electrocatalysis

Fig. 12 Reconstruction of single-crystalline Cu(100) surface during hydrogen evolution reaction in 0.1 M HClO4 solution. As indicated by red arrows, starting from the top left image at  0.58 V vs saturated calomel electrode (SCE) shifting toward HER potentials, clear hydrogen-induced surface reconstruction is observed. Adapted with permission from reference, Matsushima, H.; Taranovskyy, A.; Haak, C.; Gründer, Y.; Magnussen, O. M. Reconstruction of cu (100) Electrode Surfaces during Hydrogen Evolution. J. Am. Chem. Soc. 2009, 131, 10362–10363. Copyright (2009) American Chemical Society.

subjected to oxidation and reduction cycles, the surface undergoes significant transformations.52 Typically, the surface adsorption and desorption processes at the platinum surface can be monitored by voltammetric experiments. These are highly sensitive to the atomic scale surface sites participating in the reaction. Specifically, this has been extensively investigated for low index singlecrystalline platinum surfaces. In addition to voltammetric methods, surface imaging techniques such as above introduced scanning tunneling microscopy (STM) are utilized to gain more information about the surface structure during electrode polarization. As an example, consider the Pt(111) surface during the oxidation and reduction cycling. Fig. 11 illustrates the STM images of Pt(111) surface during potential cycling. Initially, the surface is clean, and only a few islands of small Pt clusters are observed. These initial small islands are produced when the surface is annealed during Pt(111) preparation. After few cycles (ca. 8 cycles), the number of islands increases, and until ca. 30 cycles, they also grow (mainly in diameter). Afterward, up to 170 cycles, the islands mainly grow in height. These experimental findings suggest that the overall island formation/roughening process occurs in two steps: an early nucleation growth step, followed by the growth of already formed islands. As it is seen in the STM images (Fig. 11), during the initial step, the number of nanoscale islands/clusters continuously increases on the Pt surface. In the second phase, the islands are coalescing, and already formed clusters/islands continue to grow in size, predominantly in height. Furthermore, the extent of such structural transformations is increased with the amplitude of the applied potential. For instance, when alternating potentials of  3 V (or higher) are applied to bulk Pt wires which are immersed in alkaline electrolytes, they tend to erode relatively quickly. In this case, one can even observe Pt nanoparticles which are first formed on the wires and then released into the electrolyte.53 As shown, the surface of the catalyst materials undergoes significant structural changes during the electrocatalytic reaction. This, in turn, can have a major impact on surface activity for structure-sensitive reactions, as discussed in the previous Section 6.19.2. As another example, let us take a look at the surface reconstruction of Cu(100) during the HER.54 An in-situ STM with high temporal resolution images shown in Fig. 12 illustrates how the surface of Cu(100) electrode is reconstructed during the HER, which is induced by surface adsorbed hydrogen. When the potential is shifted from  0.58 V vs saturated calomel electrode (SCE) toward more negative potentials (see Fig. 12), the surface phase transition occurs within a few seconds (video available at ref. 54). Such surface reconstruction of the Cu(100) electrode drastically affects the kinetics of the HER. The reconstructed Cu lattice resulted in enhanced electrocatalytic performance, highlighting such phase transitions can also be beneficial for electrocatalytic activity.

Structure-reactivity relations in electrocatalysis

433

Fig. 13 Schematic description of structural changes of PtNi nanoparticles during their utilization as oxygen reduction reaction electrocatalyst. After 25 potential cycles situation is designated as activated, and stability stands for after 4 k potential cycles. The inset images show the TEM images of nanoparticles after 4 k cycles. Adapted with permission from reference, Cui, C.; Gan, L.; Heggen, M.; Rudi, S.; Strasser, P. Compositional Segregation in Shaped Pt Alloy Nanoparticles and their Structural Behaviour During Electrocatalysis. Nat. Mater. 2013, 12, 765–771. Copyright (2013) Springer Nature.

As it was illustrated in Fig. 10, the nanoparticles of pure metal undergo significant structural changes during their utilization as electrocatalysts. Such structural changes become even more apparent for alloyed bimetallic nanoparticles.55 For instance, let us consider PteNi alloys, which are well-known oxygen reduction reaction catalysts used at the cathode side of fuel cells.56 Because of the harsh acidic environment in the fuel cells, transition metal atoms in Pt-M alloys tend to leach out faster than the more noble Pt atoms, leaving strained Pt layers. This is especially true at the anodic potentials. Fig. 13 shows how PteNi octahedral nanoparticles lose their geometry over time when used to catalyze the ORR. During the initial 25 cycles, the PteNi nanoparticles are activated where a significant portion of surface Ni atoms are leached out. The activated PteNi samples show the highest electrocatalytic  ðiÞ  activity toward the ORR. This enhancement is caused by strained surface and concave structures with over coordinated (CN 7:5) Pt atoms with optimized binding energies, as discussed in the previous chapter. After the stability tests, the activities decrease significantly, Pt1.5Ni, PtNi1.5, and PtNi catalysts dropped by 16%, 45%, and 66%, respectively. Ni composition at the bulk of nanoparticles drops to Pt4Ni1. As discussed above, the catalysts’ structural changes can be detrimental to its overall performance. In the past decades, there have been various developments in preventing such abrupt structural changes over time. Herein, some approached to prevent catalyst degradation over time are briefly introduced. Therefore, pure and alloyed Pt electrocatalysts for the ORR are considered due to their widespread applications. One approach to prevent structural deformation of Pt nanoparticles is to carefully modify the surface with elaborately selected organic or inorganic surfactant materials. Of course, such an approach will reduce the number of available active sites. However, it allows tailoring toward maximum long-term performance. In a similar manner, one can tune the support material by functionalization to optimize its interaction with Pt nanoparticles for better stability.

434

Structure-reactivity relations in electrocatalysis

Fig. 14 Enhanced stability of PteRheNi octahedral nanostructured oxygen reduction reaction electrocatalysts. HAADF STEM images and EDX composition maps of PteRheNi octahedral nanoparticles (A, B, C) initial, (D, E) after 8 k cycles and (F, G) after 30 k cycles. Reprinted with permission from reference, Beermann, V.; Gocyla, M.; Willinger, E. et al. Rh-Doped Pt–Ni Octahedral Nanoparticles: Understanding the Correlation between Elemental Distribution, Oxygen Reduction Reaction, and Shape Stability. Nano Lett. 2016, 16, 1719–1725. Copyright (2016) American Chemical Society.

When dealing with alloyed Pt-based electrocatalysts, a key reason for their long-term activity degradation is the continuous leaching of the less noble solute metal. Moreover, the leached metals can poison other parts of the system leading to further performance losses. Such a phenomenon can be circumvented by pre-treating the alloyed particles in an acidic solution before their utilization. Interestingly, near-surface doping of the PteNi electrocatalysts with Rh metal had shown improved stability when they were utilized for the ORR.57 Fig. 14 illustrates the HAADF STEM images and EDX composition maps of PteRheNi octahedral nanoparticles. In comparison to bimetallic PteNi octahedral nanoparticles, the ternary structure preserved its initial structure even after 30 k potential cycling. It was found that the shape losses in PteNi nanoparticles occur mainly due to the diffusion of Pt atoms during the potential cycling. Therefore, it is assumed that the Rh atoms suppress the diffusion of Pt atoms. Hence, the presence of small amounts of Rh atoms prevents a loss in geometry at the same time maintaining the active surface structure during their longterm usage.

6.19.4

Summary

Developing stable and active electrocatalysts is crucial for building highly efficient electrochemical energy conversion devices. The electrocatalytic activity is determined by the nature of active sites, which are influenced by various factors such as the materials composition, surface structure and electrolyte composition. Combining theoretical principles and experimental observation allows predictions to disclose the optimal active phase. We have explained the concept of generalized coordination number as a versatile tool for activity predictions based on surface structure and for rational catalyst design. The noise EC-STM and its application for the direct experimental identification of electrochemically active sites were discussed briefly. Finally, the nature of structural changes of catalyst surfaces was investigated. There are three changes that need to be considered when utilizing nanostructured bimetallic

Structure-reactivity relations in electrocatalysis

435

alloyed catalyst materials: the changes in their size, shape, and composition. The approaches to diminish the effects of leaching and structural changes for alloyed Pt-based catalysts are elaborated. In the coming years, the advancements in theoretical material screening and experimental electrocatalyst synthesis will enable the development of highly efficient electrochemical energy conversion devices. Such breakthroughs are of great importance for tackling the global greenhouse emission problems and for the sustainable transition of our society toward renewable energy systems.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.

Sabatier, P. Hydrogénations et déshydrogénations par catalyse. Ber. Dtsch. Chem. Ges. 1911, 44, 1984–2001. Langmuir, I. Chemical Reactions by Surfaces. Trans. Faraday Soc. 1922, 17, 607–620. Taylor, H. S. A Theory of the Catalytic Surface. Proc. R. Soc. London, Ser. A 1925, 108, 105–111. Sabatier, P. New Synthesis of Methane. Comptes Rendus 1902, 134, 514–516. Balandin, A. A. Modern State of the Multiplet Theor of Heterogeneous Catalysis. Adv. Catal. 1969, 19, 1. Trasatti, S. Work Function, Electronegativity, and Electrochemical Behaviour of Metals: III. Electrolytic hydrogen evolution in acid solutions. J. Electroanal. Chem. Interfacial Electrochem. 1972, 39, 163–184. Nørskov, J. K.; Bligaard, T.; Logadottir, A.; et al. Trends in the Exchange Current for Hydrogen Evolution. J. Electrochem. Soc. 2005, 152, J23–J26. Langmuir, I. The Constitution and Fundamental Properties of Solids and Liquids. Part I. Solids. J. Am. Chem. Soc. 1916, 38, 2221–2295. Langmuir, I. The Adsorption of Gases on Plane Surfaces of Glass, Mica and Platinum. J. Am. Chem. Soc. 1918, 40 (9), 1361–1403. Thomas, J. M.; Thomas, W. J. Principles and Practice of Heterogeneous Catalysis, 2nd ed.; Wiley-VCH: Weinheim, 2015. Stambouli, A. B. Fuel Cells: The Expectations for an Environmental-Friendly and Sustainable Source of Energy. Renew. Sustain. Energy Rev. 2011, 15, 4507–4520. Zini, G.; Tartarini, P. Electrolysis and Fuel Cells. In Solar Hydrogen Energy Systems, Springer: Milano, 2012; pp 29–52. Levie, R. The Electrolysis of Water. J. Electroanal. Chem. 1999, 476, 92–93. Grove, W. R. On Voltaic Series and the Combination of Gases by Platinum. Philos. Mag. J. Sci. 1839, 86-87, 127–130. Sánchez-Sánchez, C. M.; Solla-Gullón, J.; Vidal-Iglesias, F. J.; et al. Imaging Structure Sensitive Catalysis on Different Shape-Controlled Platinum Nanoparticles. J. Am. Chem. Soc. 2010, 132 (16), 5622–5624. Markovic, N. M.; Sarraf, S. T.; Gasteiger, H. A.; et al. Hydrogen Electrochemistry on Platinum Low-Index Single-Crystal Surfaces in Alkaline Solution. J. Chem. Soc., Faraday Trans. 1996, 92, 3719–3725. Hoshi, N.; Asaumi, Y.; Nakamura, M.; et al. Sructural Effects on the Hydrogen Oxidation Reaction on N (111)-(111) Surface of Platinum. J. Phys. Chem. C 2009, 113, 16843– 16846. Gómez-Marín, A. M.; Feliu, J. M. Oxygen Reduction on Nanostructured Platinum Surfaces in Acidic Media: Promoting Effect of Surface Steps and Ideal Response of Pt(111). Catal. Today 2015, 244, 172–1769. Maciá, M. D.; Campiña, J. M.; Herrero, E.; Feliu, J. M. On the Kinetics of Oxygen Reduction on Platinum Stepped Surfaces in Acidic Media. J. Electroanal. Chem. 2004, 564, 141–150. Kuzume, A.; Herrero, E.; Feliu, J. M. Oxygen Reduction on Stepped Platinum Surfaces in Acidic Media. J. Electroanal. Chem. 2007, 599, 333–343. Guo, S.; Wang, E. Noble Metal Nanomaterials: Controllable Synthesis and Application in Fuel Cells and Analytical Sensors. Nano Today 2011, 6, 240–264. Tian, N.; Lu, B. A.; Yang, X. D.; et al. Rational Design and Synthesis of Low-Temperature Fuel Cell Electrocatalysts. Electrochem. Energy Rev. 2018, 1, 54–83. Ogasawara, H.; Brena, B.; Nordlund, D.; et al. Structure and Bonding of Water on Pt (111). Phys. Rev. Lett. 2002, 89, 276102. Gomez-Marin, A. M.; Rizo, R.; Feliu, J. M. Oxygen Reduction Reaction at Pt Single Crystals: A Critical Overview. Catal. Sci. Technol. 2014, 4, 1685–1698. Arulmozhi, N.; Jerkiewicz, G. Design and Development of Instrumentations for the Preparation of Platinum Single Crystals for Electrochemistry and Electrocatalysis Research. Part 2: Orientation, Cutting, and Annealing. Electrocatalysis 2017, 8, 399–413. Hernandez, S. Development of Methods for Reducing the Cost of Density Functional Theory and Time-Dependent Density Functional Theory. Doctoral dissertation, UCLA, 2015. Calle-Vallejo, F.; Loffreda, D.; Koper, M. T.; Sautet, P. Introducing Structural Sensitivity into Adsorption–Energy Scaling Relations by Means of Coordination Numbers. Nat. Chem. 2015, 7, 403–410. Jiang, T.; Mowbray, D. J.; Dobrin, S.; et al. Trends in CO Oxidation Rates for Metal Nanoparticles and Close-Packed, Stepped, and Kinked Surfaces. J. Phys. Chem. C 2009, 113, 10548–10553. Calle-Vallejo, F.; Sautet, P.; Loffreda, D. Understanding Adsorption-Induced Effects on Platinum Nanoparticles: An Energy-Decomposition Analysis. J. Phys. Chem. Lett. 2014, 5, 3120–3124. Calle-Vallejo, F.; Martínez, J. I.; García-Lastra, J. M.; Sautet, P.; Loffreda, D. Fast Prediction of Adsorption Properties for Platinum Nanocatalysts with Generalized Coordination Numbers. Angew. Chem., Int. Ed. 2014, 53, 8316–8319. Calle-Vallejo, F.; Pohl, M. D.; Bandarenka, A. S. Quantitative Coordination–Activity Relations for the Design of Enhanced Pt Catalysts for CO Electro-Oxidation. ACS Catal. 2017, 7, 4355–4359. Zhao, Z.; Chen, Z.; Zhang, X.; Lu, G. Generalized Surface Coordination Number as an Activity Descriptor for CO2 Reduction on Cu Surfaces. J. Phys. Chem. C 2016, 120, 28125–28130. Calle-Vallejo, F.; Tymoczko, J.; Colic, V.; et al. Finding Optimal Surface Sites on Heterogeneous Catalysts by Counting Nearest Neighbors. Science 2015, 350, 185–189. Calle-Vallejo, F.; Pohl, M. D.; Reinisch, D.; Loffreda, D.; Sautet, P.; Bandarenka, A. S. Why Conclusions from Platinum Model Surfaces Do Not Necessarily Lead to Enhanced Nanoparticle Catalysts for the Oxygen Reduction Reaction. Chem. Sci. 2017, 8, 2283–2289. Pohl, M. D.; Watzele, S.; Calle-Vallejo, F.; Bandarenka, A. S. Nature of Highly Active Electrocatalytic Sites for the Hydrogen Evolution Reaction at Pt Electrodes in Acidic Media. ACS Omega 2017, 2, 8141–8147. Calle-Vallejo, F.; Bandarenka, A. S. Enabling Generalized Coordination Numbers to Describe Strain Effects. ChemSusChem 2018, 11, 1824–1828. Escudero-Escribano, M.; Malacrida, P.; Hansen, M. H.; et al. Tuning the Activity of Pt Alloy Electrocatalysts by Means of the Lanthanide Contraction. Science 2016, 352, 73–76. Pedersen, A. F.; Ulrikkeholm, E. T.; Escudero-Escribano, M. Probing the Nanoscale Structure of the Catalytically Active Overlayer on Pt Alloys with Rare Earths. Nano Energy 2016, 29, 249–260. Rück, M.; Garlyyev, B.; Mayr, F.; Bandarenka, A. S.; Gagliardi, A. Oxygen Reduction Activities of Strained Platinum Core–Shell Electrocatalysts Predicted by Machine Learning. J. Phys. Chem. Lett. 2020, 1, 1773–1780. Koper, M. T. Structure Sensitivity and Nanoscale Effects in Electrocatalysis. Nanoscale 2011, 3, 2054–2073. Binnig, G.; Rohrer, H. Scanning tunneling microscopy. Surf. Sci. 1983, 126, 236–244. Itaya, K.; Tomita, E. Scanning Tunneling Microscope for Electrochemistry-a New Concept for the In Situ Scanning Tunneling Microscope in Electrolyte Solutions. Surf. Sci. 1988, 201, L507–L512. Pfisterer, J. H.; Liang, Y.; Schneider, O.; Bandarenka, A. S. Direct Instrumental Identification of Catalytically Active Surface Sites. Nature 2017, 549, 74–77.

436

Structure-reactivity relations in electrocatalysis

44. Haid, R. W.; Kluge, R. M.; Liang, Y.; Bandarenka, A. S. In Situ Quantification of the Local Electrocatalytic Activity Via Electrochemical Scanning Tunneling Microscopy. Small Methods 2020, 4, 2000710. 45. Mitterreiter, E.; Liang, Y.; Golibrzuch.; et al. In-Situ Visualization of Hydrogen Evolution Sites on Helium Ion Treated Molybdenum Dichalcogenides under Reaction Conditions. NPJ 2D Mater. Appl. 2019, 3, 1–9. 46. Kluge, R. M.; Haid, R. W.; Bandarenka, A. S. Assessment of Active Areas for the Oxygen Evolution Reaction on an Amorphous Iridium Oxide Surface. J. Catal. 2021, 396, 14–22. 47. Liang, Y.; Csoklich, C.; McLaughlin, D.; Schneider, O.; Bandarenka, A. S. Revealing Active Sites for Hydrogen Evolution at Pt and Pd Atomic Layers on au Surfaces. ACS Appl. Mater. Interfaces 2019, 11, 12476–12480. 48. Geiger, S.; Kasian, O.; Ledendecker, M.; et al. The Stability Number as a Metric for Electrocatalyst Stability Benchmarking. Nat. Catal. 2018, 1, 508–515. 49. Xue, S.; Garlyyev, B.; Auer, A.; Kunze-Liebhauser, J.; Bandarenka, A. S. How the Nature of the Alkali Metal Cations Influences the Double-Layer Capacitance of Cu, Au, and Pt Single-Crystal Electrodes. J. Phys. Chem. C 2020, 124, 12442–12447. 50. Xue, S.; Garlyyev, B.; Watzele, S.; et al. Influence of Alkali Metal Cations on the Hydrogen Evolution Reaction Activity of Pt, Ir, Au, and Ag Electrodes in Alkaline Electrolytes. ChemElectroChem 2018, 5 (17), 2326–2329. 51. Fichtner, J.; Watzele, S.; Garlyyev, B.; et al. Tailoring the Oxygen Reduction Activity of Pt Nanoparticles through Surface Defects: A Simple Top-Down Approach. ACS Catal. 2020, 10, 3131–3142. 52. Jacobse, L.; Huang, Y. F.; Koper, M. T.; Rost, M. J. Correlation of Surface Site Formation to Nanoisland Growth in the Electrochemical Roughening of Pt (111). Nat. Mater. 2018, 17, 277–282. 53. Garlyyev, B.; Watzele, S.; Fichtner.; et al. Electrochemical Top-Down Synthesis of C-Supported Pt Nano-Particles with Controllable Shape and Size: Mechanistic Insights and Application. Nano Research 2020. https://doi.org/10.1007/s12274-020-3281-z. 54. Matsushima, H.; Taranovskyy, A.; Haak, C.; Gründer, Y.; Magnussen, O. M. Reconstruction of cu (100) Electrode Surfaces during Hydrogen Evolution. J. Am. Chem. Soc. 2009, 131, 10362–10363. 55. Bergmann, A.; Roldan Cuenya, B. Operando Insights into Nanoparticle Transformations during Catalysis. ACS Catal. 2019, 9, 10020–10043. 56. Cui, C.; Gan, L.; Heggen, M.; Rudi, S.; Strasser, P. Compositional Segregation in Shaped Pt Alloy Nanoparticles and their Structural Behaviour during Electrocatalysis. Nat. Mater. 2013, 12, 765–771. 57. Beermann, V.; Gocyla, M.; Willinger, E.; et al. Rh-Doped Pt–Ni Octahedral Nanoparticles: Understanding the Correlation between Elemental Distribution, Oxygen Reduction Reaction, and Shape Stability. Nano lett. 2016, 16, 1719–1725.