Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies 9781119870524

Citation preview

Catalysis for a Sustainable Environment

Catalysis for a Sustainable Environment Reactions, Processes and Applied Technologies Volume 1

Edited by Professor Armando J. L. Pombeiro Instituto Superior técnico Lisboa, Portugal

Dr. Manas Sutradhar

Universidade Lusófona de Humanidades e Tecnologias Faculdade de Engenharia Lisboa, Portugal

Professor Elisabete C. B. A. Alegria Instituto Politécnico de Lisboa Departamento de Engenharia Química Lisboa, Portugal

This edition first published 2024 © 2024 John Wiley and Sons Ltd All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/ permissions. The right of Armando J.L. Pombeiro, Manas Sutradhar, and Elisabete C.B.A. Alegria to be identified as the author of the editorial material in this work has been asserted in accordance with law. Registered Offices John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Editorial Office The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com. Wiley also publishes its books in a variety of electronic formats and by print-on-demand. Some content that appears in standard print versions of this book may not be available in other formats. Trademarks: Wiley and the Wiley logo are trademarks or registered trademarks of John Wiley & Sons, Inc. and/ or its affiliates in the United States and other countries and may not be used without written permission. All other trademarks are the property of their respective owners. John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book. Limit of Liability/Disclaimer of Warranty In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. A catalogue record for this book is available from the Library of Congress Hardback ISBN: 9781119870524; ePub ISBN: 9781119870630; ePDF ISBN: 9781119870623; oBook ISBN: 9781119870647 Cover image: © Sasha Fenix/Shutterstock Cover design by Wiley Set in 9.5/12.5pt STIXTwoText by Integra Software Services Pvt. Ltd, Pondicherry, India

v

Contents About the Editors Preface xv

xiii

VOLUME 1 1

Introduction 1 Armando J.L. Pombeiro, Manas Sutradhar, and Elisabete C.B.A. Alegria Structure of the Book 2 Final Remarks 4 Part I

2

2.1 2.2 2.3 2.4 2.5 2.6 2.6.1 2.6.2 2.7

3 3.1 3.2 3.3 3.4 3.5 3.6

Carbon Dioxide Utilization

5

Transition from Fossil-C to Renewable-C (Biomass and CO2) Driven by Hybrid Catalysis 7 Michele Aresta and Angela Dibenedetto Introduction 7 The Dimension of the Problem 8 Substitutes for Fossil-C 8 Hybrid Catalysis: A New World 11 Hybrid Catalysis and Biomass Valorization 13 Hybrid Catalysis and CO2 Conversion 16 CO2 as Building Block 16 CO2 Conversion to Value-added Chemical and Fuels via Hybrid Systems 17 Conclusions 21 References 21 Synthesis of Acetic Acid Using Carbon Dioxide 25 Philippe Kalck Introduction 25 Synthesis of Methanol from CO2 and H2 26 Carbonylation of Methanol Using CO2 28 Carbonylation of Methane Using CO2 31 Miscellaneous Reactions, Particularly Biocatalysis 31 Conclusions 32 References 32

本书版权归John Wiley & Sons Inc.所有 ftoc_Vol1.indd 5

26-12-2023 20:08:10

vi

Contents

4

4.1 4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.3 4.4 4.5

5

5.1 5.1.1 5.1.2 5.1.3 5.1.4 5.1.5 5.1.6 5.1.7 5.1.8 5.2 5.3 5.4 5.5

6

6.1 6.1.1 6.1.2 6.1.3 6.1.4 6.2 6.2.1 6.2.1.1 6.2.1.2

New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources: A New Catalytic Challenge 35 Ana B. Paninho, Malgorzata E. Zakrzewska, Leticia R.C. Correa, Fátima Guedes da Silva, Luís C. Branco, and Ana V.M. Nunes Introduction 35 Cyclic Carbonates from Bio-based Epoxides 37 Bio-based Epoxides Derived from Terpenes 39 Bio-based Vinylcyclohexene Oxide Derived from Butanediol 41 Bio-based Epichlorohydrin Derived from Glycerol 42 Epoxidized Vegetable Oils and Fatty Acids 42 Cyclic Carbonates Derived from Carbohydrates 44 Cyclic Carbonates Derived from Bio-based Diols 46 Conclusions 50 Acknowledgements 50 References 50 Sustainable Technologies in CO2 Utilization: The Production of Synthetic

Natural Gas

55

M. Carmen Bacariza, José M. Lopes, and Carlos Henriques CO2 Valorization Strategies 55 CO2 to CO via Reverse Water-Gas Shift (RWGS) Reaction 56 CO2 to CH4 56 CO2 to CxHy 57 CO2 to CH3OH 58 CO2 to CH3OCH3 58 CO2 to R-OH 59 CO2 to HCOOH, R-COOH, and R-CONH2 60 Target Products Analysis Based on Thermodynamics 60 Power-to-Gas: Sabatier Reaction Suitability for Renewable Energy Storage 61 CO2 Methanation Catalysts 63 Zeolites: Suitable Supports with Tunable Properties to Assess Catalysts’s Performance 64 Final Remarks 68 References 69 Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis 73 Denzil Moodley, Thys Botha, Renier Crous, Jana Potgieter, Jacobus Visagie, Ryan Walmsley, and Cathy Dwyer Introduction 73 Sustainable Aviation Fuels (SAF) via Fischer-Tropsch-based Routes 73 Introduction to FT Chemistry 75 FT Catalysts for SAF Production 79 Reactor Technology for SAF Production Using FTS 81 State-of-the-art Cobalt Catalysts 82 Catalyst Preparation Routes for Cobalt-based Catalysts 85 Precipitation Methodology – a Short Summary 85 Preparation Methods Using Pre-shaped Supports 85

Contents

6.2.1.2.1 6.2.1.2.2 6.2.1.2.3 6.2.1.2.4 6.2.2 6.2.3 6.2.3.1 6.2.3.2 6.2.3.3 6.2.3.4 6.2.3.5 6.2.3.6 6.2.3.7 6.2.4 6.3 6.3.1 6.3.2 6.3.3 6.3.4 6.3.5 6.3.6 6.3.7 6.3.8 6.3.9 6.4 7 7.1 7.2 7.2.1 7.2.1.1 7.2.1.2 7.2.1.3 7.2.2 7.3 7.4

Support Modification  85 Cobalt Impregnation  85 Calcination  86 Reduction  88 Challenges for Catalysts Operating with High Carbon Efficiency: Water Tolerance  88 Strategies to Increase Water Tolerance and Selectivity for Cobalt Catalysts  90 Optimizing Physico-chemical Support Properties for Stability at High Water Partial Pressure  90 Stabilizing the Support by Surface Coating  91 Impact of Crystallite Size on Selectivity  91 Metal Support Interactions with Cobalt Crystallites of Varying Size  92 The Role of Reduction Promoters and Support Promoters in Optimizing Selectivity  94 Role of Pore Diameter in Selectivity  96 Effect of Activation Conditions on Selectivity  98 Regeneration of Cobalt PtL Catalysts- Moving Toward Materials Circularity  100 An Overview of Fe Catalysts: Direct Route for CO2 Conversion  101 Introduction  101 Effect of Temperature  102 Effect of Pressure  103 Effect of H2:CO Ratio  104 Catalyst Development  104 Stability to Oxidation by Water  104 Sufficient Surface Area  105 Availability of Two Distinct Catalytically Active Sites/phases  105 Sufficient Alkalinity for Adsorption and Chain Growth  106 Future Perspectives  106 References  108 Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives  117 Maurizio Peruzzini, Fabrizio Mani, and Francesco Barzagli Introduction  117 Catalytic Synthesis of Urea  119 Urea from CO2 Reductive Processes  120 Electrocatalysis  120 Photocatalysis  122 Magneto-catalysis  123 Urea from Ammonium Carbamate  124 Catalytic Synthesis of Urea Derivatives  127 Conclusions and Future Perspectives  133 Part II Transformation of Volatile Organic Compounds (VOCs)  139

8 8.1 8.1.1 8.1.2

Catalysis Abatement of NOx/VOCs Assisted by Ozone  141 Zhihua Wang and Fawei Lin NOx/VOC Emission and Treatment Technologies  141 NOx/VOC Emissions  141 NOx Treatment Technologies  142

vii

viii

Contents

8.1.2.1 8.1.2.2 8.1.2.3 8.1.2.4 8.1.3 8.1.3.1 8.1.3.2 8.1.3.3 8.1.3.4 8.1.3.5 8.2 8.2.1 8.2.1.1 8.2.1.2 8.2.1.3 8.2.1.4 8.2.2 8.2.2.1 8.2.2.2 8.3 8.3.1 8.3.1.1 8.3.1.2 8.3.2 8.3.2.1 8.3.2.2 8.3.2.3 8.3.3 8.3.3.1 8.3.3.2 8.3.3.3 8.3.4 8.4

SNCR  142 SCR  142 SCR Catalysts  142 Ozone-assisted Oxidation Technology  142 VOC Treatment Technologies  143 Adsorption  143 Regenerative Combustion  143 Catalytic Oxidation  144 Photocatalytic Oxidation  144 Plasma-assisted Catalytic Oxidation  144 NO Oxidation by Ozone  144 NO Homogeneous Oxidation by Ozone  145 Effect of O3/NO Ratio  145 Effect of Temperature  145 Effect of Residence Time  145 Process Parameter Optimization  146 Heterogeneous Catalytic Deep Oxidation  146 Catalytic NO Deep Oxidation by O3 Alone  146 Catalytic NO Deep Oxidation by Combination of O3 and H2O  148 Oxidation of VOCs by Ozone  150 Aromatics  150 Toluene  150 Benzene  153 Oxygenated VOCs  154 Formaldehyde  154 Acetone  154 Alcohols  155 Chlorinated VOCs  155 Chlorobenzene  155 Dichloromethane  155 Dioxins and Furans  156 Sulfur-containing VOCs  157 Conclusions  157 References  157

9

Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions  161 Elisabete C.B.A. Alegria, Manas Sutradhar, and Tannistha R. Barman Introduction  161 Benzene  162 Toluene  167 Ethylbenzene  171 Xylene  172 Final Remarks  175 Acknowledgments  176 References  176

9.1 9.2 9.3 9.4 9.5 9.6

Contents

10

10.1 10.2 10.2.1 10.2.2 10.2.3 10.2.4 10.2.5 10.2.6 10.2.7 10.3 10.4

Catalytic Cyclohexane Oxyfunctionalization  181 Manas Sutradhar, Elisabete C.B.A. Alegria, M. Fátima C. Guedes da Silva, and Armando J.L. Pombeiro Introduction  181 Transition Metal Catalysts for Cyclohexane Oxidation  182 Vanadium Catalysts  182 Iron Catalysts  186 Cobalt Catalysts  189 Copper Catalysts  191 Molybdenum Catalysts  198 Rhenium Catalysts  199 Gold Catalysts  200 Mechanisms  201 Final Comments  202 Acknowledgments  203 References  203 Part III Carbon-based Catalysis  207

11

11.1 11.1.1 11.1.2 11.2 11.2.1 11.3 11.3.1 11.3.2 11.3.3 11.4 12

12.1 12.2 12.3 12.3.1 12.3.2

Carbon-based Catalysts for Sustainable Chemical Processes  209 Katarzyna Morawa Eblagon, Raquel P. Rocha, M. Fernando R. Pereira, and José Luís Figueiredo Introduction  209 Nanostructured Carbon Materials  209 Carbon Surface Chemistry  210 Metal-free Carbon Catalysts for Environmental Applications  212 Wet Air Oxidation and Ozonation with Carbon Catalysts  212 Carbon-based Catalysts for Sustainable Production of Chemicals and Fuels from Biomass  214 Carbon Materials as Catalysts and Supports  214 Cascade Valorization of Biomass with Multifunctional Catalysts  216 Carbon Catalysts Produced from Biomass  219 Summary and Outlook  220 Acknowledgments  221 References  221 Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes  225 Giulia Tuci, Andrea Rossin, Matteo Pugliesi, Housseinou Ba, Cuong Duong-Viet, Yuefeng Liu, Cuong Pham-Huu, and Giuliano Giambastiani Introduction  225 The H2S Selective Oxidation to Elemental Sulfur  226 Alkane Dehydrogenation  231 Alkane Dehydrogenation under Oxidative Environment: The ODH Process  231 Alkane Dehydrogenation under Steam- and Oxygen-free Conditions: The DDH Reaction  237

ix

x

Contents

12.4

Conclusions  240 Acknowledgments  241 References  241

13

Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and, Wastewater Treatment: Cutting-Edge Solutions for a Sustainable World  247 Clara Pereira, Diana M. Fernandes, Andreia F. Peixoto, Marta Nunes, Bruno Jarrais, Iwona Kuźniarska-Biernacka, and Cristina Freire Introduction  247 Hybrid Carbon-metal Oxide Electrocatalysts for Energy-related Applications  249 Oxygen Reduction Reaction (ORR)  249 Oxygen Evolution Reaction (OER)  254 Hydrogen Evolution Reaction (HER)  257 CO2 Reduction Reaction (CO2RR)  259 Biomass Valorization over Hybrid Carbon-metal Oxide Based (Nano)catalysts  261 Advanced (Photo)catalytic Oxidation Processes for Wastewater Treatment  266 Heterogeneous Fenton Process  266 Heterogeneous photo-Fenton Process  271 Heterogeneous electro-Fenton Process  277 Photocatalytic Oxidation  281 Advanced Catalytic Reduction Processes for Wastewater Treatment  288 Conclusions and Future Perspectives  291 Acknowledgments  292 References  292

13.1 13.2 13.2.1 13.2.2 13.2.3 13.2.4 13.3 13.4 13.4.1 13.4.2 13.4.3 13.4.4 13.5 13.6

VOLUME 2 About the Editors  xiii Preface  xv Part IV

Coordination, Inorganic, and Bioinspired Catalysis  299

14

Hydroformylation Catalysts for the Synthesis of Fine Chemicals  301 Mariette M. Pereira, Rui M.B. Carrilho, Fábio M.S. Rodrigues, Lucas D. Dias, and Mário J.F. Calvete

15

Synthesis of New Polyolefins by Incorporation of New Comonomers  323 Kotohiro Nomura and Suphitchaya Kitphaitun

16

Catalytic Depolymerization of Plastic Waste  339 Noel Angel Espinosa-Jalapa and Amit Kumar

17

Bioinspired Selective Catalytic C-H Oxygenation, Halogenation, and Azidation of Steroids  369 Konstantin P. Bryliakov

Contents

18

Catalysis by Pincer Compounds and Their Contribution to Environmental and Sustainable Processes  389 Hugo Valdés and David Morales-Morales

19

Heterometallic Complexes: Novel Catalysts for Sophisticated Chemical Synthesis  409 Franco Scalambra, Ismael Francisco Díaz-Ortega, and Antonio Romerosa

20

Metal-Organic Frameworks in Tandem Catalysis  429 Anirban Karmakar and Armando J.L. Pombeiro

21

(Tetracarboxylate)bridged-di-transition Metal Complexes and Factors Impacting Their Carbene Transfer Reactivity  445 LiPing Xu, Adrian Varela-Alvarez, and Djamaladdin G. Musaev

22

Sustainable Cu-based Methods for Valuable Organic Scaffolds  461 Argyro Dolla, Dimitrios Andreou, Ethan Essenfeld, Jonathan Farhi, Ioannis N. Lykakis, and George E. Kostakis

23

Environmental Catalysis by Gold Nanoparticles  481 Sónia Alexandra Correia Carabineiro

24

Platinum Complexes for Selective Oxidations in Water  515 Alessandro Scarso, Paolo Sgarbossa, Roberta Bertani, and Giorgio Strukul

25

The Role of Water in Reactions Catalyzed by Transition Metals  537 A.W. Augustyniak and A.M. Trzeciak

26

Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts  559 Skyler Markham, Debbie C. Crans, and Bruce Atwater

27

Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions  577 Ana P. Carvalho, Angela Martins, Filomena Martins, Nelson Nunes, and Rúben Elvas-Leitão

VOLUME 3 About the Editors  xiii Preface  xv Part V 28

Organocatalysis  609

Sustainable Drug Substance Processes Enabled by Catalysis: Case Studies from the Roche Pipeline  611 Kurt Püntener, Stefan Hildbrand, Helmut Stahr, Andreas Schuster, Hans Iding and Stephan Bachmann

xi

xii

Contents

29

Supported Chiral Organocatalysts for Accessing Fine Chemicals  639 Ana C. Amorim and Anthony J. Burke

30

Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis  659 Kotohiro Nomura and Nor Wahida Binti Awang

31

Modern Strategies for Electron Injection by Means of Organic Photocatalysts: Beyond Metallic Reagents  675 Takashi Koike

32

Visible Light as an Alternative Energy Source in Enantioselective Catalysis  687 Ana Maria Faisca Phillips and Armando J.L. Pombeiro Part VI

Catalysis for the Purification of Water and Liquid Fuels  717

33

Heterogeneous Photocatalysis for Wastewater Treatment: A Major Step Towards Environmental Sustainability  719 Shima Rahim Pouran and Aziz Habibi-Yangjeh

34

Sustainable Homogeneous Catalytic Oxidative Processes for the Desulfurization of Fuels  743 Federica Sabuzi, Giuseppe Pomarico, Pierluca Galloni, and Valeria Conte

35

Heterogeneous Catalytic Desulfurization of Liquid Fuels: The Present and the Future  757 Rui G. Faria, Alexandre Viana, Carlos M. Granadeiro, Luís Cunha-Silva, and Salete S. Balula Part VII Hydrogen Formation, Storage, and Utilization  783

36

Paraformaldehyde: Opportunities as a C1-Building Block and H2 Source for Sustainable Organic Synthesis  785 Ana Maria Faísca Phillips, Maximilian N. Kopylovich, Leandro Helgueira de Andrade, and Martin H.G. Prechtl

37

Hydrogen Storage and Recovery with the Use of Chemical Batteries  819 Henrietta Horváth, Gábor Papp, Ágnes Kathó, and Ferenc Joó

38

Low-cost Co and Ni MOFs/CPs as Electrocatalysts for Water Splitting Toward Clean Energy-Technology  847 Anup Paul, Biljana Šljukić, and Armando J.L. Pombeiro



Index  871

xiii

About the Editors Armando Pombeiro is a Full Professor Jubilado at Instituto Superior Técnico, Universidade de Lisboa (ULisboa), former Distant Director at the People’s Friendship University of Russia (RUDN University), a Full Member of the Academy of Sciences of Lisbon (ASL), the President of the Scientific Council of the ASL, a Fellow of the European Academy of Sciences (EURASC), a Member of the Academia Europaea, founding President of the College of Chemistry of ULisboa, a former Coordinator of the Centro de Química Estrutural at ULisboa, Coordinator of the Coordination Chemistry and Catalysis group at ULisboa, and the founding Director of the doctoral Program in Catalysis and Sustainability at ULisboa. He has chaired major international conferences. His research addresses the activation of small molecules with industrial, environmental, or biological significance (including alkane functionalization, oxidation catalysis, and catalysis in unconventional conditions) as well as crystal engineering of coordination compounds, polynuclear and supramolecular structures (including MOFs), non-covalent interactions in synthesis, coordination compounds with bioactivity, molecular electrochemistry, and theoretical studies. He has authored or edited 10 books, (co-)authored ca. 1000 research publications, and registered ca. 40 patents. His work received over. 30,000 citations (over 12,000 citing articles), h-index ca. 80 (Web of Science). Among his honors, he was awarded an Honorary Professorship by St. Petersburg State University (Institute of Chemistry), an Invited Chair Professorship by National Taiwan University of Science & Technology, the inaugural SCF French-Portuguese Prize by the French Chemical Society, the Madinabeitia-Lourenço Prize by the Spanish Royal Chemical Society, and the Prizes of the Portuguese Chemical and Electrochemical Societies, the Scientific Prizes of ULisboa and Technical ULisboa, and the Vanadis Prize. Special issues of Coordination Chemistry Reviews and the Journal of Organometallic Chemistry were published in his honor. https://fenix.tecnico.ulisboa.pt/homepage/ist10897

xiv

About the Editors

Manas Sutradhar is an Assistant Professor at the Universidade Lusófona, Lisbon, Portugal and an integrated member at the Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa, Portugal. He was a post‐doctoral fellow at the Institute of Inorganic and Analytical Chemistry of Johannes Gutenberg University of Mainz, Germany and a researcher at the Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa. He has published 72 papers in international peer review journals (including three reviews + 1 reference module), giving him an h-index 28 (ISI Web of Knowledge) and more than 2250 citations. In addition, he has 11 book chapters in books with international circulation and one patent. He is one of the editors of the book Vanadium Catalysis, published by the Royal Society of Chemistry. His main areas of work include metal complexes with aroylhydrazones, oxidation catalysis of industrial importance and sustainable environmental significance, magnetic properties of metal complexes, and bio-active molecules. The major contributions of his research work are in the areas of vanadium chemistry and oxidation catalysis. He received the 2006 Young Scientist Award from the Indian Chemical Society, India and the Sir P. C. Ray Research Award (2006) from the University of Calcutta, India. https://orcid.org/0000-0003-3349-9154 Elisabete C.B.A. Alegria is an Adjunct Professor at the Chemical Engineering Department of the Instituto Superior de Engenharia de Lisboa (ISEL) of the Polytechnic Institute of Lisbon, Portugal. She is a researcher (Core Member) at the Centro de Química Estrutural (Coordination Chemistry and Catalysis Group). She has authored 86 papers in international peered review journals and has an h-index of 23 with over 1600 citations, four patents, five book chapters, and over 180 presentations at national and international scientific meetings. She was awarded an Honorary Distinction (2017–2020) for the Areas of Technology and Engineering (Scientific Prize IPL-CGD). She is an editorial board member, and has acted as a guest editor and reviewer for several scientific journals. Her main research interests include coordination and sustainable chemistry, homogeneous and supported catalysis, stimuli‐responsive catalytic systems, green synthesis of metallic nanoparticles for catalysis, and biomedical applications. She is also interested in mechanochemistry (synthesis and catalysis) and molecular electrochemistry. https://orcid.org/0000-0003-4060-1057

xv

Preface Aiming to change the world for the better, 17 Sustainable Development Goals (SDGs) were adopted by the United Nations (UN) Member States in 2015, as part of the UN 2030 Agenda for Sustainable Development that concerns social, economic, and environmental sustainability. Hence, a 15-year plan was set up to achieve these Goals and it is already into its second half. However, the world does not seem to be on a good track to reach those aims as it is immersed in the Covid-19 pandemic crisis and climate emergency, as well as economic and political uncertainties. Enormous efforts must be pursued to overcome these obstacles and chemical sciences should play a pivotal role. Catalysis is of particular importance as it constitutes the most relevant contribution of chemistry towards sustainable development. This is true even though the SDGs are integrated and action in one can affect others. For example, the importance of chemistry and particularly catalysis is evident in several SDGs. Goal 12, addresses “Responsible Consumption and Production Patterns” and is aligned with the circularity concept with sustainable loops or cycles (e.g., in recycle and reuse processes that are relevant within the UN Environmental Program). Goal 7 addresses “Affordable and Clean Energy” and relates to efforts to improve energy conversion processes, such as hydrogen evolution and oxygen evolution from water, that have a high environmental impact. Other SDGs in which chemistry and catalysis play an evident role with environmental significance include Goal 6 (“Clean Water and Sanitation”), Goal 9 (“Industry, Innovation and Infrastructure” 13 (“Climate Action”), Goal 14 (“Life Below Water”), and Goal 15 (“Life on Land”). The book is aligned with these SDGs by covering recent developments in various catalytic processes that are designed for a sustainable environment. It gathers skilful researchers from around the world to address the use of catalysis in various approaches, including homogeneous, supported, and heterogeneous catalyses as well as photo- and electrocatalysis by searching for innovative green chemistry routes from a sustainable environmental angle. It illustrates, in an authoritative way, state-of-the-art knowledge in relevant areas, presented from modern perspectives and viewpoints topics in coordination, inorganic, organic, organometallic, bioinorganic, pharmacological, and analytical chemistries as well as chemical engineering and materials science. The chapters are spread over seven main sections focused on Carbon Dioxide Utilization, Transformation of Volatile Organic Compound (VOCs), Carbon-based Catalysts, Coordination, Inorganic, and Bioinspired Catalysis, Organocatalysis, Catalysis for the Purification of Water and Liquid Fuels,and Hydrogen Formation, Storage, and Utilization. These sections are gathered together as a contribution towards the development of the challenging topic.

xvi

Preface

The book addresses topics in (i) activation of relevant small molecules with strong environmental impacts, (ii) catalytic synthesis of important added value organic compounds, and (iii) development of systems operating under environmentally benign and mild conditions toward the establishment of sustainable energy processes. This work is expected to be a reference for academic and research staff of universities and research institutions, including industrial laboratories. It is also addressed to post-doctoral, postgraduate, and undergraduate students (in the latter case as a supplemental text) working in chemical, chemical engineering, and related sciences. It should also provide inspiration for research topics for PhD and MSc theses, projects, and research lines, in addition to acting as an encouragement for the development of the overall field. The topic Catalysis for Sustainable Environment is very relevant in the context of modern research and is often implicit, although in a non-systematic and disconnected way, in many publications and in a number of initiatives such as international conferences. These include the XXII International Symposium on Homogeneous Catalysis (ISHC) that we organized (Lisbon, 2022) and that to some extent inspired some parts of this book. In contrast to the usual random inclusion of the topic in the literature and scientific events, the applications of catalytic reactions focused on a sustainable environment in a diversity of approaches are addressed in this book. The topic has also contributed to the significance of work that led to recent Nobel Prizes of Chemistry. In 2022, the Nobel Prize was awarded to Barry Sharpless, Morten Meldal, and Carolyn Bertozzi for the development of click chemistry and bioorthogonal chemistry. The set of criteria for a reaction or a process to meet in the context of click chemistry includes, among others, the operation under benign conditions such as those that are environmentally friendly (e.g., preferably under air and in water medium). In 2021, the Nobel Prize was awarded to Benjamin List and David W.C. MacMillan for the development of asymmetric organocatalysis, which relies on environmentally friendly organocatalysts. The book illustrates the connections of catalysis with a sustainable environment, as well as the richness and potential of modern catalysis and its relationships with other sciences (thus fostering interdisciplinarity) in pursuit of sustainability. At last, but not least, we should acknowledge the authors of the chapters for their relevant contributions, prepared during a particularly difficult pandemic period, as well as the publisher, John Wiley, for the support, patience, and understanding of the difficulties caused by the adverse circumstances we are experiencing nowadays and that constituted a high activation energy barrier that had to be overcome by all of us… a task that required rather active catalysts. We hope the readers will enjoy reading its chapters as much as we enjoyed editing this book. Armando Pombeiro Manas Sutradhar Elisabete Alegria

1

Introduction Armando J.L. Pombeiro1, Manas Sutradhar 2, and Elisabete C.B.A. Alegria3 1 Centro de Química Estrutural and Departamento de Engenharia Química, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal 2 Faculdade de Engenharia, Universidade Lusófona - Centro Universitário de Lisboa, Campo Grande 376, Lisboa, Portugal Centro de Química Estrutural and Departamento de Engenharia Química, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal 3 Departamento de Engenharia Química, ISEL, Instituto Politécnico de Lisboa, Portugal Centro de Química Estrutural and Departamento de Engenharia Química, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal

The relevance of catalysis in our lives is well-documented by its involvement in the industrial production chain for the manufacture of most products, such as petrochemicals, fine chemicals for pharmaceuticals, polymers, fertilizers, and bio-produced materials. Catalysis is also prominent in many biological transformations and connects several areas of chemical and related sciences from different perspectives (e.g. chemistry and energy, chemistry and the environment, Chemistry at the Interface of Biology, pharmacology and medicine, functional biomaterials, materials sciences, and chemical engineering). Catalysis plays a key role in achieving the United Nations (UN) Sustainable Development Goals (SDGs), namely those of environmental significance as mentioned in the Preface of this book. With the aim of ending poverty, protecting the environment and promoting prosperity, the UN embraced 17 SDGs in 2015 and encouraged countries, industries, and organizations around the world to adopt these goals. These actions include, for example, the development of sustainable forms of energy and its storage, the application of green chemistry principles in industrial processes, the recycling of resources, orientation towards a circular economy, the use of low-cost raw materials and of carbon from biomass, the conversion of CO2 and CO from flue gases, and the mitigation of air pollution. The Covid-19 pandemic forced a long period of reflection about the value of human relations and of human interactions with the environment. The pandemic provided a unique opportunity to join efforts towards achieving the above aims of the UN 2030 Agenda that includes the SDGs. However, efforts concerning direct human interactions do not seem to be paving a promising path. Let us hope that the harmonization of human actions with the need for a healthy and sustainable environment will be more successful despite of the difficulties already experienced by initiatives such as the UN Paris Agreement on Climate Change that aims to limit global warming by reducing greenhouse gas emissions. One major environmental concern is pollution. Control of this pollution is a main objective that can be accomplished by work that can be described as environmental catalysis. For example, work Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 1, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

2

  Introduction

in this field aims to contribute towards the reduction of emissions of environmentally unacceptable compounds such as CO2, volatile organic compounds (VOCs), nitrogen oxides (NOx), sulfur oxides (SOx), and CO. It involves the use of catalytic cleanup technologies for this purpose, as well as the conversion of VOCs, liquid and solid waste treatment, and the conversion of greenhouse gases. It also addresses, for instance, the application of catalysis under eco-friendly conditions, the use of catalytic technologies for waste minimization, catalyst recycling, and the development of new catalytic routes for selective synthesis of valuable products. Additional important developments of environmental interest include the use of energy-efficient catalytic processes (which are assisted by low power microwave radiation or ultrasound), catalysis in the reduction of the environmental impact, and catalysis to produce clean fuels. Sustainability is a relevant concern for all of these and green chemistry routes that protect or are compatible with protecting the environment should be pursued. Sustainable processes should replace conventional chemical syntheses and transformations by minimizing the formation of byproducts or waste and bypassing the use of conventional and polluting organic solvents under eco-friendly reaction conditions. Working for a sustainable future is a current challenge and Catalysis for a Sustainable Environment can aid in these efforts.

Structure of the Book This book brings together researchers whose contributions to the development of environmentally sustainable catalytic processes are well recognized. Throughout the chapters, the authors give their perspectives on state-of-the-art approaches and address innovative methodologies in relevant areas of homogeneous, supported, and heterogeneous catalysis, as well as photo-, electro- and magnetocatalysis from a sustainable viewpoint. The book consists of 38 chapters spread over 7 sections (Parts) as follows: (I) Carbon Dioxide Utilization, (II) Transformation of Volatile Organic Compounds (VOCs), (III) Carbon-based Catalysis, (IV) Coordination, Inorganic, and Bioinspired Catalysis, (V) Organocatalysis, (VI) Catalysis for the Purification of Water and Liquid Fuels, and (VII) Hydrogen Formation, Storage, and Utilization. The book has an interdisciplinary character illustrating relevant areas in coordination, inorganic, organic, organometallic, bioinorganic, and pharmacological chemistry, as well as nanochemistry, chemical engineering, and materials science. As addressed in Part I, the use of renewable carbon sources such as biomass and CO2 (Chapter 2) for the manufacture of chemicals such as acetic acid or urea (Chapters 3 and 7), and fuels (Chapters 2 and 5) is a highly active field of research and a step towards a circular carbon economy. Recently, efforts focused on combining CO2 with bio-based resources has highlighted the importance of catalysis in the process viability of converting inactive substrates (Chapter 2). Hybrid catalysis, based on the integrated use of robust and selective chemocatalysts, stands out as one of the important trends for CO2 conversion to added-value chemicals (Chapter 2). The formation of cyclic carbonates by catalytic reactions of CO2 with promising bio-based resources (epoxides, carbohydrates, and diols) is analysed (Chapter 4). The methanation of CO2 (Sabatier reaction) to produce electricity is also addressed, with an emphasis on the performance of zeolite-supported catalysts and on structure-reactivity relations (Chapter 5). The application of Fischer-Tropsch catalysis to convert green H2 and sustainable carbon into kerosene range hydrocarbons for aviation fuels is presented, with one of the indirect routes involving the conversion of CO2 plus green H2 into CO and water via the reverse displacement of water gas shift (Chapter 6). The elimination of nitrogen oxides (NOx) (Chapter 8) and VOCs (Chapters 8 and 9), critical precur-

Structure of the Book

perspective, the most promising NOx and VOCs treatment technologies are highlighted, including catalytic ozonation of NO (Chapter 8) and advanced oxidation processes for VOCs (Chapter 9). Apart from the elimination of aromatic VOCs (Chapter 9) and saturated hydrocarbons (cyclohexane) (Chapter 10), the selective functionalization of these compounds to added-value organic products is discussed. Carbon-based materials have been successfully applied to specific reactions, either as catalysts or as catalyst supports and various synthetic strategies are available, as illustrated in Part III. The versatility of carbon materials is related to their capacity to maximize surface area and be easily functionalized by replacing carbon atoms with heteroatoms such as S, N, O, P, or B, allowing control of their electronic properties (Chapter 11). Methodologies for the introduction of these heteroatoms into carbons and synthetic methods for nanostructured carbons are reviewed and their use as catalysts is discussed in the context of sustainable production of fuels and chemicals, energy conversion, and environmental protection (Chapter 11). State-of-the-art approaches focussed on metal-free carbon-based catalysts for gas-phase industrial oxidation processes (H2S oxidation to sulphur and alkane dehydrogenation) are presented (Chapter 12). The application of emerging ecosustainable carbon-metal oxide (nano)catalysts as electrocatalysts, as catalysts for biomass valorization, and as (photo)(electro)catalysts for water and wastewater treatment is covered (Chapter 13). The use of metal-based coordination compounds for the development of sustainable catalytic protocols is discussed mainly in Part IV for many (technological) processes of organic synthesis. These include hydroformylation (Chapter 14), synthesis of ethylene copolymers by incorporation of sterically encumbered olefins and cyclic olefins (Chapter 15), depolymerization of plastic waste (catalytic hydrogenation, hydrogenolysis, hydrosilylation, and hydroboration) (Chapter 16), tandem reactions (Chapter 20), carbene transfer from diazocarbenes to C–H and C–C π-bonds (Chapter 21), synthesis of organic scaffolds of pharmaceutical importance (Chapter 22), oxidation reactions (Chapters 23–25), cross-coupling reactions (Chapter 26), and Friedel Crafts acylation (Chapter 27). The contribution of pincer complexes to the development of environmentally friendly systems, particularly in hydrogenation and dehydrogenation reactions or in the transformation of CO2 into valuable products such as methanol, is highlighted (Chapter 18). The synergistic cooperation of metals in heterometallic complexes is also emphasized for various homogeneous catalytic processes (Chapter 19). Rules governing the regio- and stereoselectivity of catalytic functionalizations in the presence of biologically-inspired transition metal-based catalysts are addressed to provide mechanistic insights into selective bioinspired C-H oxygenations, halogenations, and azidations of steroids and terpenoids (Chapter 17). Several chapters address the significance of particular components of reactions. The significance of Au nanoparticles as catalysts in oxidation reactions (e.g. of CO and VOCs) and in the water-gas shift is described (Chapter 23). The use of platinum (Pt) complexes in water and in micellar catalysis is also highlighted illustrating the lowering of the E-factor in various organic transformations (Chapter 24). The importance of using water for enhanced activities and selectivities at any level (chemo, regio, or enantio) is shown in a range of catalytic reactions (Chapter 25). The significance of speciation chemistry in the optimization of catalysts for the Suzuki-Miyaura coupling and towards the development of greener catalysts is shown (Chapter 26). The contribution of structure-property relationships for a better design of zeolite catalysts in Friedel Crafts acylation reactions is also treated (Chapter 27). Other important reactions catalysed by coordination compounds are also treated in previous (e.g. Chapters 9 and 10) or following (e.g. Chapters 34, 35, and 38) parts in different contexts. In Part V, recent advances in green and sustainable organocatalysis are addressed, focussed on the reduction of energy consumption, increasing efficiency and selectivity, reducing wastes, and optimizing resource use. Industrial (Roche Pharmaceutical Division) design of sustainable new

3

4

  Introduction

The advantages brought by organocatalytic reactions using immobilized catalysts for asymmetric synthesis of fine chemicals and the effectiveness of relevant organocatalysts that have been used in recent years are discussed (Chapter 29). Syntheses of long-chain aliphatic polyesters, especially by condensation polymerization (of dicarboxylic acids with diols) and acyclic diene metathesis polymerization are described, and the closed-loop chemical recycling (and upcycling) is highlighted (Chapter 30). The development of noble metal-free organic photocatalysts as potential electron sources and as an alternative to hazardous alkali metals is challenging, and an overview, combining organic photocatalysis with electrolysis in organic synthesis, is provided (Chapter 31). The combination of chiral organocatalysis with photoredox chemistry can allow the development of novel enantioselective reactions and softening of reaction conditions, and different modes of organocatalytic activation combined with photocatalysis are addressed (Chapter 32). Catalytic processes for purification of water and liquid fuels are described in Part VI. For example, water is susceptible to pollution by a large number of contaminants from various sources and the application of sustainable materials and technologies for wastewater treatment is addressed, with a focus on the photocatalytic route (Chapter 33). There has been growing concern about the environmental impact of the emission of carbon dioxide and other pollutants. The most successful desulphurization technologies involving functional materials are described (Chapters 34 and 35). Part VII concerns the formation (and storage) of hydrogen in different contexts. Examples are given of homogeneous metal-catalysed reactions with paraformaldehyde used as a source of hydrogen in water for transfer hydrogenation reactions, including the reduction of C=C, C=O, –CC–, and –CN bonds (Chapter 36). Energy storage is a shortcoming of the use of renewable fuels and the storage of chemical energy using hydrogen batteries is discussed (Chapter 37). Readily available and chemically stable storage materials, as well as solvents and catalysts, are relevant for the long-term and large-scale storage of hydrogen (Chapter 37). Finally, a brief overview is given on the replacement of fossil fuels by hydrogen as a synthetic fuel. Although water splitting for high purity H2 production is tempting, the associated high energy consumption hampers its application. The potential significance of low-cost mono- and bimetallic metalorganic frameworks/coordination polymers (MOFs/CPs) as bifunctional electrocatalysts is addressed in terms of hydrogen evolution (HER) and oxygen evolution (OER) reactions (Chapter 38).

Final Remarks Catalysis provides key tools for the development of a sustainable environment and towards the promotion of quality of life, as illustrated herein. This can be pursued by developing innovative catalytic materials and technologies to improve resource use and foster sustainable processes and products. The preparation of this book was initiated during the Covid-19 pandemic, a period favourable to a meditation assessing the interrelationship between humans and the environment. The book covers several recent catalytic studies that are aimed at improving environmental sustainability. Its contents are intended to give a vision of the challenges and future directions for the development of efficient and eco-sustainable catalytic processes. We hope that it will also serve as an inspiration and an incentive to the application of catalysis towards environmental sustainability.

5

Part I Carbon Dioxide Utilization

7

2 Transition from Fossil-C to Renewable-C (Biomass and CO2) Driven by Hybrid Catalysis Michele Aresta1,2 and Angela Dibenedetto1,2,3 1

Interuniversity Consortium on Chemical Reactivity and Catalysis, CIRCC, Via Celso Ulpiani, 27, Bari, Italy IC2R srl, Lab H124, Tecnopolis, Valenzano (BA), Italy 3 METEA and Department of Chemistry, University of Bari Aldo Moro, Via E. Orabona, 4, Bari, Italy 2

2.1  Introduction Moving to a de-fossilized economy is a must for our society, which is in need of reducing the impact of anthropogenic activities on climate to avoid a point of non-return [1–7] that may cause multiple disasters. The use of fossil-C over the last two centuries for feeding the chemical and power industries has produced a continuous release of waste heat and greenhouse gases (GHGs) into the atmosphere and CO2 has been attributed a central role in driving the climate change. The direct heating of the atmosphere (only an average 33% of the chemical energy of fossil-C is used in the conversion into electric or mechanical energy, the rest being lost to the atmosphere as heat in the temperature range 150–900+ °C) has caused an increase in the concentration of water vapor, a GHG more powerful than CO2. The two GHGs together with others such as methane, nitrogen oxides (NOx), and chlorofluorocarbons (CFCs) are reinforcing the natural greenhouse effect, contributing to an increase of the average planet temperature that should be maintained below 2 °C above the average temperature of 1990 to prevent irreversible changes to our planetary ecosystem. Since 1981, the temperature of oceans and land has increased at a rate of 0.18 °C/decade, more than doubling the increase observed during the previous century (1880–1980) of 0.08 °C/decade [8]. Such climate change is causing sudden, violent meteorological events all over the world, while the rise of the level of the oceans, caused by the transfer of water from land to the seas (melting of ice), is a serious menace for coastal areas that could be submersed [9]. In the last three decades, the capture of CO2 from point sources has been considered as a way to reduce climate change, but this has had no effect and therefore the only way to save our planet is to drastically reduce the extraction and use of fossil-C in all its forms to simultaneously lower the discharge of heat, water vapor, and GHGs to the atmosphere. Such big change requires a global agreement and action. The Conference of Parties (COP) 2016 through the Paris Agreement has provided the basis for common action that, after some initial important uncertainties, seems now to have the convinced cooperation of all major actors based on developments at the COP in Glasgow in 2021.

Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 1, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

8

2  Transition from Fossil-C to Renewable-C (Biomass and CO2 ) Driven by Hybrid Catalysis

The need is impellent for implementation of the agreement by all industrialized societies to meet the target as demonstrated by action taken to prevent the enlargement of the ozone hole, which banned over 200 ozone-depleting substances (ODS; mainly CFCs and hydro-CDCs [HCFCs]) [10]. Various policy measures have been introduced to limit or phase out the consumption of ODSs since this agreement was reached. As a result, global consumption of ODSs declined by 98.5% between 1986 and 2018, meaning that the release of 343,000 ozone-depleting potential tonnes was avoided between 1986 and 2002 [11]. With the committed participation of all countries, the damage has started to be repaired in three decates. This was possible because substitutes of CFCs were developed, meaning that the delivery of dangerous species to the atmosphere was stopped. This problem has a dimension that is not equivalent to that of fossil-C, which is larger by several orders of magnitude. However, substitutes for fossil-C must be found rapidly, even if this is not so easy considering the dimension of the problem. Time plays a key role in such change; the sooner we start finding and adopting solutions, the more quickly solutions will be implemented and start to become effective. But first, let us consider how much fossil-C we use currently to illustrate the magnitude of the problem.

2.2  The Dimension of the Problem As of 2018, our society has consumed 13,978 Mtoileq (Mtoileq means that all the energy consumed is expressed as oil burned). Table 2.1 shows that only the Confederation of Independent States (CIS) has shown an apparent decrease of energy consumption (figures in parentheses) over the period of 1990–2018, during which time CIS varied in composition. However, an increase of 20% in energy consumption is observed for the CIS from 2000 to 2018. Whereas North America, Europe, and Japan were the major consumers of energy until the 2000s, the highest increase since that time has been observed for developing countries with India, China, the Middle East, and Africa leading the world in growth of energy consumption. Table 2.2 shows the contribution of various forms of fossil-C to the overall energy budget by region or country. It is obvious that solving the energy source problem is much more complex than substituting ODSs and an efficient solution will be effective only if the following conditions are met: ● ●

Global cooperation is implemented. There is global agreement upon an effective defossilization of the energy and chemical sectors.

If there is random national engagement in limiting the causes of climate change, the effects will be very limited or zero.

2.3  Substitutes for Fossil-C How and where will it be possible to find substitutes for fossil-C? The main targets involve the use of perennial energy sources such as solar, wind, hydro, and geothermal (SWHG) as well as biomass. The latter is considered to be renewable carbon, but alone cannot cover the parts of applications that cannot be decarbonized. Whereas we can imagine decarbonizing the energy sector, it will not be possible to decarbonize the chemical industry, the polymer industry, and part of the fuel sector as our current way of life is based on carbon. However, biomass alone cannot cover the need of carbon-based goods and fuels. As a consequence, an additional source of carbon will be necessary. This is CO2, the most abundant

3770

Europe

538

168

Mt/y COAL

Country

Bcm LNG

Mt/y OIL

349

787

Asia

982

India

624

USA

217

Germany

935

976

North America

234

Russia

186

South Africa

432

243

Latin America

189

Japan

3,770

Europe

538

168

Mt/y COAL

Country

Bcm LNG

Mt/y OIL

Bcm: billion cubic meters.

China

Country

349

787

Asia

982

India

624

USA

217

Germany

935

976

North America

234

Russia

186

South Africa

432

243

Latin America

189

Japan

Table 2.2  Yearly production of coal, gas, and oil in the various parts of the world (2018).

Bcm: billion cubic meters.

China

Country

150

129

Na

50

Pacific

Na

50

1496

539

125

Turkey

1496

539

Middle East

Poland

129

125

Turkey

Middle East

Poland

Pacific

South Korea

150

South Korea

Table 2.1  Total energy consumption as Mtoilequivalent by region or country.and its segmentation by countries and regions.

691

671

CIS

113

Australia

691

671

CIS

113

Australia

398

151

Africa

109

Indonesia

398

151

Africa

109

Indonesia

10

2  Transition from Fossil-C to Renewable-C (Biomass and CO2 ) Driven by Hybrid Catalysis

source of carbon we have at hand (ca. 830 GtC are available in the atmosphere). The wise use of perennial energies (SWHG) and renewable-C (biomass and CO2) will sustain our society development in future years. The former will represent the inexhaustible reserve of primary energy, whereas the second will represent the source of carbon for dedicated uses. We do not believe that fossil-C will go down to zero in near or far future, but we believe that its use will be sensibly reduced (perhaps to one-fourth of actual by 2050) with a greatly beneficial impact on our environment, even if the decrease is not very rapid. In 2020, SWHG covered approximately 29% of the total amount of electric energy consumed in the world, representing a 2 point increase with respect to 2019 (27%). During this time, the use of other fuels decreased. Bioenergy use in industry grew 3%, but the lower oil demand due to the Covid-19 pandemic caused a decline in biofuels used to blend oil-derived fuels. In 2021, electricity from perennial sources was set to expand by more than 8% to reach 8 300 TWh, the fastest year-on-year growth since the 1970s. China alone should account for almost half of the global increase, followed by the United States of America (USA), the European Union (EU), and India. Photovoltaic (PV) and wind are set to contribute two-thirds of such growth [12]. The perennial sources will deliver electrons [13] and, thus, will contribute to electricity production. But electricity alone will not solve all problems of the human society. Electrons will be distributed through dedicated lines to industries, cities, public buildings, and private houses. They will be also used in applications such as electrified transport (trains) or city-aerial electric buses and even in cars using batteries. Electrons will be used to produce H2 from water (electrolysis). And hydrogen can be used in transport (H2-fuelled cars). However, some important transport sectors (such as maritime transport and aviation) and some industrial sectors (fine chemicals, polymers, goods used daily, and similar products) will remain out of reach. Therefore, one can foresee that our society will use a blend of energy sources and vectors by 2050 in which fossil-C will have a decreased share. Our forecast is that fossil-C will decrease from actual 81% of global consumption to perhaps 20%. This will cause the decrease of direct fossil-CO2 emission from actual 37 Gt/y to 8–9 Gt/y. Various scenarios can be found in the literature and not all agree on the future role of fossil-C. Claims range from a total defossilization to a continued use of fossil-C, but at a reduced rate. Scheme 2.1 shows the transformation of energy consumption expected in coming years as illustrated by the U.S. Energy Information Association (EIA-USA) [14], including the roles of perennial and renewable energy sources. The major contribution to the growth of use of energy and goods will be given by developing countries in which economic growth will drive the demand for energy. Therefore, Asia (China and India), the Middle East, and Africa will be major actors. One can expect that different sources will be exploited in different regions according to the local reserves and availability of SWHG energy. Perennial energy sources should be preferentially exploited everywhere for reducing climate impacts. Despite some very optimistic scenarios that foresee zero emissions, the EIA scenarios show that fossil-C will still be significantly in use even in 2050. Therefore, a realistic scenario is depicted in Scheme 2.2 with the use of various sources of energy and vectors by application. As shown in Scheme 2.2, the intensity of use of fossil-C will decrease in all sectors and specially in the production of thermal energy, chemicals for industry, special fuels, and materials. The field of major interest for this chapter is the change of raw materials in the chemical industry, special fuels, and materials, where the big shift will take place as depicted in Scheme 2.3. This shift will require new catalysts able to convert raw materials richer in oxygen and innovative technologies [15]. For this purpose, hybrid catalysis will play a major role.

2.4  Hybrid Catalysis: A New World

Scheme 2.1  Expected growth of energy consumption by 2050 (a), the contribution by the various energy sources (b), and the use by selected regions worldwide (c). (EIA data).

Chemicals Fossil-C Biomass Ren-C

Electric Energy

Thermal Energy

Special Fuels

Materials

Fossil-C Biomass Ren-C SWGH

Fossil-C Biomass Ren-C SG

Fossil-C Biomass Ren-C

Fossil-C Biomass Ren-C

Industry-Civil-Land transport

Avio-Maritime

Scheme 2.2  Correlation of the intensity of use of primary sources of energy according to end users (Ren-C=Renewable-carbon; S=Solar, W=Wind, G=Geothermal, H=Hydro).

2.4  Hybrid Catalysis: A New World Innovation in catalysis will be a necessity and this will be the major driver of the change in the chemical industry. In fact, moving from fossil-C to renewable carbon (ren-C) as source of raw materials will introduce the use of substrates richer in oxygen (hydrocarbons [HCs] and synthesis gas [syngas] will be substituted by carbohydrates and CO2) (Scheme 2.3). This will result in a

11

12

2  Transition from Fossil-C to Renewable-C (Biomass and CO2 ) Driven by Hybrid Catalysis

growing demand for green-H2, unless a major change is made to move closer to nature by using water and CO2 as source of a myriad of chemicals. Such a great step will require new technologies and new catalysts. Hybrid catalysis is able to combine biotechnology and chemical processes using enzymatic- and chemo-catalysis and this will be the solution. Learning from nature and combining innovation with existing solid knowledge of chemical processes will provide a new attitude to developing new catalytic systems able to use the new substrates. Hybrid catalysis (Figure 2.1) is the integration of chemo-, electro-, and biotec(enzymatic)-processes. The integration can concern either two sectors at one time (chemo-electro catalysis, bioelectro catalysis, or bio-chemo catalysis), or even all three sectors (bio-chemo-electro catalysis). In following sections, the advantages of such integration will be discussed with some examples. The basic principle is that in the integration one technique will do what the other(s) cannot do or will do it better (more selectively or faster or with higher conversion). The overall target is to develop innovative processes based on ren-C that are more economical on all levels (atoms, energy, infrastructure, operational, raw materials, etc.) in the short or medium term with respect to fossil-C based processes that have had over one century of optimization before arriving at today’s cost levels. Fulfilling such a goal would mean reducing waste (solid-liquid) production and CO2 emission to the atmosphere. Currently, the chemical industry produces over 90% of the goods used by our society by using catalysts, and over 40% of the entire world economy depends on catalysis [16, 17]. The process integration is not trivial, due to the peculiar properties of chemo- and enzymaticsystems (or microorganisms) and the complexity of their mutual interaction (deactivation of From Hydrocarbons-HC From Carbon Monoxide

H(CH2)nH CO

to

Carbohydrates

to

Carbon Dioxide

R(HCOH)nR′ CO2

Scheme 2.3  Major changes in raw materials composition in the chemical industry as ren-C (biomass and CO2) takes the place of fossil-C and synthesis gas (syngas) derived from it) (Red=Today, Blue=Future).

Figure 2.1  Hybrid catalysis: integration of chemo-, electro-, and enzymatic catalysis.

2.5  Hybrid Catalysis and Biomass Valorization

enzymes) and interaction with electrons (energetics). The integration of enzymes (microorganisms) and chemo-catalysts offers the opportunity of combining the stereospecificity and selectivity of the former with the versatility of the latter, while taking advantage of the modular acid-base, redox, and nucleophilic-electrophilic character that will drive the interaction of the substrates. In fact, enzymes are superior to chemo-catalysts (homogeneous catalysts, essentially) in the synthesis of optically active compounds with high selectivity towards one of the isomers. Although the use of asymmetric ligands makes homogeneous catalysts prone to the production of a high excess of one of the isomers, nevertheless enzymes are much more effective due to their structure. How easy will it be and how long will it take to realize such integration at the application level, with a simultaneous reduction of investment costs (CAPEX), reduction in operational costs (OPEX), and increase in selectivity and rate of production? The timing is not exactly predictable, but it is time now to invest resources (personnel and financial) in this field. It must be pointed out that a real hybrid catalyst should act in a single pot. So far, examples of combinations of catalytic stages were used in which two different catalytic systems act in two separate, consequent stages, reaching a result that each alone would not be able to touch. These are cases of combined more than integrated catalytic systems. Notably, hybrid catalysis has mostly been applied to the conversion of bio-sourced molecules since its appearance, which is more complex than steps used in chemical processes and, this is increased to reduce the impact of chemo-catalysis, even due to the spent catalyst disposal and recovery. The reader will not be surprised if the examples discussed in the next section are based on the conversion of bio-sourced substrates. In following paragraphs, the state of the art (SotA) and perspective applications will be discussed, highlighting the power and potential of hybrid catalysis.

2.5  Hybrid Catalysis and Biomass Valorization Land biomass is in general a solid material made of a variety of single molecules and linkages. It can be classified by large as cellulosic and oily biomass, the former formed by cellulose (35–50%), hemicellulose (20–30%), and lignin (10–25%) and the latter formed by long chain fatty esters of glycerol. Sugars, amines, aminoacids, organic aromatic and aliphatic moieties, polymeric species, esters (with long- and short-chain acids), and ethers (aliphatic and aromatic) are among the species most frequently present in various kinds of biomass. This means that the raw-biomass conversion implies the interaction of the catalyst with a variety of linkages, such as: C–C, C–O, C–N, C–H, O–H, N–H, C–S, S–O, S–N, P–C, P–O, O–E, and C–E (where E is an element different from C, H, N, O, P, S). Chemical catalysts are designed to be quite specialist and may act on a specific bond, leaving the others unaltered. Chemo-catalysts (homogeneous, supported, and heterogeneous) have been developed and are used in the chemical industry primarily for carrying out a defined linkage cleavage/formation targeting selectivity in processes such as: the conversion/valorization of hydrocarbons to bulk or fine chemicals, the conversion of syngas to Cn species (Fischer-Tropsch reaction), the hydrogenation of/addition to unsaturated C–C (double and triple) bonds, and C–C coupling, working in a liquid or gaseous phase. Chemo-catalysts primarily act singularly on individual molecules and are less able to attack compact structures. The use of multifunctional catalysts or assembled catalysts might guarantee a concerted action on the complex system, but the effect could reduce selectivity and result in the formation of a variety of products. A recent example of such catalytic activity on solid systems is plastics depolymerization [18], an infrequent and difficult process.

13

14

2  Transition from Fossil-C to Renewable-C (Biomass and CO2 ) Driven by Hybrid Catalysis

Direct biomass conversion using chemocatalysis is not a common process either, being limited to the basic thermal depolymerization reaction carried out under alkaline conditions at high temperature [19]. Conversely, enzymes and microorganisms can operate in consortia on biomass by being able to act, in parallel or in cascade, on the various linkages and to convert biomass at low temperatures (even room temperature) into simple monomers (sugars) or their derivatives [20]. When solid biomass is used with microorganisms, a preliminary depolymerization of the biomass may be necessary in order to produce entities that can interact effectively with the internal of the cell where enzymes are present. The use of microorganisms often results to be more effective than the use of enzymes especially in the case the latter require a cofactor that is regenerated in the whole cell but requires external action when free enzymes are used. Sometimes, cell membranes, where cofactors are located, are used with free enzymes to allow the regeneration of cofactors. The combination of enzymes and microorganisms with chemo-catalytic systems into a hybrid system can result in a reinforced action as one could do what the other is not able to do. Attempts to use hybrid catalysis started in the first decade of 2000s. It is a recent science with less than 100 examples of application to a variety of systems. Herein, some cases will be discussed and commented upon. There are still some different ways of defining a hybrid catalytic system. As mentioned above, three different systems are combined by two or else they can act together: chemo-catalysis, electrocatalysis, and bio-catalysis. At the start, a simple combination of two steps-two pots (one chemical step and one based on biosystems or even on electrochemistry, or else an electrochemical and a bioprocess) was defined to be a hybrid catalytic system. Strictly speaking, a hybrid catalyst should act in a single pot, using all actors in one step. This is not trivial as chemicals can affect enzymes, as already mentioned. To avoid such negative influence, enzymes and chemical species should be supported in such a way that they do not get in touch and can maintain their efficiency. It must be said that even the simple coupling of chemo-catalysts and enzymes affords interesting results when dealing with optically active products or products having a complex structural architecture. However, the ultimate target of hybrid catalysis is to perform in one pot and possibly in one step a complex sequence of catalytic steps. Let us see where we are and where we can go, by showing a few selected examples, even making the comparison with a single technology (chemical, biotech, or electrochemistry). As a first case, we wish to compare the chemical and hybrid route to produce 5-hydroxymethylfurfural (5-HMF), a platform molecule used for the synthesis of a number of different chemicals. Figure 2.2 compares the chemical route [21–23] for the conversion of glucose into 5-HMF to the hybrid route [24] that integrates enzymes and chemo-catalysts. The chemical conversion of glucose into 5-HMF is a two-step process that needs in a first step a basic catalyst (in our proprietary process, we use a mixed oxide with basic properties) for the isomerization of glucose into fructose and in the second step an acid catalyst (an acid mixed oxide) for the dehydration of fructose into 5-HMF. The two steps can be combined into a single reactor, as shown in Figure 2.2, but remain well separated because of the different reaction conditions. To increase the production of 5-HMF an extraction using dimethylcarbonate has been used [23] that allows an interesting conversion yield (c. 68%). It is worth noting that this technique allows DMC to be recycled by evaporation under vacuo so that 5-HMF can be recovered as a pure white solid. Altogether, a conversion of glucose to 5-HMF equal to 53% is reached, working in the temperature interval 80–170 °C across the two steps. The hybrid route still requires two pots because of the different pH required by each of the two steps. The working temperature can be as high as 70 °C and a conversion of 88% of glucose was observed with a 31% of 5-HMF. When milder conditions are used, a lower overall yield is obtained. Such a hybrid technique (indeed, it is not a real hybrid

2.5  Hybrid Catalysis and Biomass Valorization

Figure 2.2  Conversion of glucose into 5-hydroxymethylfurfural (5-HMF). Chemo-catalytic vs hybrid process. The methyl isobutyl ketone (MIBK) phase and the adduct with boronic acid allow the transfer of fructose from the basic to the acid compartment with high efficiency.

system, but more a combined bio-chemo catalytic system) has been applied to several other reactive systems [25–28] with interesting results. Another example of glucose conversion into an added-value product is the chemo-enzymatic synthesis of D-mannitol. In this case, glucose is first isomerized enzymatically into fructose (see Figure 2.2) and the latter is then hydrogenated to D-mannitol with 46% yield and 92% ee [29, 30]. The combination of enzymes and chemo-catalysts has a key role when optically active compounds are synthesized due to the highly specific activity of enzymes. As mentioned above, whole cells can also be used in combination with metal systems. One of the major benefits of this approach is that co-factors required by some enzymes (typical of enzymes involved in redox reactions) can easily be regenerated in the whole cell. This has considerable advantage over finding other approaches to regenerate co-factors. This is the case when NADH is oxidized to NAD+. It is not trivial to reduce the reactant back to the active 1,4-isomer, as either the 1,6-isomer is formed or dimers (neither of which is active). Another advantage of using the whole cell is that this allows the costly enzyme extraction-purification step to be avoided [31–39]. An interesting case is the synthesis of catechols that have applications in diversified fields such as coatings, antioxidants and cosmetics, aromas, and antiseptics [40] and for which hybrid systems have been compared to pure chemical or biotechnological processes. Chemical processes are complex and require drastic conditions, affording low yields and organic waste. A fully enzymatic process, based on the use of a tolune di-oxygenase (TDO) and a cis-hydrol dehydrogenase (DHDH), also has a poor yield as the latter inhibits TDO. A hybrid system that couples TDO to a chemo-hydrogenation using Pd/C has been shown to work much better with yields of catechol higher than 50%. This is an interesting case study because of the multiple cross-influences of the chemo- and bio-systems and because it combines two antithetical processes: an oxidation (TOD) and a reduction (hydrogenation with Pd/C).

15

16

2  Transition from Fossil-C to Renewable-C (Biomass and CO2 ) Driven by Hybrid Catalysis

Figure 2.3  Conversion of bioglycerol into trimethylene carbonate (TMC) using hybrid catalysis [41].

Another interesting example of the benefits of coupling biotechnology and catalysis is the conversion of glycerol into propene carbonate (Figure 2.3). In this case, a one-pot synthesis was not possible because some components of the biological system strongly affected the catalysts. The examples discussed above show that it is possible to build and use such integrated biotech chemo systems, the most difficult step is to keep them working within the same pot or reaction environment. They also show the locations of bottlenecks and barriers to work around. One of the key issues is that chemicals can deactivate enzymes. When heterogeneous catalysts are used, bio-components (proteins and others) can cover the catalysts surface, reducing or nullifying their activity. When homogeneous catalysts are used, the metal center can be affected by the coordination of bio-molecules that prevents its activity. A solution can be the heterogenization of the enzymes or of the catalysts that can be supported on a stable matrix so that they do not touch. Nevertheless, biomolecules present in the reaction medium can always affect the catalysts. In fact, such an appealing field of catalysis requires thought to make it work and many trials are needed to build the best reactive system.

2.6  Hybrid Catalysis and CO2 Conversion Within a circular economy framework, the utilization of CO2 as building block for chemicals or materials or else as source of carbon for fuels has a key role. In fact, carbon capture and utilization (CCU) is part of several national and international programmes, and the carbon circular economy (CCE) is more and more frequent in research and industrial perspectives. Notably, the use of CO2 as a building block or alternatively as a source of carbon for fuels requires quite different energy inputs with the latter also requiring hydrogen as a co-reagent (Figure 2.4). The two classes of reactions will be discussed separately in the following two sections, in part because they have received quite different attention for the development of hybrid-catalysis approaches.

2.6.1  CO2 as Building Block Only very few examples of hybrid-catalysis applied to the conversion of CO2 into chemicals have been reported. It must be mentioned that sometimes the term hybrid has been given a different meaning to the one used here. For example, the use of metallorganic catalysts grafted on silica or

2.6 Hybrid Catalysis and CO2 Conversion

Figure 2.4  Uses of CO2 in chemical transformations. Processes shown above the dashed line require moderate (or zero) energy input and can be performed even in an energy system based on fossil-C. Processes shown below the dashed line are processes that are energy intensive and in addition demand hydrogen.

supported on other substrates has been defined as hybrid catalysis [42]. We consider such catalytic systems as heterogenized catalysts more than as hybrid systems, which should couple two or three components involving chemo-, electro-, or bio-catalysts. Sometimes the hybrid catalysts are named differently, such as abiotic–biological hybrid systems [43]. A recent review of carboxylation reactions can be found in Refs [44, 45]. One of the earliest applications of hybrid catalysis is represented by the electro-catalytic incorporation of CO2 onto organic substrates with the use of soluble metal-catalysts. In the early 1990s, Dunach and Perichon investigated this field in detail with the carboxylation of several organic substrates [46]. Conversely, an early example of coupling a chemo- and an enzymatic-system is represented by the carboxylation of phenol to yield p-hydroxybenzoic acid (pHBA) (Figure 2.5) [47, 48]. p-HBA is a monomer for polymers that have peculiar application in the production of materials used for specialistic applications. The hybrid catalytic system implies the conversion of phenol into the phosphate derivative (step 1 in Figure 2.5), followed by the enzymatic carboxylation using a phenol carboxylase enzyme extracted from Thauera aromatica. In one case (right), the reaction occurs at ambient temperature and pressure using enzymes encapsulated in a low-melting agar. In the second case, a cut-off membrane was used that avoids enzyme loss. The reaction occurs in a CO2 atmosphere and a solution of phenyl-phosphate is slowly and continuously moved over the enzymes. Both the chemo- and the enzymatic-steps are carried out in the same pot and, therefore, such a process can be considered as a one-pot two-step process. The process is very selective and only the para-isomer is produced. Interestingly, the reaction can also occur in supercritical conditions using SC–CO2 as solvent and reagent [49].

2.6.2  CO2 Conversion to Value-added Chemical and Fuels via Hybrid Systems The requirements for a sustainable conversion of carbon dioxide to value-added chemical products are becoming of technological and societal importance [50, 51]. For example, methanol, a potential

17

18

2  Transition from Fossil-C to Renewable-C (Biomass and CO2 ) Driven by Hybrid Catalysis OPO32–

OH

OH

P2O5

ENZYME

NaOH

CO2

TON = 1600 TOF = 60 COOH

Even in SC-CO2!

REAGENT + COSUBSTRATES + ENZYME

Reagent + Cosubstrates + Enzymes

Membrane reactor

Inglobated into Low-melting agar Product

PRODUCT

Figure 2.5  Carboxylation of phenol using hybrid catalysis. In the first step, a chemical reaction converts phenol into phenol-phosphate. In the second step the enzyme phenol carboxylase, extracted from Thaurea aromatica, performs the carboxylation. This is a two step-one pot hybrid catalysis [48] / Elsevier.

platform molecules because of its available applications in the fields of fuels and chemicals in the future, can be obtained by using an hybrid systems (enzymatic/photocatalytic) in water. The process, carried out at room temperature, is driven by a pool of dehydrogenase (DH) enzymes [52–56] that use NADH as a cofactor. ate DH ald DH CO2 + 3 NADH F → F → ADH  → CH3OH + 3 NAD+

(2.1)

As depicted in Eq. 2.1, three mol of NADH (6e-) are consumed to produce one mol of methanol. So, the multi-enzymatic cascade reaction can be improved at large scale only if the cofactor is completely regenerated. Ru-modified-ZnS has been used as photocatalyst coupled to bioglycerol as reducing agent [53], or to water/bioglycerol as H-donor, and Rh(III)-complex has been used as an e−–H+ transfer [54, 55] for increasing the CH3OH/NADH molar ratio. Moreover, it has been reported that coupling microorganisms (or enzymes) with a photochemical system can generate electrons for NAD+ reduction [56]. The approaches used to CO2 reduction driven by electrical and/or solar inputs using different catalytic systems (homogeneous [57], heterogeneous [58], and enzymatic [59]) highlight some aspects in this area to be considered, such as: ● ● ● ●

The chemoselective conversion of CO2 to a single product. The minimization of the competitive reduction of protons to hydrogen. The long-term stability under environmentally friendly aqueous conditions. The fact that unassisted light-driven CO2 reduction does not require external electrical bias and/ or sacrificial chemical quenchers.

Certainly, the homogeneous and heterogeneous catalysts used for CO2 conversion have two drawbacks: the product selectivity (not very high) and sensitivity to an aqueous solution. On the other hand, enzymes are characterized by high specificity but are not very resistant under reaction conditions.

2.6 Hybrid Catalysis and CO2 Conversion

The establishment of a hybrid bio-inorganic system for solar-to-chemical conversion of CO2 and water to value added chemicals is very appealing. Bioelectrochemical systems (BESs) are unique systems capable of converting chemical energy into electrical energy (and vice-versa) while employing microbes as catalysts. BESs include microbial fuel cells (MFCs), microbial electrolysis cells (MECs), microbial electrosynthesis systems (MES), and other components [19]. MES has been specifically designed for cathodic CO2 conversion and can utilize CO2 as substrate and convert it into chemicals (organic acids, alcohols) and materials using electroactive microorganisms (Figure 2.6) [60]. The electroactive microorganisms can utilize electrons from a cathode or photosensitizer to reduce CO2 to multi-carbon products with low energy input and high transformation efficiency [61] Interestingly, the cathode can be produced by using biocompatible material such as copper foam coated with reduced graphene oxide. The electroactive biofilms Sporomusa ovata, deposited on graphene oxide-coated copper foam cathode, is able to produce acetate from CO2 (Figure 2.7) [62].

+

–

CO2

Electrotrophic community

MES

Short chain carboxylic acids

e–

e– O2

H+ CEM

e–

e+

H

CO2 +

Planktonic community (homoacetogenic, solventogenic and chain elongating microorganisms)

H2O e–

e–

H2

Short and medium chain carboxylic acids/alcohols

H+

Alcohols ANODE

BIOCATHODE

Organic acids Bioplastics

Figure 2.6  Microbial electrosynthesis system (MES) for CO2 conversion [60] / Elsevier / CC BY-4.0.

V

e–

e–

H2 Reduced graphene oxide

4H+ + O2

e–

e–

S. ovata

e– H+

H+

e–

CO2

S. ovata S. ovata S. ovata

2H2O

e–

e–

S. ovata

Acetate

Anode PEM Reduced graphene oxidecoated copper foam cathode

Figure 2.7  Microbial electrosynthesis system (MES) with a composite cathode [62] / Elsevier.

19

2  Transition from Fossil-C to Renewable-C (Biomass and CO2 ) Driven by Hybrid Catalysis

Microorganisms able to harvest electrons directly from the cathode can be classified as acetogens or methanogens and these include Methanobacterium palustre, Sporomusa sphaeroides, Sporomusa silvacetica, Clostridium ljungdahlii, Clostridium aceticum, and Moorella thermoacetica [63, 64]. Conversely, other microorganisms can transfer electrons indirectly, without contact with inorganic catalysts or the cathode. Three main indirect electron transfer pathways can be considered: redox mediators, single-carbon compounds, and H2 (Moorella thermoacetica or Clostridium formicoaceticum, Clostridium acetobutylicum, Nitrosomonas europaea [65–67]). Most autotrophic microorganisms, such as Methanosarcina barkeri [68] and Methanococcus maripaludis [69], use intermediates via H2 to reduce CO2 to CH4 through the Calvin-Benson pathway. For example, by using platinum or an earth-abundant substitute, α-NiS, as a biocompatible hydrogen evolution reaction (HER) electrocatalyst, and Methanosarcina barkeri as a biocatalyst for CO2 fixation, a robust and efficient electrochemical CO2 conversion to CH4 (up to 86% overall Faradaic efficiency for ≥7 d) has been demonstrated [68]. By using this system, hydrogen generated in situ at the cathode reacts with carbon dioxide to afford methane in presence of the M. barkeri biocatalyst.

(a)

V

e–

(b)

e– CH4

H2O

V

e–

e– CH4

H2O

CH4 (CnHn)

CH4 (CnHn)

H+ Biogas 2H+ +O2

CO2

Anode

Cathode

Anode

(c)

e–

HCO3–

e– CH4 (CnHn)

CO1

CH4

Cathode

Biogas

Cathode IEM Anode IEM Cathode

CO2 Anode

e–

O2 HCO2

CO2

CH4 Biogas

COD

OH–

NH4–

H

+

HCO3–

(CnHn)

CO1

V e–

COD

CH4 H+

Biogas

IEM

Biogas

(d) V

e– CH4

CO2

2H+ +O2

NH4–

20

H– Na+

Cl23 Cl–

CO1 CO2 Cl– Cl–

H2O2

Na+ Regeneration Absorption

Cathode

Figure 2.8  Different types of Microbial electrosynthesis system (MES) for biogas upgrading. a) Monodimensional system. b) Two-dimensional system. c) Triple compartments configuration with anode, cathode, and regenerative unit (IEM: ion exchange membrane). d) Four compartments configuration with anode, regeneration, absorption and cathode compartment [70] / Elsevier / CC BY-4.0.

References

Recently, BES systems have been applied to biogas upgrading. Multidimensional electrode materials have shown superior performance in supplying the electrons for the reduction of CO2 to CH4. Most of the studies on the biogas upgrading process include hydrogen (H2) mediated electron transfer mechanisms in BES biogas upgrading. [Figure 2.8] [70] Of interest is also the photosynthetic biohybrid system proposed by Liu et al. [71] showing a silicon nanowire array coupled with anaerobic bacterium, S. ovata, that is able to use the photoinduced electrons to reduce CO2 to acetate (6 g/L, faradaic efficiency of up to 90%.). The latter can be then used as substrate and further converted into chemicals by using a different engineered bacterial strain (Escherichia coli).

2.7  Conclusions Hybrid catalysis is a recent discipline that combines biotechnology and chemical (electrochemical) catalysis for the conversion of the non-fossil raw materials such as CO2 and biomass-derived platform molecules into added value chemicals or energy products. Much has still to be discovered for defining its real potential, but early examples clearly show that it can contribute to reducing the use of fossil-C and to mitigating environmental impact, even using waste as raw materials. Competent integration of these approaches are essential for its rapid development and for making hybrid catalysis a main actor in the production system in both the chemical and energy sectors.

References 1 Aresta, M. and Dibenedetto, A. (2004). Catal. Today 98: 455. 2 Aresta, M. and Dibenedetto, A. (2007). Dalton Trans. 2975. 3 Aresta, M., Dibenedetto, A., and Angelini, A. (2013). J. CO2 Util. 3-4: 65. 4 Aresta, M., Dibenedetto, A., and Angelini, A. (2014). Synthesis of organic carbonates, Chapter 2. In: Advances in Inorganic Chemistry: CO2 Chemistry. 66 (ed. M. Aresta and R. Van Eldik), 25–81. Cambridge: Academic Press. 5 Nocito, F. and Dibenedetto, A. (2020). Curr. Opin. Green Sustain. Chem. 21: 34–43. 6 Dibenedetto, A. and Nocito, F. (2020). ChemSusChem. 13: 6219–6228. 7 Aresta, M. and Dibenedetto, A. (2021). The Carbon Dioxide Revolution. Amsterdam: Springer. 8 Lindsey, R. and Dahlman, L. (2021). Climate change: global temperature. NOAA’s 2020 Annual Climate Report, March 15. https://www.climate.gov/news-features/understanding-climate/ climate-change-global-temperature (accessed 30 May 2023). 9 Aresta, M. and Dibenedetto, A. (2021). The atmosphere, the natural cycles, and the greenhouse effect, Chapter 3. In: The Carbon Dioxide Revolution (ed. M. Aresta and A. Dibenedetto), 31–43. Amsterdam: Springer. 10 Montreal Protocol (1987). 11 European Environment Agency. Consumption of ozone-depleting substances. https://www.eea. europa.eu/ims/consumption-of-ozone-depleting-substances (accessed 30 May 2023). 12 International Energy Agency. https://www.iea.org/reports/renewables-2022 (accessed 30 May 2023). 13 Aresta, M. and Dibenedetto, A. (2021). The alternative, carbon-free primary energy sources and relevant technologies, Chapter 5. In: The Carbon Dioxide Revolution (ed. M. Aresta and A. Dibenedetto), 6100000–72. Amsterdam: Springer.

21

22

2  Transition from Fossil-C to Renewable-C (Biomass and CO2 ) Driven by Hybrid Catalysis

14 Midcontinent Independent Systems Operator, Inc. (2021). MISO Futures Report. https://cdn. misoenergy.org/MISO%20Futures%20Report538224.pdf. (accessed 30 May 2023). 15 Aresta, M., Bocarsly, A.B., and Dibenedetto, A.E. (eds.) (2022). Beyond Current Research Trends in CO2 Utilization. Lausanne: Frontiers Media SA. doi: 10.3389/978-2-88974-601-9. 16 Aresta, M., Dibenedetto, A., and Dumeignil, F. (2015). Catalysis, growth and society, CCh. 1. In: Biorefineries: An Introduction (ed. M. Aresta, A. Dibenedetto, and F. Dumeignil), 5–12. Berlin: DeGruyter. 17 Aresta, M., Dibenedetto, A., and Dumeignil, F. (2021). Transition from the linear to the circular economy. The role of biorefinery and catalysis, Ch. 1. In: Biorefinery: From Biomass to Chemicals and Fuels. Towards Circular Economy, 2e (ed. M. Aresta, A. Dibenedetto, and F. Dumeignil), 7–24. Berlin: De Gruyter. 18 Quaranta, E., Fini, P., Nocito, F., and Dibenedetto, A. (2021). J. Hazard. Mater. 403: 123957. 19 Heuson, E. and Dibenedetto, A. (2021). Hybrid catalysis: bridging two words to create energy, Ch. 15. In: Biorefinery: From Biomass to Chemicals and Fuels. Towards Circular Economy, 2e (ed. M. Aresta, A. Dibenedetto, and F. Dumeignil), 493–536. Berlin: De Gruyter. 20 Blair, E.M., Dickson, K.L., and O’Malley, M.A. (2021). Curr. Opin. Microbiol. 64: 100–108. 21 Souzanchi, S., Nazari, L., Rao, K.T.V. et al. (2019). Catal. Today 319: 76–83. 22 Dibenedetto, A., Aresta, M., Pastore, C. et al. (2015). RSC Adv. 5: 26941. 23 Dibenedetto, A., Aresta, M., di Bitonto, L., and Pastore, C. (2016). ChemSusChem. 9 (1): 118–125. 24 Gimbernat, A., Guehl, M., Lopes Ferreira, N. et al. (2018). Catalysts 8: 335. 25 Pesci, L., Baydar, M., Glueck, S., et al. (2017). Org. Process. Res. Dev. 21: 85–93. 26 Shahidi, F. and Naczk, M. (2003). Phenolics in Food and Nutraceuticals. Boca Raton: CRC Press. 27 Straathof, A.J.J. (2003). Biotechnol. Bioeng. 83: 371–375. 28 Ballesteros, A., Bornscheuer, U., Capewell, A. et al. (1995). Biocatal. Biotransform. 13: 1–42. 29 Kieboom, T. (2007). Integration of biocatalysis with chemocatalysis: cascade catalysis and multi‐step conversions. In: Concert Catalysis for Renewables, 2007 (ed. G. Centi and R.A. Van Santen), 273–297. New York City: Wiley. 30 Makkee, M., Kieboom, A.P.G., Bekkum, H.V., and Roels, J.A. (1980). J. Chem. Soc. Chem. Commun. 19: 930–931. 31 Yang, D., Ma, C., Peng, B. et al. (2020). Ind. Crops. Prod. 153: 112580. 32 Zhang, P., Liao, X., Ma, C. et al. (2019). ACS Sustain. Chem. Eng. 7: 17636–17642. 33 Qin, L.-Z. and He, Y.-C. (2020). Appl. Biochem. Biotechnol. 190: 1289–1303. 34 Huang, Y., Liao, X., Deng, Y., and He, Y. (2019). Catal. Commun. 120: 38–41. 35 Xue, -X.-X., Ma, C.-L., Di, J.-H. et al. (2018). Bioresour. Technol. 268: 292–299. 36 Di, J., Ma, C., Qian, J. et al. (2018). Bioresour. Technol. 262: 52–58. 37 He, Y., Ding, Y., Ma, C. et al. (2017). Green Chem. 19: 3844–3850. 38 He, Y.-C., Jiang, C.-X., Jiang, J.-W. et al. (2017). Bioresour. Technol. 238: 698–705. 39 He, Y.-C., Jiang, C.-X., Chong, G.-G. et al. (2017). Bioresour. Technol. 245: 841–849. 40 Berberian, V., Allen, C.C.R., Sharma, N.D. et al. (2007). AdV. Synth. Catal. 349: 727–739. 41 Dibenedetto, A., Aresta, M., Di Bitonto, L., and Dubois, J.L. (2013). Patent EP 13192912.7. 42 Calabrese, C., Giacalone, F., and Aprile, C. (2019). Catalysts 9: 325. 43 Li, J., Tian, Y., Zhou, Y. et al. (2020). Trans. Tianjin Univ. 26: 237–247. 44 Tommasi, I. (2017). Catalysts 7: 380. 45 Nocito, F. and Dibenedetto, A. (2022). Carboxylation reaction based on the direct and indirect use of CO2: sustainable syntheses of C-CO2, O-CO2 and N-CO2 bonds. In: Industrial Arene Chemistry: Markets, Technologies, Processes and Case Studies of Aromatic Commodities (ed. J. Mortier), 807–847. New York City: Wiley. ISBN 9783527347841.

References

46 Derien, S., Clinet, J.C., Dunach, E., and Perichon, J. (1993). J. Organ. Chem. 58 (9): 2578–2588 (and other papers by the same authors). 47 Aresta, M., Quaranta, E., Liberio, R. et al. (1998). Tetrahedron 54 (30): 8841–8846. 48 Aresta, M. and Dibenedetto, A. (2002). Rev. Mol. Biotechnol. 90 (2): 113–128. 49 Dibenedetto, A., Lo Noce, R., Pastore, C. et al. (2006). Env. Chem. Lett. 3 (4): 145. 50 Lewis, N.S. and Nocera, D.G. (2006). Proc. Natl. Acad. Sci. USA 103: 15729–15735. 51 Armaroli, N., Artero, V., Centi, G. et al. (2016). Solar driven chemistry. EUCheMS white paper. 52 Obert, R. and Dave, B.C. (1999). J. Am. Chem. Soc. 121: 12192–12194. 53 Dibenedetto, A., Stufano, P., Macyk, W. et al. (2012). ChemSuschem. 5 (2): 373. 54 Aresta, M., Dibenedetto, A., Baran, T. et al. (2014). Beilstein J. Org. Chem. 10 (1): 2556–2565. 55 Aresta, M., Dibenedetto, A., and Macyk, W. (2015). Hybrid (enzymatic and photocatalytic) systems for CO2-water co-processing to afford energy rich molecules. In: From Molecules to MaterialsPathways to Artificial Photosynthesis (ed. E.A. Rozhkova and K. Ariga), 113–169. Amsterdam: Springer. 56 Schlager, S., Dibenedetto, A., Aresta, M. et al. (2017). Energy Technol. 5: 1. 57 Benson, E.E., Kubiak, C.P., Sathrum, A.J., and Smieja, J.M. (2009). Chem. Soc. Rev. 38: 89–99. 58 Zhu, W., Michalsky, R., Metin, O. et al. (2013). J. Am. Chem. Soc. 135: 16833–16836. 59 Parkinson, B.A. and Weaver, P.F. (1984). Nature 309: 148–149. 60 Dessì, P., Rovira-Alsina, L., Sánchez, C. et al. (2021). Biotechnol. Adv. 46: 107675. 61 Zhu, H., Dong, Z., Huang, Q. et al. (2019). RSC Adv. 9: 34095–34101. 62 Aryal, N., Wan, L., Overgaard, M.H. et al. (2019). Bioelectrochemistry 128: 83–93. 63 Cheng, S., Xing, D., Call, D.F., and Logan, B.E. (2009). Environ. Sci. Technol. 43: 3953–3958. 64 Nevin, K.P., Hensley, S.A., Franks, A.E. et al. (2011). Appl. Environ. Microbiol. 77: 2882–2886. 65 Song, J., Kim, Y., Lim, M. et al. (2011). ChemSusChem. 4: 587–590. 66 Jeon, B.Y., Jung, I.L., and Park, D.H. (2012). Adv. Microbiol. 02: 332–339. 67 Khunjar, W.O., Sahin, A., West, A.C. et al. (2012). PLoS ONE 7: e44846. 68 Nichols, E.M., Gallagher, J.J., Yude Su, C.L. et al. (2015). Proc. Natl. Acad. Sci. USA 112 (37): 11461–11466. 69 Kracke, F., Wong, A.B., Maegaard, K. et al. (2019). Nat. Commun. 2: 45. 70 Aryal, N., Zhang, Y., Bajracharya, S. et al. (2022). Chemosphere 291: 132843. 71 Liu, C., Gallagher, J.J., Sakimoto, K.K. et al. (2015). Nano Lett. 15 (5): 3634–3639.

23

25

3 Synthesis of Acetic Acid Using Carbon Dioxide Philippe Kalck University of Toulouse UPS-INP (France), Laboratoire de Chimie de Coordination du CNRS UPR 8241, Composante ENSIACET de l’Institut National Polytechnique de Toulouse, 4 allée Emile Monso, TOULOUSE Cedex 4, France

3.1  Introduction Ethanoic acid, or acetic acid, is produced at 85% by methanol carbonylation. In 2020, approximately 18 million tons was produced, and its annual growth is approximately 5% [1]. Methanol, which had a production capacity of 110 million metric tons in 2021 [2], is one of the first building blocks in a wide variety of synthetic products and is also used as a fuel additive [3]. The main use of acetic acid is related to its transformation in vinyl acetate monomer (ethyl-, propyl-, n-butyl- and isobutylacetate) and acetic anhydride and its use as solvent in the synthesis of terephthalic acid [4]. Vinyl acetate allows the production of latex emulsion resins for applications in paints, adhesives, paper coatings, and textile treatments. Acetic acid is the second aliphatic compound, after methanol, that is produced from CO/H2 chemistry. This syngas couple results from the gasification (a partial oxidation) of coal (49%), carbon-based fuel oil, refinery residue, naphta (37%), natural gas (9%), petroleum coke (3%), and biomass/waste (2%) [5]. The CO/H2 mixture contains little amounts of CO2. In addition, all of its contaminants must be purified (such as H2S, COS, NH3, HCN, Cl2, tars, alkali particulates, and various trace metals such as Hg, As, Cd, and Se). When synthesized from this CO/H2 syngas mixture, methanol constitutes an abundant and lowcost precursor in the synthesis of acetic acid, provided that the carbonylation process (Eq. 3.1) is highly selective and the catalyst is recycled in addition to the use of purification.

CH3OH + CO → CH3COOH

(3.1)

This catalysis today uses rhodium or iridium metals and it appears that these two platinum group metals (PGMs), which are efficiently recycled, are far from reaching their greatest demand in the automotive industry and petroleum refining. Indeed, the demand for PGMs in the future will be mainly focused on hydrogen production in the next generation of energy technologies [6]. Since the 1913 discovery that acetic acid could be synthesized from methanol and carbon monoxide, intense research has been performed [7–11]. Several years later, Reppe et al. at I.G. Farben and later at BASF patented the catalytic activity of iron, cobalt, and nickel in the presence of an iodide copper salt to carbonylate methanol into acetic acid and methylacetate with most examples Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 1, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

26

3  Synthesis of Acetic Acid Using Carbon Dioxide

focusing on nickel at 230–340 °C and 180–200 bar [12]. BASF and later British Celanese conducted intense work that resulted in many patents involving the use of primarily iron with low activity and cobalt, as well as nickel carbonyls in the presence of iodine or iodide salts, to perform the carbonylation reaction at high pressures and temperatures [13]. The damage by corrosion encountered when using iodide promoters was only solved at the end of the 1950s, when the highly resistant Hastelloy® Mo-Ni alloys were developed [14]. Much work has been done to carbonylate methanol into acetic acid in large industrial plants under low pressures at reasonable temperatures to easily recover and recycle the catalyst in addition to obtaining high selectivity to simplify the purification process. After cobalt, rhodium and iridium complexes are presently used for the industrial production of acetic acid in plants, which today are as large as 1,500 tons/day. Many studies have been conducted with the objective of obtaining attractive catalysts based on abundant and inexpensive metals, to immobilize them on specific supports, and especially to avoid metal and iodide promoters. With global awareness that we use excessive amounts of fossil fuels, whose combustion generates the greenhouse gas CO2m the preferred approach is not only to avoid the production of this gas, but essentially to convert it into fuels and chemicals—the carbon capture and utilization (CCU) option [15]. The global consumption of petroleum and liquid fuels was 96.9 million barrels per day, natural gas 3,918 109 m3, and coal 7,906 106 tons, generating 34 gigatons (34 109 tons) of CO2 emissions in 2021 [16, 17]. Thus, the new strategy is not only to minimize the CO2 emissions, but above all to consume large quantities of CO2 in the synthesis of fuels and chemicals. With such a strategy, recent approaches involve the use of methanol produced from CO2 and H2 as an energy and a dihydrogen carrier. In addition, methanol carbonylation can be performed with CO2 as CO surrogate. Moreover, the direct synthesis of acetic acid starting from methane and carbon dioxide begins to give encouraging results. In addition, enzymatic catalysis and microbial electrosynthesis are promising approaches to obtain large amounts of carbon monoxide, methanol, and acetic acid using large-scale reactors. In the present chapter, we focus on the use of CO2 to synthesize acetic acid, and we analyze the more recent strategies, after a brief historical perspective that focuses on the reaction mechanisms.

3.2  Synthesis of Methanol from CO2 and H2 In 1905, Sabatier and Senderens (who studied hydrogenation catalyzed by metals) discovered that copper was able to decompose methanol into hydrogen and CO/CO2 and, somewhat less efficiently, to form methanol from CO/CO2/H2 [18]. As the technical history of methanol production proceeded [3, 19, 20], industrial developments resulted in the production of methanol from syngas generated from the gasification of coal. This process was first used in Germany in 1923 using a [Cr2O3-ZnO] zinc chromite catalyst operating at 300 bar and 300–400 °C [3]. As syngas contains significant amounts of CO2, the two hydrogenation reactions are exothermic in the sense of methanol formation (Eqs. 3.2 and 3.3), and are favored by low temperatures and high pressures.

CO + 2H2 → CH3OH ∆H = −90.6 kJ / mol

(3.2)



CO2 + 3H2 → CH3OH ∆H = −49.5 kJ / mol

(3.3)

3.2 Synthesis of Methanol from CO2 and H2

Under these conditions, it is necessary to take into account the water-gas shift reaction (Eq. 3.4), which produces carbon dioxide and dihydrogen.

CO + H2O → CO2 + H2 ∆H = −41.2 kJ / mol

(3.4)

For syngas conversion into methanol, two reviews describing the century of evolution of catalysts [3, 20] show that the high pressure (300 bar) and high temperature (300–400 °C) process using Cr2O3–ZnO evolved toward the use of Cu/ZnO/Al2O3 and Cu/ZnO/Cr2O3 as catalysts operating at 30–120 bar and achieving selectivity greater than 99.5%. Today, methanol is produced from syngas arising from fossil sources, particularly through coal gasification. Beyond the large investigations on the syngas/methanol route that resulted in ~600 inventions per year during the last decade, 380 patents have been awarded between 2016–2020 for the CO2/methanol process [20]. On the initiative of Olah to improve methanol economy [21], the VulcanolTM process was developed and a 4,000 tons/year plant was built in Iceland by Carbon Recycling International (CRI) to use the energy provided by a high temperature volcano to perform water electrolysis in a geothermal power station and to react dihydrogen with CO2 recovered from the volcano [22]. Since 2012, this George Olah Renewable Methanol Plant has been able to transform 5,500 tons/year of CO2. Today, the industrial production of CRI from CO2 industrial emissions has reached 110,000 tons/year and 100,000 tons/year in its China and Norway plants, respectively. The catalyst is Cu/ZnO/Al2O3. With a H2/CO2 ratio of 3 in the feed, at 220 °C, 35 bar, and 0.2 g of catalyst, a selectivity of 73.21% in CH3OH is obtained for 20%Cu/ZnO–Al2O3 and 25.83% in CO due to the reverse water gas shift reaction, whereas for a 40%Cu/ZnO–Al2O3 catalyst the selectivity in CH3OH is 94.96% and in CO only 4.04% [23]. Small amounts of CH4, dimethylether, and formaldehyde are detected. The addition of small quantities of palladium to copper significantly increases the productivity in methanol. For example, 2%Pd–20%Cu/ZnO–Al2O3 provides 98.20% of CH3OH and 1.15% of CO. Thus, the addition of palladium facilitates the reduction of copper oxide species due to the spillover effect occurring between Pd and CuO [23]. A recent paper underlines that the literature provides evidence that zinc oxide particles on a copper surface are significantly more active for producing methanol than the classically used Cu/ ZnO catalyst [24]. The authors used a preferential chemical vapor deposition method (PCVD) to deposit Zn on copper nanoparticles rather than on an alumina support. The specific ZnO/Cu structure obtained by PCVD is more selective in methanol than the catalyst produced by the impregnation method with an activity of 930 vs 765 µmol CH3OH/gCu/min (65% vs 45% selectivity) [24]. Similarly, the use of an inverse ZnO/Cu/MgO catalyst shows that the reduced Zn species at the interface plays an important role in favoring methanol synthesis [25]. Cu/MgO catalysts in a semicontinuous reactor allow a 76% conversion of CO2 to be reached and an overall selectivity of 59% in methanol [26]. Other catalytic systems are still being explored. Thus, Cu/ZnO catalysts have been deposited over ZSM-5 acid zeolites by chemical vapor impregnation, giving a modest yield of 192 mmol/ kgcat.h CH3OH and (CH3)2O when compared to the catalyst prepared by oxalate gel precipitation (1322 mmol/kgcat.h) [27]. Magnetic manganese ferrites nanoparticles of MnFe2O4 (average grain size 67nm) and bismuth-doped Bi–MnFe2O4 (84.5nm) were developed in a fixed-bed reactor in which 17 fingers were designed to perform independent reaction cores for hydrogenation of CO2. Operating at 220 °C and using external magnets to recycle the catalysts, Bi–MnFe2O4 at 1 bar of CO2/H2 (1/3) gave a 22% conversion and a 61% selectivity in methanol [28]. Copper-nickel impregnation methods of reduced-graphene oxide were used to prepare CuNi–rGO catalysts that have a

27

28

3  Synthesis of Acetic Acid Using Carbon Dioxide

Figure 3.1  Metal organic frameworks with terephtalate linkers for 6 Ce(IV) and 3 Ce(IV)-3 Ti(IV) metal atoms.

good CO2 activation ability. The conversion in methanol in a fixed-bed flow reactor is 7.87% at 225 °C and 40 bar with a methanol selectivity of 98.7% [29]. These performances were slightly higher than CuNi deposited on graphene oxide (CuNi–GO) or on ammonia modified graphene oxide. The addition of nickel promotes the reduction of Cu2+ in Cu+ and Cu0. Due to the synthesis of metal-organic frameworks (MOFs) of high stability and great capacity to adsorb various gaseous compounds, having 6 zirconium atoms for which the linker is terephtalic acid and called Zr-UiO-66[30], photocatalytic performances have been explored for CO2 transformation into methanol using Ce–UiO–66 and Ce0.5Ti0.5–UiO–66–NH2 (Figure 3.1) catalysts. In a first step, CO2 is adsorbed on the semiconducting catalyst, which has the property of being porous with a high surface area. Next, the photocatalytic reduction is induced by light that generates electrons in the conduction band as well as holes in the valence band [31]. In addition facilitating the reducibility of Ce4+ to Ce3+ due to the substitution of the 2-aminoterephtalate linker with the terephthalate one, the low-lying 3d orbitals of Ti facilitate the photoreduction reaction. Thus, Ce0.5Ti0.5–UiO–66–NH2 induces a better photocatalytic reduction of CO2 to CH3OH (see Eq. 3.5), through HCOOH, with minimal H2 evolution compared with Ce–UiO–66 [32].

CO2 + 6H+ + 6e− → CH3OH + H2O

(3.5)

3.3  Carbonylation of Methanol Using CO2 Currently, the carbonylation of methanol with carbon monoxide is catalyzed by rhodium and iridium. It is necessary to operate in the presence of HI to transform methanol into CH3I, which reacts with the metal center in an oxidative addition steps, as shown in Figure 3.2 for rhodium [11]. Such a homogeneous carbonylation could be carried out using CO arising from the catalytic transformation of CO2 or by using CO2 directly as a CO surrogate. Indeed, recent studies show that the reverse water-gas shift reaction (RWGS, Eq. 3.6) used to convert CO2 with hydrogen into CO and H2O is presently very selective[33].

CO2 + H2 → CO + H2O

(3.6)

For this endothermic reaction (ΔH298=42.1 kJ/mol), the conversion is operated at atmospheric pressure and temperatures above 300 °C. The aim is to avoid the co-production of methanol and methane, and even the formation of alkenes/alkanes through the Fischer-Tropsch reaction [34, 35]. It is thus necessary to operate under precise reaction conditions involving the catalyst particle size, particle dispersion, the nature of the support and its surface energy, the temperature, and the presence of various additives. For instance, 1%Rh/Al2O3 and its Ba-containing additive

3.3 Carbonylation of Methanol Using CO2

Figure 3.2  Carbonylation scheme of methanol in homogeneous phase using [RhI2(CO)2]- as the active species.

produce high selectivity toward CH4 below 500 °C, with a maximum 60% yield at 400 °C. However, K-containing 1%Rh/Al2O3 converts CO2 to CO alone in the 300–700 °C range [36]. The use of Ru-Fe nanoparticles supported on samarium-doped ceria (Sm–CeO2) at 2 wt.%, particularly Ru45Fe55, allowed a 100% CO selectivity to be obtained above 500 °C [37]. Moreover, FeOx nanoparticles supported on Co3O4 give the same performance, but between 200 and 400 °C due to the reducibility of Co3O4 and the generation of oxygen vacancies as well as a strong metal-support interaction [38]. A recent review on the reverse water-gas shift reaction over supported catalysis underlines the performances of perovskite-type oxides to induce good catalytic performances and high CO selectivity, eliminating the methanation side-reaction [39]. The Cu/CeO2 systems allows catalytic efficiency to be increased [40, 41]. Recently, a terpyridine-based iron polypyridine complex [Fe(tpyPY2Me)]2+ was shown to perform the electrochemical reduction of CO2 into CO at very low applied overpotentials, with high selectivity and high rates [42]. The carbonylation reaction can be performed with a 70% yield based on methanol using the couple CO2/H2 [43, 44]. The catalytic reaction needs the two precursors [Ru3(CO)12]/ [Rh2(OCOCH3)4] with the imidazole ligand in the presence of LiI at 200 °C and 80 bar in 1,3-dimethyl-2-imidazolidinone as the solvent. Five catalytic cycles produce approximately the same reactivity and the turnover number reaches 1,000. The alcohol formation is negligible. CH3I formed by the reaction of LiI on methanol produces the oxidative addition to the Rh species; after coordination of CO2, the migratory methyl reaction generates a Rh–COOCH3 bond that is hydrogenated into CH3COOH by the ruthenium catalyst [43]. Further studies revealed that the [Rh2Cl2(CO)4]/LiCl/LiI/4-methylimidazole system is still more efficient than the Rh/Ru previous one [45] because the yield in acetic acid is 82% and the turnover frequency 26.2 h–1 at 180 °C and 50 bar CO/50 bar H2. Here also, the oxidative addition of CH3I generates an Rh–CH3 species and the CO2 coordination followed by the methyl migratory insertion leads to a Rh–COOCH3 species. Isotopic studies support the conclusion that the reaction pathway does not proceed via the CO route.

29

30

3  Synthesis of Acetic Acid Using Carbon Dioxide

Interestingly, two efficient catalysts constituted of MnOx nanoparticles supported on mesoporous Co3O4 or one Cu-nanocrystal encapsulated within a single Zr(IV)-based MOF promote a cobalt catalyst hydrogenation of CO2 that gives rise to methanol with a 100% selectivity (200 °C, 10 bar) and it is therefore possible to synthesize acetic acid exclusively from carbon dioxide [46].

CH3OH + CO2 + H2 → CH3COOH + H2O

(3.7)



CO2 + 3H2 → CH3OH + H2O

(3.8)

The George Olah Plant is currently operating with copper-zinc oxide-alumina to produce about 5,000 tons/year of methanol in Iceland and the size of the plant could reach 35,000 tons/year in a few years [47]. Through an electrocatalytic procedure, it is possible to convert carbon dioxide directly into acetic acid. With copper(I) complexes containing the N,N,N’,N’-tetra(2-pyridyl)-2,6-pyridine-diamine and N,N,N’,N’-tetra(2-pyridyl)-biphenyl-4,4ʹ-diamine ligands supported on carbon-doped boron nitride, in 1-ethyl-3-methyl-imidazolium tetrafluoroborate/LiI/water solution, the Faradic efficiency can reach 80% [48]. Recently, it was discovered that a ruthenium, or cobalt, nickel and copper complex containing a porphyrin, or polypyridine, or phtalocyanine ligand immobilized on a photoactive semiconductor support, such as N-doped carbon, graphite carbon nitride, natural allotrope S8 sulfur, and metal oxides of titanium, nickel, cobalt, iron, zinc, and bismuth is an efficient catalyst for the synthesis of acetic acid by the carbonylation of methanol by CO2 [49]. As claimed in this patent, the yield in acetic acid for the Co-phtalocyanine/S8 system in N,N-dimethylacetamide, under ambient temperature and pressure. After 24 hours of irradiation by a 20W LED, this reaction reached 60% and the selectivity was 81% [50]. After completion of the reaction and filtration of the photocatalyst, three recycling runs provide the same performance. No carbon monoxide is detected in the gaseous products, but oxygen, ethylene, and the water in the liquid phase allow the authors to propose that CO2 is the carbonylation vector rather than CO. Moreover, in a separate experiment, the introduction of styrene allows the production of cylopropylbenzene. Thus, the formation of :CH2 carbene from methanol is considered the first step of the catalytic cycle, allowing reaction with the photoactivated cobalt complex. As shown in Figure 3.3, the carbene-cobalt species reacts with CO2 to Figure 3.3  Proposed catalytic cycle for the synthesis of acetic acid from methanol and carbon dioxide under photoactivation conditions.

3.5  Miscellaneous Reactions, Particularly Biocatalysis

generate the [Co–CH2COO.]* intermediate, which reacts with 2 H+ and 2 e− arising from water and completes the catalytic cycle to produce acetic acid and oxygen [50]. No cobalt leaching occurs.

3.4  Carbonylation of Methane Using CO2 The simultaneous use of CH4 and CO2 as raw materials to synthesize value-added chemicals, focusing on acetic acid in this chapter, represents an attractive route for efficiently reducing carbon dioxide. This reaction, provided it is selective, is a green and 100% atom-efficient route (Eq. 3.9), although it is thermodynamically not favorable.

CH4 + CO2 → CH3COOH ∆G° = 71.0 kJ / mol

(3.9)

In a recent review, the use of metals, metal oxides, and metal modified zeolites has been examined. Cu- and Zn-based catalysts are active for this co-conversion, which remains as low as 8–9%. The knowledge of the active sites requires further investigation, even if the metal site activates the C-H bond of methane and the second component provides a Lewis acid site or oxygen vacancy to aid at the CO2 adsorption and activation [51]. As the thermodynamic yield remains quite low, the authors recommend exploring tandem reactions in two reactors, or nonconventional activation processes such as plasma catalysis.

3.5  Miscellaneous Reactions, Particularly Biocatalysis Enzymatic methods have been explored to extend the metabolic CO2 processes in cells [52]. The reversible reduction of CO2 into CO can be performed with high turnover numbers on the [NiFe4S4] cluster of a [NiFe] carbon monoxide dehydrogenase, with the carbon atom being bonded to nickel and one oxygen atom bonded to one iron atom [53, 54]. Microbial electrosynthesis (MES) is a novel electrochemical reduction technique in which microorganisms are attached to the electrodes to catalyze the CO2 reduction and this could show environmental benefits [55]. In the two-chamber MES reactor, the anode produces, by water oxidation, protons that are transferred through an ion exchange membrane to the biocatalyzed cathode (Eqs. 3.10 and 3.11) [56].

4H2O → 2O2 + 8H+ + 8e−

(3.10)



+ − − 2HCO− 3 + 9H + 8e → CH3COO + 4H2O

(3.11)

In situ grown Mo2C- or nickel-phosphide-modified electrodes improves MES efficiency, for instance, through the attachment of microorganisms such as Acetobacterium and Arcobacter or Clostridium ljungdahlii [57, 58]. Various acetogenic microorganisms can be used [59], and a continuous headspace gas recirculation allows the reaction to reach a cathodic efficiency of 50% [60]. Recent improvements have been achieved but high yields are still necessary for large amounts of CO2 [61–65]. In particular, microbial electrosynthesis of an H-type (H–cell), composed of a cathodic and an anodic chamber separated by a cation exchange membrane, allow an increase in the CO2 gas dissolution and thus improves significantly the production of acetate, even when electric supply is interrupted [66–68]. Increasing the CO2 flow rate from 5 to 20 mL/min results in the H-cell increasing the acetate production rate from 45 to 270 mg/L∙day.

31

32

3  Synthesis of Acetic Acid Using Carbon Dioxide

3.6  Conclusions Recycling CO2 as a C1 building block through a CCU technology appears to be a convenient approach to perform the synthesis of acetic acid using new strategies that would be efficient and economic in comparison with the present industrial processes [69]. In the context of the production of hydrogen by renewable energy, especially solar or wind, hydrogenation of carbon dioxide by a new generation of catalysts and plants to produce methanol will give large amounts of this sustainable reactant. Its carbonylation, by CO or directly by CO2, would yield acetic acid. The direct condensation reaction of carbon dioxide and methane, which is a green atom efficient route, needs to reach higher yields to compete efficiently with the methanol pathway. Enzymatic methods, particularly microbial electrosyntheses, are also powerful methods that are emerging as efficient approaches for the conversion of CO2 into acetic acid.

References 1 Shah, K. (2014). Acetic acid: overview and market outlook. Indian Petrochem Conference. 2 Methanol Institute. (2021). The Methanol Industry (2021). https://www.methanol.org/themethanol-industry (accessed 26 October 2021). 3 Sheldon, D. (2017). Johnson Matthey Technol. Rev. 61: 172–182. 4 Le Berre, C., Serp, P., Kalck, P., and Torrence, G.P. (2014). Acetic acid. In: Ullmann’s Encyclopedia of Industrial Chemistry (ed. B. Elvers). New York City: Wiley. 5 Casleton, K.H., Breault, R.W., and Richards, G.A. (2008). Combust. Sci. Tech. 180: 1013–1052. 6 Hughes, A.E., Haque, N., Northey, S.A., and Sarbjit, G. (2021). Resources 10: 93–134. 7 BASF (1913 8 March). Verfahren zur Darstellung von Kohlenwasserstoffen und deren Derivaten. DE 293787 Patent. Granted on 23 Aug 1916. 8 Mittasch, A. and Schneider, C. (1914, 11 February). Producing compounds containing carbon and hydrogen. US Patent 1,201,850 to BASF (19 May 1970). Granted on 17 October 1916. 9 Fischer, F. (1925). Ind. Eng. Chem. 17: 574–576. 10 von Kutepow, N., Himmele, W., and Hohenschutz, H. (1965). Chemie-Ing. Techn. 37: 383–388. 11 Kalck, P., Le Berre, C., and Serp, P. (2020). Coord. Chem. Rev. 402: 213078. 12 Reppe, W., Kröper, H., and Pistor, H.-J. (1941). Verfahren zur Herstellung von Carbonsaüren und deren estern. DE 763693 Patent to I.G. Farben (March 16, 1941). Granted on April 16 1953. 13 Mullen, A. (1980). Carbonylations catalyzed by metal carbonyls – Reppe reactions. In: New Syntheses with Carbon Monoxide (ed. J. Falbe), 243–308. New York City: Springer Verlag. 14 Arpe, H.J. (2010). Industrial Organic Chemistry, 5e. New York City: Wiley. 15 Aresta, M., Dibenedetto, A., and Quaranta, E. (2016). J. Catal. 343: 2–45. 16 World Energy Outlook (2021). Int. Energy Agency 2021: 1–386. 17 Short-Term Energy Outlook (2021). U.S. Energy Information Administration December 2021: 1–53. 18 Sabatier, P. and Senderens, J.-B. (1905). Ann. Chim. Phys. 4: 319. 19 Mondal, U. and Ganapati, D.Y. (2021). Green Chem. 23: 8361–8405. 20 Bisotti, F., Fedeli, M., Prifti, K. et al. (2021). Ind. Eng. Chem. Res. 60: 16032–16053. 21 Olah, G.A. (2005). Angew. Chem. Int. Ed. 14: 2636–2639. 22 Olah, G.A. and Prakash, G.K.S. (2010). Producing methanol and its products exclusively from geothermal sources and their energy. WO Patent 2010/011504 A2 to University of Southern California. 23 Ciesielski, R., Shtyka, O., Zakrzewski, M. et al. (2020). Kinet. Catal. 61: 623–630.

References

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

Saedy, S., Newton, M.A., Zabilskiy, M. et al. (2022). Catal. Sci. Technol. 12: 2703–2716. Pandit, L., Boubnov, A., Behrendt, G. et al. (2021). ChemCatChem. 13: 4120–4132. Kleiber, S., Pallua, M., Sienbenhofer, M., and Lux, S. (2021). Energies 14: 4319. Tariq, A., Ruiz Esquius, J., Davies, T.E. et al. (2021). Top. Catal. 64: 965–973. Bibi, M., Ullah, R., Sadiq, M. et al. (2021). Catalysts 11: 592. Wang, C., Fang, Y., Liang, G. et al. (2021). J. CO2 Util. 49: 101542. Hafizovic Cavka, J., Jakobsen, S., Olsbye, U. et al. (2008). J. Am. Chem. Soc. 130: 13850–13851. Ramyashree, M.S., Shanmuga Priya, S., Freudenberg, N.C. et al. (2021). J. CO2 Util. 43: 101374–101395. Payra, S., Ray, S., Sharma, R. et al. (2022). Inorg. Chem. 61: 2476–2489. Pahija, E., Panaritis, C., Gusarov, S. et al. (2022). ACS Catal. 12: 6887–6905. Wang, D., Xie, Z., Porosoff, M.D., and Chen, J.G. (2021). Chem 7: 2277–2311. Liu, J., Song, Y., Guo, X. et al. (2022). Chinese J. Catal. 43: 731–754. Büchel, R., Baiker, A., and Pratsinis, S.E. (2014). Appl. Catal. A: Gen. 477: 93–101. Panaritis, C., Edake, M., Couillard, M. et al. (2018). J. CO2 Util. 26: 350–358. Panaritis, C., Yan, S., Couillard, M., and Baranova, E.A. (2022). J. CO2 Util. 56: 101824. Chen, X., Chen, Y., Song, C. et al. (2020). Front. Chem. 8: 709. Li, M., Pham, T.H.M., Oveisi, E. et al. (2021). ACS Appl. Energy Mater. 4: 12326–12335. Xu, J., Li, L., Pan, J. et al. (2022). Adv. Sustainable Syst. 6: 2100439. Loipersberger, M., Derrick, J.S., Chang, C.J., and Head-Gordon, M. (2022). Inorg. Chem. 61: 6919–6933. Qian, Q., Zhang, J., Cui, M., and Han, B. (2016). Nat. Com. 7: 11481–11487. Qian, Q., Han, B., Zhang, J., and Cui, M. (2016). Method for synthesizing acetic acid by methanol, carbon dioxide and hydrogen. CN Patent 107008502 to INST Chemistry CAS (January 27 2016) granted on August 4 2017. Cui, M., Qian, Q., Zhang, J. et al. (2017). Green Chem. 19: 3558–3565. Liu, W.-C., Baek, J., and Somorjai, G.A. (2018). Top. Catal. 61: 530–541. Olah, G.A. (2013). Angew. Chem. Int. Ed. 52: 104–107. Sun, X., Zhu, Q., Kang, X. et al. (2017). Green Chem. 19: 2086–2091. Jain, S.L., Saini, S., Khatri, P.K. et al. (2021), Preparation of acetic acid via photocatalytic hydrocarboxylation of methanol with CO2 under visible light irradiation. Patent: India, IN202011011197 A 2021-09-17; WO2021186459 A1 2021-09-23. Saini, S., Samal, P.P., Krishnamurty, S. et al. (2021). Green Chem. 23: 9048–9060. Tu, T., Nie, X., and Chen, J.G. (2021). ACS Catal. 11: 3384–3401. Shi, J., Jiang, Y., Jiang, Z. et al. (2015). Chem. Soc. Rev. 44: 5981–6000. Fesseler, J., Jeoung, J.-H., and Dobbek, H. (2015). Angew. Chem. Int. Ed. 54: 8560–8564. Jeoung, J.-H., Fesseler, J., Domnik, L. et al. (2022). Angew. Chem. Int. Ed. 61: e202117000. Das, S., Diels, L., Pant, D. et al. (2020). J. Electrochem. Soc. 167: 155510. Gadkari, S., Mirza Beigi, B.H., Aryal, N., and Sadhukhan, J. (2021). RSC Adv. 11: 9921–9932. Huang, H., Huang, Q., Song, T.-S., and Xie, J. (2020). Energy Fuels 34: 11299–11306. Wang, G., Huang, Q., Song, T.-S., and Xie, J. (2020). Energy Fuels 34: 8666–8675. Nevin, K.P., Hensley, S.A., Franks, A.E. et al. (2011). Appl. Environ. Microbiol. 77: 2882–2886. Mateos, R., Sotres, A., Alonso, R.M. et al. (2019). Energies 12: 3297. Vassilev, I., Dessi, P., Puig, S., and Kokko, M. (2022). Bioresour. Technol. 348: 126788. He, Y., Wang, S., Han, X. et al. (2022). ACS Appl. Mater. Interf. 14: 23364–23374. Zhou, M., Zeng, C., Liu, G. et al. (2022). Sci. Total Environ. 836: 155724. Yang, H.-Y., Hou, -N.-N., Wang, Y.-X. et al. (2021). Sci. Total Environ. 790: 148128.

33

34

3  Synthesis of Acetic Acid Using Carbon Dioxide

65 Gharbi, R., Gomez Vidales, A., Omanovic, S., and Tartakovsky, B. (2022). J. CO2 Util. 59: 101956. 66 Del Pilar Anzola Rojas, M., Mateos, R., Sotres, A. et al. (2018). Energy Conv. Management 177: 272–279. 67 Del Pilar Anzola Rojas, M., Zaiat, M., Gonzalez, E.R. et al. (2018). Bioresour. Technol. 266: 203–210. 68 Del Pilar anzola rojas, M., Zaiat, M., Gonzalez, E.R. et al. (2021). Process Biochem. 101: 50–58. 69 Krishnan, U.J.N. and Jakka, S.C.B. (2022). Mater. Today: Proceedings. 58 (3): 812–822.

35

4 New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources A New Catalytic Challenge Ana B. Paninho1, Malgorzata E. Zakrzewska1, Leticia R. C. Correa1, Fátima Guedes da Silva2, Luís C. Branco1, and Ana V. M. Nunes1 1 2

LAQV-REQUIMTE, Departamento de Química, Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, Caparica, Portugal Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, Lisboa, Portugal

4.1  Introduction The worldwide ambition to considerably decrease anthropogenic CO2 emissions involves large efforts in terms of adoption of renewable energies, improvements in energy efficiency, increasing the usage of fuels with low or no carbon content (e.g. hydrogen), and, in cases where emissions cannot be avoided, the implementation of CO2 capture technologies [1]. This latter strategy will generate large amounts of pure compressed CO2, which will go to geological storage unless alternative uses are developed [2]. The possibility of using emitted CO2 as a carbon source for the manufacturing of chemicals and materials is a highly active field of research that aims at recycling (waste) carbon back into the value chain [3]. In fact, despite its greenhouse effect, CO2 is also nontoxic, non-flammable, essential to life, and the earth’s natural source of carbon. Furthermore, as a source of carbon, CO2 does not compete with land for food production and is virtually an inexhaustible source capable of meeting the growing world demand for chemicals and materials. The strategy known as CO2 capture and utilization (CCU) appeared first as a strategy to offset the costs of CO2 capture and sequestration (CCS) into geological formations, trying to make value from CO2 that could pay for capture, purification, and storage technologies. However, the wide range of CO2 applications in chemical synthesis led to the realization that CCU has significant potential for emissions reduction by recycling carbon and displacing fossil fuel usage [4]. Using CO2 as a source of carbon is actually one of the oldest topics in chemical synthesis and most reactions have been known for many years. The topic has recently gained a renewed interest due to the increasing atmospheric concentration of CO2 and consequent climatic impact. As illustrated in Figure 4.1, CO2 conversion approaches include both reductive pathways (high-energy intensive) and non-reductive pathways (low-energy intensive). In the reductive pathway, CO2 is converted into energy vectors such as methane, methanol, formic acid, or syngas, in which the oxidation state of the carbon is reduced. On the other hand, the non-reductive pathway involves the total incorporation of the CO2 moiety into an organic

Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 1, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

36

4  New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources

Figure 4.1  Reductive and non-reductive pathways for CO2 conversion into valuable products.

substrate for the production, for example, of carbonates, carbamates, and urea, without affecting the oxidation state of the carbon center [5]. There are few examples of large-scale CO2 utilization technologies that have been commercialized. The largest application is in urea production by reacting CO2 with ammonia at around 185– 190 °C and at a pressure range of 18–20 MPa, directly utilizing 115 Mt of CO2 per year. Another well-known application is in the synthesis of salicylic acid (an aspirin precursor) by heating sodium phenoxide with carbon dioxide under 100 bar and 125 ºC and then treating the product with acid. This application consumes 29 Kt of CO2 per year. Another example is in the production of cyclic carbonates from the reaction between CO2 and epoxides, directly using 40 Kt of CO2 each year [5]. The cycloaddition reaction between epoxides and CO2 is the most straightforward method to obtain cyclic carbonates and is considered an example of a sustainable process beause it is a 100% atom economic reaction that uses CO2 (a waste) instead of phosgene (used in the conventional process). In contrast with previous examples, the reaction for cyclic carbonate production does not occur unless an efficient catalyst is used (quaternary ammonium or phosphonium salt). The main problem with this catalytic system is that is requires harsh operational conditions that give rise to an overall net carbon emitting process [6]. In this context, the search for new catalysts is a very active field of research with several papers published every year. Cyclic carbonates are naturally occurring [7], highly stable [8], and versatile molecules [9] with a wide range of applications and therefore a considerably growing market. Hence, CO2 conversion into cyclic carbonates has become one of the most studied applications in the field of non-reductive transformation of CO2. Cyclic carbonates are considered green solvents primarily due to their low vapor pressure, low toxicity, and high flash point. Furthermore, they naturally biodegrade by hydrolysis into low toxicity products (the corresponding diol and CO2). Ethylene and propylene carbonates are produced from CO2 at an industrial scale and find applications as polar aprotic solvents [10], as chemical intermediates, and in the formulation of electrolytes for lithium batteries [11]. One of the most promising and well-explored applications for cyclic carbonates is their use as monomers for polymer production, namely in the production of isocyanate-free polyurethanes [12].

4.2  Cyclic Carbonates from Bio-based Epoxides

Epoxidized vegetable oils and fatty acids

Carbohydrates

O C Natural derived epoxides

O

O

Bio-based diols

100% Renewable cyclic carbonates applications

Electrolytes Green solvents Materials

Chemical intermediates

Figure 4.2  Renewable substrates that can be combined with CO2 to produce cyclic carbonates.

It is important to note that using CO2 as a carbon source in a process does not necessarily means an environmental advantage over conventional routes and a careful analysis is essential to determine the achieved reduction in the carbon footprint. This reduction has to account for the amount of CO2 actually converted, but also the amount not emitted owing to the process improvement in terms of reactants used, waste produced, and energy consumed [5, 13]. Recently, many research groups have focused on the combination of CO2 with bio-based substrates, namely those derived from biomass and waste biomass [14–17]. This strategy (Figure 4.2) is highly attractive especially in the context of the production of fully sustainable new chemicals and materials for the chemical industry. However, most bio-based substrates are much less reactive, turning catalysis into a key factor in particular with respect to the techno-economic viability of the process. Another important factor to take into account for all potential bio-based substrates is related with their availability (and cost), which should match the desired market size of the targeted application of the carbonate product. This chapter will present the most promising substrates and catalytic systems developed to obtain organic cyclic carbonates from CO2, analyse new trends, and focus on the future challenges.

4.2  Cyclic Carbonates from Bio-based Epoxides The coupling reaction between CO2 and epoxides is as the most investigated method to obtain cyclic carbonates. It is a 100% atom efficient reaction (Figure 4.3) and constitutes an established field of research due to the intensive development of improved stable catalysts and new concepts [11]. Most investigated catalytic systems are homogeneous metal-based complexes, combined with a nucleophile, very often common organic salts. Several papers have been published that extensively

37

38

4  New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources

Figure 4.3  100% atom efficient cycloaddition reaction between CO2 and epoxides.

review the most promising catalytic systems [18–22]. The reaction mechanism starts with the coordination of the oxygen atom of the epoxide with the metal center of the catalyst, followed by the opening of the epoxide ring promoted by a nucleophilic attack by the halide species, which gives rise to a metal-alkoxide. The metal-alkoxide reacts with CO2 to form a metal-carbonate intermediate that further undergoes a back-biting reaction that closes the carbonate ring. Depending on the metal catalyst, the nature of the epoxide, the nature of the halide, and the operating conditions used, the successive incorporation of CO2 and epoxide molecules on the metal-carbon intermediate is favoured (over the back-biting reaction), resulting in the formation of polycarbonates [19, 20]. The most explored epoxides are derived from fossil fuels and include ethylene, propylene, and styrene oxide. The high toxicity and volatility of propylene oxide and, particularly, of ethylene oxide (which is a gas at room temperature) present a huge safety concern. Alternatively, bio-based routes are being developed for the production of ethylene and propylene. However, these paths are not yet at the stage of industrial implementation [11, 15]. In fact, life cycle analysis studies reporting on the production of carbonates from CO2 and epoxides point out the production of the epoxide as the bigger contributor to overall environmental impacts [23]. Due to increasing awareness of environmental challenges and recent developments in green chemistry and sustainable development, there is interest in the replacement of fossil-based epoxides by bio-based derived epoxides for the production of 100% renewable carbonates [15]. However, these bio-derived species are much more challenging due to their complex structure, with considerably lower reactivities reported, making catalysis a major player in the future development of the field. If we consider the bio-based industrial production of carbonates, the use of renewable-based epoxides is still rather limited, although some of these compounds (e.g. epoxidized fatty acid methyl esters and glycidol) are already produced in industrially relevant quantities as the main products and by-product of biodiesel production, respectively [24, 25]. Terpenes are another class of bio-based compounds that can be selectively oxidized to the corresponding epoxides and most are produced by plants and microorganisms as mixtures of isomeric compounds, requiring mild conditions for further interconversion [26]. Terpenes have gained increasing attention in the field of polymer science and sustainable materials. Terpene compounds have been applied for a long time in the preparation of fragrances, flavors, and pharmaceuticals. Some terpenes, such as pinene, carvone, myrcene, and limonene are presently obtained from turpentine oil, a by-product of paper pulping process and extraction from citrus fruits. Furthermore, terpenes have a large structural diversity, which allowed the possibility of producing complex cyclic carbonates by relatively easy transformations such as oxidation followed by coupling with CO2 [27]. Limonene is the most common terpene, with the (R)-limonene enantiomer being the main compound found in the oil extracted from orange peel (a by-product of the citrus industry). Its abundance and low cost make it an excellent choice as a biorenewable epoxide monomer for the reaction with CO2 [26, 27].

4.2  Cyclic Carbonates from Bio-based Epoxides

4.2.1  Bio-based Epoxides Derived from Terpenes Terpenes are readily available at a large scale, with worldwide production of about 350,000 tonnes/ year [26]. Although copolymerization reactions are out of the scope of this chapter, it is worth mentioning that the first successful reaction of a terpene-derived epoxide with CO2, came from Coates et al. in 2004, who reported the successful copolymerization of limonene oxide and CO2 using β-diiminate zinc complexes as catalysts [28]. Since then, several other examples of studies on terpene derived carbonates have been reported [17]. Fiorani et al. [29] manage to selectively obtain limonene cyclic carbonate over the polycarbonate product [30] in the presence of an Al(III) triphenolate complex in conjunction with bis(triphenylphosphine)iminium chloride (PPNCl) as an active binary homogeneous catalyst system. The authors observed a change in the reaction selectivity by increasing the temperature from 40 ºC to 85 ºC, with the higher temperature selectively giving rise to the cyclic product. This result is in accordance with previous studies showing that the formation of the cyclic product is thermodynamically more favourable, with the polycarbonate (the kinetic product) favoured at lower temperatures [11]. The authors also observed that the cis isomer reacted faster than the trans isomer in the copolymerization process and that the opposite occurs for the cycloaddition reaction. The cyclic carbonate formation starting from the trans isomer seems to be the favoured process. 1,2-Limonene oxide (a mixture of cis and trans) is the most studied compoundn due to the higher reactivity of limonene endocyclic double bond. In fact, limonene presents two double bonds and both can be epoxidized to 1,2-limonene oxide, 8,9-limonene oxide, or dipentene oxide as illustrated in Figure 4.4. Considering the challenging nature of these coupling reactions, the isolated yields of the targeted bio-carbonates were reasonable and in most cases in the range of 50–60% [29]. The best result was obtained for trans-limonene oxide with 73% conversion and 57% yield. The others,

Figure 4.4  Structures of the R-limonene enantiomer, limonene-based oxides, and limonene-based cyclic carbonates.

39

40

4  New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources

(bi)cyclic terpene oxides (including carvone oxide, limonene diepoxide, and menthene oxide), exhibited a similar reactivity to give the respective terpene (mono)carbonates with 45–52% yield and high chemoselectivity under the conditions of 85  oC and 1MPa for 66 hours. This study also involved the investigation of terpenes derived from citronellyl acetate, geranyl acetate, linalyl acetate, and neryl acetate, from which moderately high yields of cyclic carbonate were obtained. The conversion of myrcene gave a complex reaction mixture, resulting in a low yield of cyclic carbonate due to additional side reactions likely involving the conjugated double-bond. Recently, another study with an efficient complex aluminum catalyst (1 mol%) in combination with 3 mol% of Bu4NCl as a cocatalyst and solvent-free was published [31]. The synthesis of terpene limonene cyclic carbonate occurred with the maximum conversion of 66% after 66 hours. Another innovation in the field was the use of scorpionate ligands. Martínez et al. developed a lanthanum heteroscorpionate catalyst complex that has shown good catalytic activity for the synthesis of cyclic carbonates from epoxides and CO2 (conversions around 60%) [32]. The catalytic process requires low catalyst loading and mild reaction pressure conditions (100 oC, 10bar, 16h) for the synthesis of cyclic carbonates. In 2020, Fernández-Baeza et al. investigated an efficient scorpionate ligand as a catalyst at 80  oC and 20 bar [33]. Scorpionate zirconium amides were studied in combination with TBABr as a cocatalyst and good yields were obtained (56% after 24 hours). In parallel, Longwitz et al. developed a calcium-based catalyst system. Different crown ethers were tested as ligands in combination with various co-catalysts for the possible activation of CO2. They found that the addition of triphenyl phosphane as a co-catalyst leads to a significant increase in the activity, similar or even higher compared with the use of organic superbases like 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) and 1,5,7-triaza-bicyclo-[4.4.0]dec-5-ene (TBD) [34]. After optimization, they were able to fully convert the limonene epoxide with good chemoselectivity towards the respective carbonate, isolating 80% of the desired product, although harsher reaction conditions were necessary (75 oC, 50 bar, 48 hours). Bahr et al. performed some of the first work with this reaction under a metal-free catalyst. Limonene dioxide and CO2 were catalysed by TBABr alone (3 mol%), resulting in almost complete conversion at 140  oC and 30 bar after 55 hours [35]. The reaction was also performed using SiO2supported pyrrolidinopyridinium iodide (SiO2-PPI) as a bifunctional heterogeneous catalyst. However, due to the steric hindrance of the limonene decarbonate, the heterogeneous catalyst has shown a significantly lower conversion (78% after 120 hours under the same reaction conditions). Later, Morikawa et al. improved this metal-free system. Once more, halides salts were selected as catalysts (TBABr, TBACl, and TBAI), at 100  oC, 30 bar, and 20 hours [36]. The best result was obtained using 10 mol% of TBACl, providing conversions of 76% (trans-limonene oxide), 51% (mixture cis/trans limonene oxide), and 19% (cis-limonene oxide), which are in line with the results obtained through metal-based catalyst. In 2019, a detailed kinetic study of the limonene oxide/CO2 reaction with TBAC as a catalyst was performed [37]. This work established the higher reactivity of the trans isomer in the formation of limonene carbonate. The reaction kinetics shows the first-order dependence with respect to all the reaction components (1,2- limonene oxide, CO2, and TBACl). An innovative step that contributes to these aspects is given by the conversion from batch to continuous processes. The continuous flow production of cyclic carbonates from CO2 and limonene oxide was explored by Paninho and Nunes using a biphasic system scCO2 + ionic liquid [38]. In different approach, alginate aerogels were explored as support for catalysts for limonene carbonate production from CO2 and limonene oxide [39]. This approach takes advantage of both homogeneous and heterogeneous catalysis together with the utilization of a bio-based support with high catalyst loadings.

4.2  Cyclic Carbonates from Bio-based Epoxides

4.2.2  Bio-based Vinylcyclohexene Oxide Derived from Butanediol 4-Vinylcyclohexene oxide (VCHO) can be prepared through the dimerisation of bio-1,3-butadiene, which is produced from 1,4- or 2,3-butanediol that are both commercially available from fermentation of C5 and C6 sugars [15]. The synthesis of cyclic carbonate from VCHO and CO2 was first investigated without using any solvent in the presence of ionic liquid as a catalyst [40]. Ionic liquids based on 1-alkylmethylimidazolium salts of different alkyl groups (ethyl, butyl, hexyl, and octyl) and different anions (Cl−, BF4−, and PF6−) were tested as catalysts. The conversion of VCHO was affected by the structure of the imidazolium salt ionic liquids, specifically, the ones with the cations of bulkier alkyl chain length and with increasing nucleophilic anion showed better reactivity. Reaction temperature, pressure, and the use of a zinc halide co-catalyst enhanced the addition of CO2 to VCHO. Semibatch operation with a continuous supply of CO2 showed higher VCHO conversion than the batch operation. The use of zinc halide co-catalyst with the ionic liquid enhanced the reactivity due to the cooperative action of both the acidic site (Zn) and basic site (Br−). Afterward, Pescarmona et al. evaluated the versatility of pyridylamino-bisphenolate Fe(III) complexes as a catalyst with this challenging substrate, VCHO combined with CO2 [41]. The catalytic activity and product selectivity in the CO2–VCHO reaction (60 °C, 80 bar, 18 hours), followed similar trends to those observed for the CO2–CHO coupling reactions. An increase in selectivity towards the cyclic carbonate was observed with the higher relative amounts of co-catalyst. As expected, the results of this study also showed that higher selectivity towards cyclic carbonate was attained by increasing the reaction temperature. Poly(vinylcyclohexene carbonate) was also achieved using OAc- and Cl- as nucleophiles and a Fe-complex/nucleophile with 18% and 48% of epoxide conversion, respectively, and with the latter nucleophile giving significantly higher conversion. The higher nucleophilicity and smaller size of the chloride compared to the acetate anion explain the observed difference in activity. Remarkably, the selectivity of the reaction between VCHO and CO2 could be completely switched by using Bu4NBr as a co-catalyst and performing the reaction with a catalyst-to-co-catalyst ratio of 1:10. Under these conditions, full selectivity towards the cyclic carbonate was achieved, with an excellent epoxide conversion of 92%. These results show that tuning the type and relative amount of co-catalyst combined with the FeIIIX[pyridylamino-bis(phenolate)] complexes, allows efficiently switching the reaction selectivity between exclusive polymeric or exclusive cyclic carbonate formation. This demonstrates that FeIIIX[pyridylamino-bis(phenolate)] complexes are versatile catalysts for this reaction. In 2021, bimetallic zinc complexes bearing macrocyclic thioetherphenolate [OSSO]-type ligands were investigated as catalysts, assisted by onium salts, for CO2 fixation into cyclic carbonates from a broad range of epoxides, including the biomass-derived epoxide, VCHO [42]. The obtained results for vinylcyclohexene carbonate showed good yields (higher than 50%) and selectivity, but high temperatures were required (130  oC). Furthermore, these catalysts allowed resistance to many kinds of impurity, good recyclability with negligible losses in catalytic activity, and high stability to moisture and oxygen. Kamphuis et al. also extended their work with amino-tris(phenolate)-based metal complexes to internal epoxides such as cyclohexene oxide and VCHO. In this work, amino-tris(phenolate)-based complexes incorporating group(IV) metal centers (titanium and zirconium) in combination with tetrabutylammonium halides (iodide, bromide, and chloride) were used and only a moderate conversion was achieved when VCHO was used as substrate (17% of conversion for the corresponding cyclic carbonate product) [43]. It was also observed that when VCHO was used, the selectivity towards the cyclic carbonate was not complete (92% of selectivity) and a significant fraction of polycarbonate product was also observed from the Fourier transform infrared (FTIR) spectra. This is a

41

42

4  New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources

well-known phenomenon, which presumably originates from the geometric ring strain of the two interconnected rings in the cyclic carbonate product, rendering it less thermodynamically favourable [44]. Such difference in selectivity can be explained by the fact that a higher nucleophile-tometal ratio was used and the replacement of chloride (from PPNCl) to bromide (from TBABr) as a nucleophile, with both features promoting the selectivity towards the cyclic carbonate product [45].

4.2.3  Bio-based Epichlorohydrin Derived from Glycerol Epichlorohydrin (ECH) is one of the more commercially important aliphatic epoxides used extensively as an industrial intermediate, laboratory reagent, and insecticide [46]. ECH is also used as starting material in the preparation of glycidyl ether based epoxides, which have been employed as substrates in the CO2/epoxide polymerisation reaction [15]. The production of of ECH from bio sources is still limited, but ECH can be derived from bio-based glycerol in a two-step process. However, the use of HCl as a chloride source decreases the greenness of this option [47]. More specifically, the glycerol fraction can be obtained from crude glycerol, resulting in residuals rich in fatty acids (fatty acid fraction). The refined glycerol can be used to produce ECH according to Dow’s Glycerol-to-Epichlorohydrin patent [48]. Then, the oxirane group from obtained ECH can be used to produce epoxide monomers. Kim et al. investigated metal-organic frameworks (MOFs) for the CO2 cycloaddition to epichlorohydrin under solvent-free reaction conditions [49]. Aluminium fumarate (Al fumarate) is a MOF, low-cost, and environmentally friendly material because it is produced in water, and the process uses an abundant and simple aluminium salt with low-priced fumaric acid as an organic linker. When Al fumarate was tested as a catalyst with Bu4NBr for the CO2 coupling reaction with ECH, high conversions of chloropropene carbonate (≥ 96%) with excellent selectivities (≥ 97%) were obtained after 6 hours of reaction at 10 bar and 50 oC. The recovery of the Al fumarate catalyst was also explored, but it showed a steady decline in catalytic activity during the recycling runs, as reported for other metal-organic framework catalysts [50]. In 2019, Werner et al. studied the use of polystyrene-supported bifunctional ammonium and phosphonium salts for converting glycidyl methacrylate into the corresponding cyclic carbonate with yields above 95% [51]. However, the activity of the catalyst against bio-based epoxides was not tested. A life cycle analysis (LCA) was performed for this work, estimating that transforming glycidyl methacrylate into its cyclic carbonate slightly reduced the overall process carbon footprint. The importance of using bio-based starting materials was also highlighted, in this LCA it was estimated that total CO2 emissions could be reduced further (by 47% overall) if EHC from the biobased epicerol process was used.

4.2.4  Epoxidized Vegetable Oils and Fatty Acids Vegetable oils and their derived fatty acids contain unsaturated groups, which can be epoxidized, and these represent a promising bioresource substrate to cycloadition with CO2. Vegetable oils have a global production of over 200 Mt per year [17]. Epoxides derived from vegetable oils can offer several direct and indirect advantages over their petroleum-based counterparts, including reduced fossil-fuel use, environmental pollution decreasing, energy savings, improved biodegradability, employment generation, and savings of foreign currency. Beyond that, part of this interest is also related to their easy preparation [52] and availability through high production worldwide [53]. Vegetable oils vary significantly depending on the type of fatty acids present in the oil. For example, the composition of soybean oil includes stearic acid, oleic acid, and linoleic acid, whereas castor oil contains ricinoleic acid with a hydroxyl group located internally. The fatty acids profile

4.2  Cyclic Carbonates from Bio-based Epoxides

usually determines the properties of the initial vegetable oils. Vegetable oil with a higher ratio of unsaturated fatty acids may have a lower melting temperature. The degree of unsaturation of vegetable oils is closely associated with crosslinking capacity, which serves to modify the mechanical properties of the obtained materials [54]. Due to their capacity to react with a diversity of substances (e.g. acids, alcohols, thiols, polyfunctional amines, and other functional groups), epoxides derived from vegetable oils are becoming a potential precursor to produce bio-based products, which then can be applied as green materials [17, 55]. A variety of vegetable oils such as castor oil, soybean oil, linseed oil, karanja oil, canola oil, hemp oil, cottonseed oil, nahar oil, rapeseed oil, and palm oil can be used to produce epoxides [56]. The epoxidized soybean oil (ESO) has been commercially used in large amounts, around 45 million tons (in 2013), because of its ease of production and availability at a relatively low cost. In addition, epoxidized methyl soyate and epoxidized allyl soyate are also available on the market [57]. Linseed oil is also widely used as a plasticizer or additive in the processing of poly (vinyl chloride) (PVC) [58]. Epoxidized vegetable oils can also be applied for the coupling reaction with CO2 to obtain carbonates. The possibility of coupling CO2 with epoxidized vegetable oils to afford compounds that are attractive as additives or chemical intermediates in the synthesis of non-isocyanate polyhydroxyurethanes (NIPU) can conjugate highly sought-after CO2 valorization with the exploitation of bio-based substrates. This conjugation of CO2 valorization with the exploitation of bio-based substrates has been reported [59]. Among the possible applications, the obtained bio-based carbonates can be applied as plasticizers for PVC [60] and as building blocks for the synthesis of nonisocyanate polyurethanes (NIPUs) [61]. Vegetable oils can be used as monomers in the synthesis of NIPUs by polyaddition or polycondensation reaction with amines [62]. This reaction avoids the use of the highly toxic isocyanate monomers, which use is becoming more restricted by European regulation [63]. Although they show promising potential, there are some drawbacks hampering the use of epoxidized vegetable oils and CO2 as feedstocks to produce carbonated vegetable oils. The CO2 molecule (which is considered almost an inert molecule due to its high stability), and the relatively lower epoxide content and reacting activity compared with petroleum-based counterparts is a disadvantage and remains a challenge [64]. Fortunately, these obstacles can potentially be overcome by designing the appropriate catalytic system. Different work on the cycloaddition reaction of CO2 to sterically hindered oleochemical internal epoxides have been reported engaging different catalytic systems with moderate reaction conditions. As previously seen for the other bio-epoxides, among the catalysts employed, TBABr is the most thoroughly investigated and studied for vegetable oils [65]. However, an efficient application of this catalyst requires reaction temperatures above 100 ºC, high to very high CO2 pressures (between 25 and 100 bar), high amounts of catalyst (2–7 mol%), and long reaction times (typically >24 hours). Consequently, research on new effective, efficient, cheaper, and more thermally robust catalysts is greatly needed [66]. The addition of metal-based Lewis acidic complexes to nucleophilic quaternary salts has the effect of strongly accelerating the carbonation of internal epoxides. Different complexes were reported to promote the carbonation of epoxidized methyl oleate (EMO) in the presence of nucleophilic halide salts in the 70–100 °C temperature range under 5–10 bar CO2. Chen et al. developed a catalytic system comprising an iron(II) pyridine-bridged complex and a nucleophilic halide with various cis-fatty acid-derived epoxides for the preparation of a series of stereo-bio-derived cyclic carbonates [67]. Epoxidized methyl oleate was selected as a model substrate and tetrabutylammonium iodide (TBAI) as cocatalyst. The reactions obtained were highly stereoselective and had excellent yields even under mild reaction conditions (5 bar, 100  oC, 24

43

44

4  New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources

hours). The stereochemical divergence in the fatty acid-derived products strongly depends on the leaving ability of the halide anion in the cocatalyst. Another catalytic method for the preparation of a series of fatty acid derived biocarbonates has been reported, comprising an Al(III) aminotriphenolate complex as a catalyst [68]. The investigation started by selecting epoxidised methyl oleate as model epoxide, combined with a bromide based nucleophile (tetrabutylammonium bromide). Further experiments were conducted in the presence of chloride based nucleophiles [PPNCl  =  bis(triphenylphosphine)iminium chloride]. The use of chloride based nucleophiles proved to be beneficial to produce almost exclusively the cis-configured. It was found that the binary catalyst comprising an Al(III) aminotriphenolate complex and PPNCl allows the stereoselective conversion of methyl esters of various epoxy fatty acid derivatives under comparatively mild reaction (70–85 °C, 10 bar CO2). These conditions were then the starting point for the catalytic coupling of the bis-epoxy derivative of methyl linoleate, which was also obtained in good yield, high chemoselectivity (>99%), and excellent stereoselectivity (cis/trans = 97: 3). Inspired by the successful previous preparation of carbonates from oleic and linoleic acid precursors, the authors then shifted their attention to the use of epoxidised methyl linolenate. This synthesis of tricarbonate product proved to be more challenging; in particular the overall stereoselectivity control was markedly lower. Good conversions were also obtained but the cis/trans ratios did not exceed 71:29. In the end, the selective conversion of mono- (oleate), di(linoleate), and tris-epoxy (linolenate) substrates with sterically challenging combinations of vicinal epoxide groups has been achieved with high levels of conversion and diastereospecificity with cis/trans ratios in the products. Ascorbic acid, a bio-based and ubiquitously available compound that has recently emerged as an efficient bifunctional hydrogen bond donor (HBD) for the cycloaddition reaction of CO2 to terminal epoxides [69], was discovered as an efficient HBD for the cycloaddition of CO2 to epoxidized fatty acids esters (EFAEs) and epoxidized vegetable oils (EVOs). The experimental conditions required were mild reaction conditions of temperature (80–100 °C) and pressure (5–10 bar) in the presence of quaternary ammonium halide salts [59]. Despite the occurrence of various by-products (such as ketones, allylic alcohols, and cyclic ethers), the reaction conditions could be tuned for each class of substrates to achieve the target carbonates in high isolated yields and regioselectivities. Overall, compared to previously published molecular Lewis acids and organocatalysts able to accelerate the carbonation of epoxidized fatty acids, ascorbic acid is a sustainable, inexpensive, and ubiquitously available HBD. Its use for the synthesis of carbonates requires just 5–10 bar CO2 pressure and moderate reaction temperatures (80–100 °C). In conclusion, we have verified that vegetable oil-derived biocarbonates are a class of useful biobased intermediates and chemicals that, after process optimization, could be efficiently produced from the epoxidized vegetable oils using simple catalysts usually in the presence of a quaternary ammonium halide salt under mild conditions.

4.3  Cyclic Carbonates Derived from Carbohydrates The application of bio-based sugars and related saccharides as natural and abundant alternatives for CO2 storage, adsorption, and conversion to valuable materials have been described [70]. They occupy the major constituent of the plant-derived biomass and thus are considered a promising feedstock for constructing a sustainable process in this area [71]. In the case of CO2 fixation, Sun et al. [72] developed a superbase/cellulose catalytic system to obtain cyclic carbonates from epoxides and CO2. They reported high conversions and selectivities associated to DBU/cellulose combination. Cellulose acts as a hydrogen bond donor and the

4.3  Cyclic Carbonates Derived from Carbohydrates

superbase as the nucleophile in the activation of the epoxide. In a different study, Tamboli et al. [73], reported the use of chitosan/DBU dissolved in 1-mesyl-3-methylimidazolium (mesylMIM)based ionic liquids for preparation of dimethyl carbonate (DMC) from methanol and CO2. Carrera et al reported [74] the use of monosaccharides, oligosaccharides, or a polysaccharide-activated by combination with adjustable proportion of liquid DBU or TMG as organic superbases, for CO2 capture. In general, it required low superbase ratios that lead to highly viscous solutions with hampered capacity for CO2 mass transfer and poor performances in capture performance. From a different perspective, an excess of the organic superbase would lead to high dilution of the capture agent with associated limitation on the wt% of CO2 uptake. After the optimization for maximal performance, it is important consider that D-mannose:DBU (0.625/1 in equivalents) leads to 13.9 wt% of CO2 uptake and 3.3/5 alcohol functionality converted to carbonates. It is also important to consider effective stirring to overcome the increase of viscosity with the progress of reaction, which limits CO2 uptake. In another study, Carrera et al [75] reported the use of stable gels of cellulose prepared at room temperature by the combination of cellulose and tetramethylguanidine (TMG) in different ratios (1:1, 1:2, 1:3 in equivalents of alcohol groups of cellulose per number of molecules of TMG). The concentration of cellulose in the gel, temperature, frequency of oscillation, and shear rate were optimized considering their possible use as matrices for CO2 capture. In a different study, Eftaiha et al. [76] used chitin acetate in DMSO for CO2 capture. The mechanism involves activation of the alcohol groups by DMSO followed by conversion into carbonates that are stabilized by the ammonium groups available in chitin acetate. A very interesting concept was reported by Sehaqui et al. [77] for the preparation of cellulose-polyethyleneimine foams and further study of its properties on CO2 capture from air. There are other studies for carbon dioxide absorption by the use of chitin or chitosan dissolved in ionic liquids. Carbon spheres were prepared from alginate and chitosan, after thermal treatment between 673 to 1073 K, leading to a good capacity for CO2 adsorption. The high conductivity presented by alginate-based spheres is crucial for the development of an adsorption/desorption system based on the use of electric power as a “switch” with low energetics associated with CO2 capture and release [78]. In recent years, some CO2 conversion processes have been described. One is the hydrogenation of sugars into a strong base to form a value-added product and another is focused on the preparation of cyclic carbonates using phosgene-derived reagents, and later produce a poly-glycocarbonates. Recently, the transformation of sugars and CO2 via room temperature transfer hydrogenation has been investigated. This affords three value-added products: sugar acid, sugar alcohol, and formate, using Pt/C and Pd/C under various reaction conditions. In one approach, CO2 was converted in carbonate and bicarbonate and this species promoted the hydrogenation of the glucose. Aqueous-alcoholic solutions greatly enhanced catalytic activity due to the increased H2 solubility and catalyst dispersion. Having 50% 2-PrOH solution produced the maximum formate yield of 30.5%. Using carbonate salts showed greater catalytic activity compared to bicarbonate, in the same pH (11.5–11.8), and the larger size of alkali cation also effectively enhanced productivity because of the higher solubility of salts. Importantly, the alcohols and alkali cations showed that the water structure breaking effect was more prominent on the glucose anion and formate, compared to the glucose molecule, resulting in the increase in the yields of gluconate and formate. The heterogeneity of the reaction system was confirmed by retained catalytic results during three consecutive recycle tests, and the sugar source could be extended to galactose and lactose [71]. In other approach, a protocol to replace phosgene-derivatives in low CO2 pressure was developed [17]. In this process, CO2 and an organic base (DBU) form an ionic hemi-carbonate intermediate, after which tosyl-chloride is added to yield a cyclic carbonate. Although the yields are not higher, this proces proved to be capable of providing the six-membered cyclic carbonate derivative of D-xylose [9]. In a subsequent work, the process was improved as to allow the

45

46

4  New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources

carbonation to occur in one step, and higher yields of the cyclic carbonate vs by products such as oligomeric species or tosylation of the alcohols were reported. The use of a weaker base such as Et3N or 2,2,6,6-tetramethylpiperidine (TMP) proved to be essential for the overall chemoselectivity. Using this strategy, the five- to eight-membered cyclic carbonates could be synthesized with higher yields comparable to the yields obtained using phosgene-reagent based syntheses [17, 79]. Gregory et al. first reported this type of reaction considering the synthesis and ring-opening polymerization (ROP) of a cyclic carbonate-functionalized mannose derivatives [80]. 1-O-Methyl-α-Dmannose was protected at positions 2 and 3 by an isopropylidene acetal. Subsequently, a six-membered cyclic carbonate was formed by subjecting the compound to DBU and CO2 followed by tosylation with TsCl and Et3N. The yield of the product (57%) was higher than in similar syntheses for D-glucose (36%) and D-xylose (41%)-derived cyclic carbonates mediated by phosgene-reagents. Indeed, a controlled polymerization of the obtained cyclic carbonate is feasible under organocatalytic conditions using TBD (1,5,7-triazabicyclo[4.4.0]dec-5-ene) and 4-methylbenzyl alcohol as initiator. An interesting candidate is 2-deoxy-D-ribose. Naturally, this sugar exists in its pyranose form, which exposes a cis-diol that can be converted into a cyclic carbonates through phosgene-based pathways [78]. Another strategy is to produce the cyclic carbonates by thymidine, one of the bases of DNA, containing the same 2-deoxy-D-ribose sugar backbone. Both phosgene-reagent and DBU-CO2mediated syntheses were unsuccessful likely due to the high ring strain of trans-fused cyclic carbonates units in furanose sugars. To relieve the ring-strain but retain polymerization potential, the secondary alcohol at position 3′ of the thymidine was tosylated. Subsequent hemi-carbonate formation at alcohol 5′ and ring-closure led to a six-membered cyclic carbonate with the stereochemistry at the 3′-position inverted. In addition, the free NH-group of thymidine required methylation so that carbonate formation would not be inhibited. As in other comparable cases, the cyclic carbonate monomer could be polymerized with a high degree of control over the molecular weight of the resultant polycarbonate [78]. Later, carbonates of glucose by inclusion of CO2 were obtained [81]. A protection group was used to produce a fully-protected glucopyranose with a benzylidene-acetal on alcohols 4 and 6, methylated anomeric alcohol, and either a methyl or a methyl triethyleneglycol group on alcohols 2 and 3. Bromination of the acetal group and hydrolysis of the resulting benzoyl group on position 4 led to a halo-alcohol derivative. Under a slightly elevated pressure of CO2 (10 bar) and in the presence of DBU, this compound was carbonated to give the glucose-based cyclic carbonate with retention of stereochemistry. Using this method could synthesize cyclic carbonates from a D-mannose and a D-galactose and polymerize them through ROP [82]. The simple one-step reaction involves CH2Br2 to generate a productive leaving group in the hemi-carbonate fragment that allows subsequent ring-closure to yield the cyclic carbonate [79–83]. This class of compounds is a recent attractive option that is sustainable in both transformations. However, they have not been explored much. One reason for this is that the focus has been on the cellulose and lignocellulose products, which have a lower added value. In the future, the catalytic methodologies need to be improved, explored, and scaled up to be used as an alternative method to convert CO2.

4.4  Cyclic Carbonates Derived from Bio-based Diols Alcohols provide an alternative bio-sourced feedstock for the production of organic carbonates [17, 84, 85]. Coupling carbon dioxide with alcohols, both simple monoalcohols to yield acyclic carbonates or diols to obtain their cyclic derivatives, suffers from unfavourable thermodynamics due to equilibrium limitation. Upon formation of a desired product, an equimolar amount of water is

4.4  Cyclic Carbonates Derived from Bio-based Diols

Figure 4.5  The direct synthesis of organic carbonates from CO2 and alcohol.

produced shifting the equilibrium towards the reactant side (Figure 4.5). This is an additional obstacle to the previously mentioned inherent stability and inertness of CO2 molecule. Many catalytic systems, either metal-based or metal-free, with the aid of physical or chemical dehydrating agents, have been tested in the synthesis of cyclic carbonates from CO2 and various alcohol substrates, including vicinal and 1,3-diols, allylic or propargylic alcohols, and halohydrins [86–88] (Figure 4.2). Diols are promising substrates that react with CO2 to produce cyclic carbonates and that can be derived from biomass. 1,3-Propanediol and 1,4-butanediol, with market volumes of 146,000 and 2 million tonnes/year [89], respectively, are produced at a commercial level by the fermentation of sugars [90–93]. The majority of 1,3-propanediol is made by the fermentation of glucose (DuPont Tate and Lyle Bio Products [93]). Bio-based 1,4-butanediol is being industrially produced by the direct fermentation of sugars using the technology developed by Genomatica, USA, in the world’s first commercial plant in Italy (Novamont) [94, 95]. However, cyclic carbonates from 1,4-butanediol are unfortunately not feasible due to the low stability of the seven-membered ring of the corresponding product. 1,2-propanediol is mainly obtained through a hydrogenolysis of glycerol (1,2,3-propanetriol) (ADM, USA; BASF/Oleon, Belgium), the main by-product of the biodiesel industry [96]. This section focuses on diols-based, five- and six-membered cyclic carbonates as illustrated in Figure 4.6 that can potentially be obtained from biomass-derived and safe feedstocks. The two hydroxy groups attached to a neighbouring sp3 carbon atoms allow for reactions involving the inherent nucleophilicity or selective electrophilic activation of one of them. There are two main approaches for the direct conversion of diols and CO2 into cyclic carbonates: The dehydrative condensation strategy, where, alongside the catalyst responsible for substrate activation, the physical (non-reactive) or chemical (reactive) dehydrating agents are used in order to trap or consume water, respectively. ● The alkylation (leaving group) strategy, where, with the effect of a base, CO2 molecule is first activated and fixed onto one of the OH groups to create a hemi-carbonate ion, which subsequently reacts with alkyl/aryl halide via nucleophilic substitution and avoids the formation of water. ●

It is important to highlight that, at present, none of these methods have been industrialised. The dehydrative condensation pathway of cyclic carbonate synthesis Currently, the most efficient catalytic system for direct synthesis of various organic carbonates from CO2 and alcohols is the ground-breaking cascade catalyst of cerium oxide (CeO2) with 2-cyanopyridine (2-CP) [97–100]. The elegance of this approach is based on a simultaneous catalytic carbonylation of alcohol with carbon dioxide and hydration of 2-cyanopyridine (a water trap) to the corresponding amide, 2-picolinamide (2-PA). The hydrolysis of the nitrile with a water formed during the reaction promotes the formation of carbonate remarkably, leading to almost

47

48

4  New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources

Figure 4.6  Alcohol-based substrates for the production of cyclic carbonates from CO2.

100% yields. The recyclability of the drying agent is performed by the dehydration of 2-PA back to 2-CP over Na2O/SiO2 catalyst, although the regeneration step is rather slow and requires further optimisation [98]. Other shortcomings include harsh process conditions, a large excess of expensive 2-CP, sensitivity of the process to the size of ceria particles, and some deactivation of the catalyst due to the adsorption of 2-PA over CeO2 surface. Another example of nitrile widely investigated as a water trap (i.e. in combination with various zinc-based catalysts [101–104]) is acetonitrile (ACN). In the presence of water, ACN is converted to acetamide, which unfortunately continues to react to produce acetic acid, diol acetates, and ammonia. Additionally, small amounts of dipropylene glycol (dehydratation of 1,2-PDO) or products of 1,2-PC decomposition/polymerisation can be found at high temperature conditions. A severe drop in selectivity is not the only limitation of this strategy though, as ACN is an unwelcome solvent from the environmental point of view due to its toxicity combined with its significant volatility at room temperature (a boiling temperature of approximately 355K) [105]. The Environmental Protection Agency (EPA) has already denied, twice, petitions to remove acetonitrile from the Toxic Release Inventory (TRI) list due to its potential to cause neurotoxicity and death in humans and its contribution to the formation of ozone in the air [106]. Other chemical water-removal agents used for cyclic carbonate preparation include i.e., benzonitrile [107], propylene oxide [108], and ketals [109], but none of these shows superior

4.4  Cyclic Carbonates Derived from Bio-based Diols

Figure 4.7  The alkylation (leaving group) strategy of cyclic carbonates synthesis from polyols and CO2.

performance. Additionally, besides the overall selectivity penalty due to a plethora of parallel reactions, every chemical method adds at least one layer of complexity to the system. The reacted desiccant needs to be separated from the reaction mixture and either regenerated or sold on a suitable market. Physical techniques such as molecular sieves or zeolites are an alternative option since they do not lead to a formation of by-products and their application and regeneration is comparatively easy [110, 111]. However, they have been less explored due to the limitation of high operating temperature and pressure required for the conversion of alcohols. The surface hydroxyl groups in molecular sieves become acidic at high temperature conditions and may decompose the carbonate. A temperature swing, understood as increase in temperature for alcohol carboxylation interchanged with a temperature decrease for water absorption, is energetically unfavourable. The alkylation (leaving group) pathway of cyclic carbonate synthesis This method bypasses the co-production of water, but the involvement of additional components (in stoichiometric amounts) leads to a formation of ensuing by-products and overall low atom economy of the process. With alkyl/aryl halides as alkylating agents, halogenated waste is produced and this restricts even further the industrial application of the process. The advantages of the method are certainly mild process conditions (ambient temperatures and low [< 10 bar] CO2 pressure) and high yields obtained (60 to 90%) [85, 88]. The first step of such an organocatalytic approach relys on the formation of an ionic hemi-carbonate intermediate upon CO2 insertion, aided by a base responsible for the deprotonation of OH group of the reacting polyol (Figure 4.7). At the same time, strong organic bases such as 1,8-diazabicyclo[5,4,0]undec-7ene (DBU), 1,5-diazabicyclo[4.3.0]non-5-ene (DBN), or 1,5,7-triazabicyclo[4.4.0]dec-1-ene (TBD) are known to form a base-CO2 adducts, increasing the nucleophilic character of the CO2 molecule [112]. In the second step, the created reactive ion undergoes an alkylation by SN2 substitution, promoted by elimination of a suitable leaving group, resulting in ring-closure and production of the desired cyclic carbonate. Additional undesired side products may occur as a result of carbonation and subsequent alkylation of both alcohol moieties (formation of acyclic bis-carbonates), or in the case when the eliminated alcoholate group (the leaving group) takes up the CO2 and as a carbonate anion reacts via SN2 with the alkyl halide present in the reaction medium [113–115]. The competition between products depends on both the polyol substrate and the nature of the base. Buchard and co-workers established the two-step procedure where DBU and alkylating agent (tosyl chloride/triethylamine) were added stepwise [114], or alternatively, they demonstrated that the use of weaker bases such as triethylamine or 2,2,6,6-tetramethylpiperidine (TMP) makes it possible to synthesise various cyclic carbonates efficiently [115]. This is because even though the first step in the generation of the hemi-carbonate intermediate is strongly disfavoured in the presence of weaker bases, the following steps of tosylation and cyclisation are strongly favoured energetically.

49

50

4  New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources

4.5  Conclusions Bio-based substrates have become promising alternatives for the production of cyclic carbonates from CO2, creating an opportunity for the development of full renewable chemicals and materials and the exploration of more complex architectures. It makes sense to speak of opportunity because the chemistry explored is generally not new. It is our society’s current need for alternative sources of carbon and to shift from linear to circular economies that makes this field of research so crucial. It is important to have in mind that the synthesis of cyclic carbonates from CO2 is not about CO2 mitigation, as its direct potential contribution does not have the scale needed, but it is entirely related to the future development of an increasing sustainable chemical industry. Future research on this field should progress based on increasing the portfolio of bio-based substrates, focusing on those highly available, as well as on the development of new catalysts that allow to overcome typical low reactivities. It will also be important to expand the applications of versatile cyclic carbonates functionality, focusing on developing new, sustainable, and safer chemical routes.

Acknowledgements The authors acknowledge Fundação para a Ciência e Tecnologia FCT/MCTES through projects PTDC/EQU-EPQ/31926/2017 and MIT-EXPL/CS/0052/2021, projects UIDB/50006/2020, UIDP/50006/2020 and LA/P/0008/2020 of the Associate Laboratory for Green Chemistry – LAQV, and project UIDB/00100/2020 of the Centro de Química Estrutural – CQE. The NMR spectrometers are part of The National NMR Facility, supported by FCT (ROTEIRO/0031/2013 PINFRA/22161/2016) (co-financed by FEDER through COMPETE 2020, POCI, and PORL and FCT through PIDDAC).

References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Gao, W., Liang, S., Wang, R. et al. (2020). Chem. Soc. Rev. 49: 8584–8686. Machado, A.S.R., Nunes, A.V.M., and Nunes da Ponte, M. (2018). J. Supercrit. Fluids. 134: 150–156. Markewitz, P., Kuckshinrichs, W., Leitner, W. et al. (2012). Energy Environ. Sci. 5 (6): 7281–7305. Hepburn, C., Adlen, E., Beddington, J. et al. (2019). Nature 575 (7781): 87–97. Aresta, M. and Dibenedetto, A. (2020). Front. Energy Res. 8: 159. Kleij, A.W., North, M., and Urakawa, A. (2017). ChemSusChem. 10 (6): 1036–1038. Zhang, H., Liu, H.B., and Yue, J.M. (2014). Chem. Rev. 114 (1): 883–898. Shaikh, A.A.G. and Sivaram, S. (1996). Chem. Rev. 96 (3): 951–976. Rollin, P., Soares, L.K., Barcellos, A.M. et al. (2021). Appl. Sci. 11 (11): 5024. Schaffner, B., Schaffner, F., Verevkin, S.P., and Borner, A. (2010). Chem. Rev. 110 (8): 4554–4581. Pescarmona, P.P. (2021). Curr. Opin. Green Sustain. Chem. 29: 100457. Yadav, N., Seidi, F., Crespy, D., and D’Elia, V. (2019). ChemSusChem. 12 (4): 724–754. Garcia‐Garcia, G., Fernandez, M.C., Armstrong, K. et al. (2021). ChemSusChem. 14 (4): 995–1015. Tappe, N.A., Reich, R.M., D’Elia, V., and Kühn, F.E. (2018). Dalton Trans. 47 (38): 13281–13313. Kamphuis, A.J., Picchioni, F., and Pescarmona, P.P. (2019). Green Chem. 21 (3): 406–448. Rehman, A., Saleem, F., Javed, F. et al. (2021). J. Environ. Chem. Eng. 9 (2): 105113.

References

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

Aomchad, V., Cristòfol, À., Della Monica, F. et al. (2021). Green Chem. 23 (3): 1077–1113. Lopes, E.J., Ribeiro, A.P., and Martins, L.M. (2020). Catalysts 10 (5): 479. North, M., Pasquale, R., and Young, C. (2010). Green Chem. 12 (9): 1514–1539. Martín, C., Fiorani, G., and Kleij, A.W. (2015). ACS Catal. 5 (2): 1353–1370. Cokoja, M., Wilhelm, M.E., Anthofer, M.H. et al. (2015). ChemSusChem. 8 (15): 2436–2454. Büttner, H., Longwitz, L., Steinbauer, J. et al. (2017). Top. Curr. Chem. 375 (3): 50. Artz, J., Müller, T.E., Thenert, K. et al. (2018). Chem. Rev. 118 (2): 434–504. Liu, W., Duan, F., and Bi, Y. (2018). RSC Adv. 8 (23): 13048–13053. Kostyniuk, A., Bajec, D., Djinović, P., and Likozar, B. (2020). Chem. Eng. J. 394: 124945. Silvestre, A.J. and Gandini, A. (2008). Terpenes: major sources, properties and applications. In: Monomers, Polymers and Composites from Renewable Resources (ed. M.N. Belgacem, A. Gandini), 17–38. Amsterdam: Elsevier. Ciriminna, R., Lomeli-Rodriguez, M., Cara, P.D. et al. (2014). Chem. Commun. 50 (97): 15288–15296. Byrne, C.M., Allen, S.D., Lobkovsky, E.B., and Coates, G.W. (2004). J. Am. Chem. Soc. 126 (37): 11404–11405. Fiorani, G., Stuck, M., Martín, C. et al. (2016). ChemSusChem. 9 (11): 1304–1311. Pena Carrodeguas, L., González‐Fabra, J., Castro‐Gómez, F. et al. (2015). Chem. A Eur. J. 21 (16): 6115–6122. de la Cruz-Martínez, F., Martínez, M.S.B., Martínez, J. et al. (2019). ACS Sustain. Chem. Eng. 7 (24): 20126–20138. Martínez, J., Fernández‐Baeza, J., Sánchez‐Barba, L.F. et al. (2017). ChemSusChem. 10 (14): 2886–2890. Fernández-Baeza, J., Sánchez-Barba, L.F., Lara-Sánchez, A. et al. (2020). Inorg Chem 59 (17): 12422–12430. Longwitz, L., Steinbauer, J., Spannenberg, A., and Werner, T. (2018). ACS Catal. 8 (1): 665–672. Bähr, M., Bitto, A., and Mülhaupt, R. (2012). Green Chem. 14 (5): 1447–1454. Morikawa, H., Minamoto, M., Gorou, Y. et al. (2018). Bull. Chem. Soc. Jpn. 91 (1): 92–94. Rehman, A., Fernández, A.M.L., Resul, M.G., and Harvey, A. (2019). J. CO2 Util. 29: 126–133. Paninho, A.I. (2018). Sustainable Intensification Strategies for the Production of Cyclic Carbonates from CO2. Master’s thesis. NOVA University of Lisbon. Paninho, A.B., Mustapa, A.N., Mahmudov, K.T. et al. (2021). Catalysts 11 (8): 872. Lee, E.H., Ahn, J.Y., Dharman, M.M. et al. (2008). Catal. Today 131 (1–4): 130–134. Taherimehr, M., Sertã, J.P.C.C., Kleij, A.W. et al. (2015). ChemSusChem. 8 (6): 1034–1042. Chen, J., Wu, X., Ding, H. et al. (2021). ACS Sustain. Chem. Eng. 9 (48): 16210–16219. Kamphuis, A.J., Tran, M., Picchioni, F., and Pescarmona, P.P. (2022). Green Chem. Eng. 3 (2): 171–179. Darensbourg, D.J. and Yarbrough, J.C. (2002). J. Am. Chem. Soc. 124 (22): 6335–6342. Taherimehr, M., Al-Amsyar, S.M., Whiteoak, C.J. et al. (2013). Green Chem. 15 (11): 3083–3090. Giri, A.K. (1997). Mutat. Res./Rev. Mut. Res. 386 (1): 25–38. Bell, B.M., Briggs, J.R., Campbell, R.M. et al. (2008). CLEAN Soil. Air Water 36 (8): 657–661. Arrowood, T. (2008). Patent Application Publication Pub. No.: US 2008/0015369 A1.1. Kim, H.S., Yu, K., Puthiaraj, P., and Ahn, W.S. (2020). Microporous Mesoporous Mater. 306: 110432. Zalomaeva, O.V., Chibiryaev, A.M., Kovalenko, K.A. et al. (2013). J Catal 298: 179–185. Guo, L., Lamb, K.J., and North, M. (2021). Green Chem. 23 (1): 77–118. Abbasov, V.M., Nasirov, F.A., Rzayeva, N.S. et al. (2018). PPor 19 (4): 427. Cui, S., Qin, Y., and Li, Y. (2017). ACS Sustain. Chem. Eng. 5: 9014–9022.

51

52

4  New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources

54 Cui, S., Borgemenke, J., Liu, Z., and Li, Y. (2019). J. CO2 Util. 34: 40–52. 55 Tan, S.G. and Chow, W.S. (2010). Polym. Plast. Technol. Eng. 49 (15): 1581–1590. 56 Karak, N. (2021). Overview of Epoxies and their thermosets. In: Sustainable Epoxy Thermosets and Nanocomposites, 1–36. American Chemical Society. 57 Zhu, J., Chandrashekhara, K., Flanigan, V., and Kapila, S. (2004). J. Appl. Polym. Sci. 91 (6): 3513–3518. 58 Stemmelen, M., Pessel, F., Lapinte, V. et al. (2011). J Polym Sci A Polym Chem 49 (11): 2434–2444. 59 Natongchai, W., Pornpraprom, S., and D’Elia, V. (2020). Asian J. Organ. Chem. 9 (5): 801–810. 60 Schäffner, B., Blug, M., Kruse, D. et al. (2014). ChemSusChem. 7: 1133–1139. 61 Nohra, B., Candy, L., Blanco, J.F. et al. (2013). Macromolecules 46 (10): 3771–3792. 62 (a) Rokicki, G., Parzuchowski, P.G., and Mazurek, M. (2015). Polym. Adv. Technol. 26 (7): 707–761. (b) Maisonneuve, L., Lamarzelle, O., Rix, E. et al. (2015). Chem. Rev. 115 (22): 12407–12439. 63 UNION, P. (2009). Regulation (EC) No 1223/2009 of the European Parliament and of the council. Official J. Eur. Union L 342: 59. 64 (a) Alves, M., Grignard, B., Méreau, R. et al. (2017). Catal. Sci. Technol. 7 (13): 2651–2684. (b) Li, Z., Zhao, Y., Yan, S., Wang, X., Kang, M., Wang, J., Xiang, H. (2008). Catal. Lett. 123 (3): 246–251. 65 Miloslavskiy, D., Gotlib, E., Figovsky, O., and Pashin, D. (2014). Int. Lett. Chem. Phys. Astron. 20: 27–29. 66 Centeno-Pedrazo, A., Perez-Arce, J., Prieto-Fernandez, S. et al. (2021). Mol. Catal. 515: 111889. 67 Chen, F., Zhang, Q.C., Wei, D. et al. (2019). J. Org. Chem. 84 (18): 11407–11416. 68 Carrodeguas, L.P., Cristòfol, À., Fraile, J.M. et al. (2017). Green Chem. 19 (15): 3535–3541. 69 Arayachukiat, S., Kongtes, C., Barthel, A. et al. (2017). ACS Sustain. Chem. Eng. 5 (8): 6392–6397. 70 Carrera, G.V.S.M., Branco, L.C., and da Ponte, M.N. (2017). Bio-inspired systems for carbon dioxide capture, sequestration and utilization (chapter 5) In: Recent Advances in Carbon Capture and Storage (ed. Y. Yun). London: Intech Open. 71 Oh, K.-R., Valekar, A.H., Cha, G.-Y. et al. (2022). J. CO2 Util. 60: 101981. 72 Sun, J., Cheng, W., Yang, Z. et al. (2014). Green Chem. 16: 3071–3078. 73 Tamboli, A.H., Chaugule, A.A., and Kim, H. (2016). Fuel 184: 233–241. 74 Carrera, G.V.S.M., Branco, L.C., and da Ponte, M.N. (2015). J. Sup. Fluids 105: 151–157. 75 Carrera, G.V.S.M., Raymundo, A., Fernandes, F.M.B. et al. (2017). Carbohydr. Polym. 169: 58–64. 76 Eftaiha, A.F., Alsoubani, F., Assaf, K.I. et al. (2016). Carbohydr. Polym. 152: 163–169. 77 Sehaqui, H., Gálvez, M.E., Becatinni, V. et al. (2015). Environ. Sci. Technol. 49: 3167–3174. 78 Ding, G., Su, J., Zhang, C. et al. (2018). ChemSusChem. 11: 2029–2034. 79 Gregory, G.L., Ulmann, M., and Buchard, A. (2015). RSC Adv. 5: 39404–39408. 80 McGuire, T.M., López-Vidal, E.M., Gregory, G.L., and Buchard, A. (2018). J. CO2 Util. 27: 283–288. 81 Gregory, G.L., Jenisch, L.M., Charles, B. et al. (2016). Macromolecules 49: 7165–7169. 82 Chen, Z., Hadjichristidis, N., Feng, X., and Gnanou, Y. (2017). Macromolecules 50: 2320–2328. 83 Pati, D., Chen, Z., Feng, X. et al. (2017). Polym. Chem. 8: 2640–2646. 84 Muzyka, C. and Monbaliu, J.-C.M. (2022). ChemSusChem. 15: e202102391. 85 Brege, A., Grignard, B., Méreau, R. et al. (2022). Catalysts 12: 124–148. 86 Honda, M., Tamura, M., Nakagawa, Y., and Tomishige, K. (2014). Catal. Sci. Technol. 4: 2830–2845. 87 Li, J.‐.Y., Zhao, Q.‐.N., Liu, P. et al. (2018). Greenhouse Gas Sci. Technol. 8: 803–838. 88 Kindermann, N., Jose, T., and Kleij, A.W. (2017). Top. Curr. Chem. 375: 15. 89 Sheldon, R.A. (2014). Green Chem. 16 (3): 950–963.

References

90 de Jong, E., Stichnothe, H., Bell, G., and Jorgensen, H. (2020). Bio-based chemicals : a 2020 update, IEA Bioenergy. Task 42: 2020: 01, Wageningen, IEA; https://www.ieabioenergy.com/ wp-content/uploads/2020/02/Bio-based-chemicals-a-2020-update-final-200213.pdf(accessed 26 September 2022). 91 Rosales-Calderon, O. and Arantes, V.A. (2019). Biotechnol. Biofuels 12: 240–299. 92 Zakrzewska, M.E., Bogel-Łukasik, E., and Bogel-Łukasik, R. (2011). Chem. Rev. 111: 397–417. 93 Biddy, M.J., Scarlata, C., and Kinchin, C. (2016) Chemicals from biomass: a market assessment of bioproducts with near-term potential. National Renewable Energy Laboratory Tech. Rep. https://www.nrel.gov/docs/fy16osti/65509.pdf (accessed 29 September 2022). 94 Yim, H., Haselbeck, R., Niu, W. et al. (2011). Nature Chem. Biol. 2011: 445. 95 Burgard, A., Burk, M.J., Osterhout, R. et al. (2016). Curr. Opin. Biotechnol. 42: 118. 96 Kong, P.S., Aroua, M.K., and Daud, W.M.A.W. (2016). Renew. Sus. Energ. Rev. 63: 533–555. 97 Honda, M., Tamura, M., Nakao, K. et al. (2014). ACS Catal. 4: 1893–1896. 98 Honda, M., Tamura, M., Nakagawa, Y. et al. (2013). ChemSusChem. 6: 1341–1344. 99 Tamura, M., Ito, K., Honda, M. et al. (2016) Sci. Rep. 6: 24038. 100 Honda, M., Tamura, M., Nakagawa, Y. et al. (2014). J. Catal. 318: 95–107. 101 Huang, S., Liu, S., Li, J. et al. (2007). Catal. Lett. 118: 290–294. 102 Zhao, X., Sun, N., Wang, S. et al. (2008). Ind. Eng. Chem. Res. 47: 1365–1369. 103 Huang, S., Liu, S., Li, J.-P. et al. (2007). J. Fuel. Chem. Technol. 35: 701–705. 104 Comerford, J., Hart, S., North, M., and Whitwood, A.C. (2016). Catal. Sci. Technol. 6: 4824–4831. 105 O’Neil, M.J. (ed.) (2013). The Merck Index – an Encyclopedia of Chemicals, Drugs, and Biologicals, vol. 14. Cambridge, UK: Royal Society of Chemistry. 106 Federal Register/Vol. 78, No. 43 / Tuesday, March 5, 2013 / Proposed Rules. 107 Da Silva, E., Dayoub, W., Mignani, G. et al. (2012). Catal. Commun. 29: 58–62. 108 Diao, Z.-F., Zhou, Z.-H., Guo, C.-X. et al. (2016). RSC Adv. 6: 32400–32404. 109 Du, Y., Kong, D.-L., Wang, H.-Y. et al. (2005). J. Mol. Catal. A Chem. 241: 233–237. 110 Aresta, M., Dibenedetto, A., Nocito, F., and Pastore, C. (2006). J. Mol. Catal. A Chem. 257: 149–153. 111 George, J., Patel, Y., Pillai, S.M., and Munshi, P. (2009). J. Mol. Catal. A Chem. 304: 1–7. 112 Villiers, C., Dognon, J.-P., Pollet, R. et al. (2010). Angew. Chem. Int. Ed. 49: 3465–3468. 113 Lim, Y.N., Lee, C., and Jang, H.Y. (2014). Eur. J. Org. Chem. 2014: 1823–1826. 114 Gregory, G.L., Ulmann, M., and Buchard, A. (2015). RSC Adv. 5: 39404–39408. 115 McGuire, T.M., López-Vidal, E.M., Gregory, G.L., and Buchard, A. (2018). J. CO2 Util. 27: 283–288.

53

55

5 Sustainable Technologies in CO2 Utilization The Production of Synthetic Natural Gas M. Carmen Bacariza, José M. Lopes, and Carlos Henriques Centro de Química Estrutural, Institute of Molecular Sciences, Departamento de Engenharia Química, Instituto Superior Técnico, Universidade de Lisboa, Avenida Rovisco Pais, Lisboa, Portugal

5.1  CO2 Valorization Strategies The Covid-19 pandemic and the consequent lockdown were responsible for a temporary reduction of approximately 5% in the CO2 emissions during the first quarter of 2020 [1]. This event was linked to the reduction of the global demand for energy, mainly in the USA, China, and the European Union [1]. However, CO2 emissions rebounded to 2018–2019 levels during 2021, partially due to the increasing energy demand of emerging markets [2]. Therefore, it is still necessary to implement strategies to reduce them in the near future. The identification and analysis of the availability of carbon dioxide emissions abatement approaches from several sources (e.g. cement production, fuel combustion, ships) has been focus of numerous research studies in recent years [3–7]. It is clear that the capture and utilization of CO2 to produce chemicals or fuels could be considered the main pathway to be followed. In recent years, several studies have dealt with the reduction of CO2 into chemicals or fuels using surplus renewable energy [8–24]. Indeed, it was reported that this strategy could supply energy and simultaneously protect the environment as the use of CO2 as a starting carbon source can help with the required emission cuts while the short-chain hydrocarbon products such as CH4, CH3OH, or even C2H6 could be used as renewable energy sources to alleviate the increasingly tense energy crisis [8–24]. The most researched area for the conversion of CO2 into fuels has been hydrogenation [25–36], with the main target products being carbon monoxide, methane, hydrocarbons, methanol, dimethyl ether, higher alcohols, formic acid/formates, or even formamides. The main drawback of this route is the high amount of hydrogen required. In this sense, many authors have agreed that CO2 recycling by conversion through hydrogenation will only be feasible if hydrogen is obtained from renewable energy sources, like solar energy or wind, via water electrolysis [37]. In fact, unless water splitting using solar energy or other similar processes is implemented, hydrogen will be produced mainly from carbonaceous materials with the co-generation of CO2, without a beneficial effect from an environmental perspective [27]. Some products of CO2 hydrogenation, such as methanol, dimethyl ether (DME) and methane, are excellent fuels in internal combustion engines,

Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 1, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

56

5  Sustainable Technologies in CO2 Utilization

while also being favorable from the perspective of storage and transport. In addition, methanol and formic acid are critical raw materials and intermediates for many chemical industries [38]. Every CO2 hydrogenation reaction requires the use of specific catalytic systems. Although both homogeneous and heterogeneous catalysts have been studied for these reactions, the homogeneous catalysts present problems in terms of recovery and regeneration. Heterogeneous catalysts are therefore preferable in terms of stability, separation, handling, and reuse, as well as for reactor design, resulting in lower costs for large-scale productions. For this reason, the next section describes the main processes of CO2 hydrogenation focusing mainly on the use of heterogeneous catalysts.

5.1.1  CO2 to CO via Reverse Water-Gas Shift (RWGS) Reaction The production of CO from CO2 through the reverse water-gas shift (RWGS) reaction (Reaction 5.1) is one of the most promising routes for CO2 conversion and takes place in many processes where CO2 and H2 are present.

CO2 �( g ) + H� 2 ( g ) �  CO( g ) + � H 2O(v)

(5.1)

Because this a reversible process, catalysts with activity toward the direct reaction are often suitable for the reverse one [39]. In fact, metal-supported catalysts containing transition (mainly Cu) or noble (mainly Pt and Pd) metals as active phase and Al2O3, SiO2, CeO2, TiO2, ZrO2, or even zeolites as supports have been the most widely applied [40]. Overall, the main aspects affecting the catalytic performances of RWGS catalysts and their ability to improve CO2 adsorption and CO desorption are: (i) varying the support nature; (ii) adding promoters such as reducible transition metal oxides or alkali metals; (iii) tuning the metal-support interactions; or (iv) inducing confinement effects for the active metal particles [40]. In the case of Cu-supported catalysts, some combinations of Cu–Ni and Cu–Zn over Al2O3 have been studied with the conclusion that the selectivity toward CO depends mainly on the Cu/metal ratio [41, 42]. As RWGS is an endothermic reaction, high temperatures would facilitate the formation of CO. However, copper nanoparticles would suffer from sintering and deactivation under these conditions. For these reasons, the use of thermal stabilizers such as iron [43, 44] or even alkaline promoters such as K2O has revealed positive effects on Cu-based catalysts [45]. Furthermore, noble metals (e.g. Pt, Ru, and Rh) have been also studied as active metals for this reaction as they typically present high ability toward H2 dissociation. The same types of promoters identified for Cu-supported catalysts have been reported for these systems. For instance, Li/Rh-Y zeolites with Li/Rh ratios higher than 10 presented CO as main product (87% of selectivity) with a selectivity to methane lower than 10% [46]. The presence of Li atoms on the surface was responsible for the creation of new active sites for CO2 adsorption [46].

5.1.2  CO2 to CH4 CO2 methanation (Sabatier reaction; Reaction 5.2), first discovered in 1902, is an important catalytic process with a range of applications including the production of synthetic natural gas (SNG) [38]. Indeed, the U.S. National Aeronautics and Space Administration (NASA) investigated the conversion of the Martian CO2 atmosphere into methane and water for fuel and astronaut life-support systems [47]. Even if the conversion of CO2 into CH4 is a thermodynamically favorable reaction, the reduction of the fully oxidized carbon to methane is an eight-electron process with

5.1  CO2 Valorization Strategies

significant kinetic limitations and, as a result, requires a catalyst to achieve acceptable rates and selectivity [38].

CO2 �( g ) +� 4 H 2 �( g ) �→ � CH 4 �( g ) + 2 H 2O(v)

(5.2)

Unlike the conversion of CO into CH4, which is already a well-established process at the industrial level, CO2 methanation has only become the focus of many studies in recent decades. This interest in using CO2 as feedstock is related to the increasing concerns about CO2 emissions combined with the increase of fuel prices. As observed in the literature [48–57], CO is the main by-product of this reaction at higher reaction temperatures (RWGS; above 400 ºC). CO2 methanation has been focus of hundreds of research studies in the last years, with thermal, photocatalytic, electrochemical or even plasma assisted processes being used [48–57]. Ni and Ru-based materials have been commonly reported for this reaction. In terms of supports, SiO2, Al2O3, Ce and Zr oxides, mesoporous materials, carbons, hydrotalcite-derived materials, and even zeolites have been applied [48–53, 58].

5.1.3  CO2 to CxHy The production of hydrocarbons from CO2 hydrogenation is generally considered to be a modification of the Fischer-Tropsch (FT) process, in which syngas (CO and H2) is transformed into hydrocarbons. The catalysts used in these reactions are usually present in compositions analogous to those used in conventional FT synthesis, but modified to achieve higher selectivity for the desired hydrocarbons and to improve CO2 activation. Several alternatives, direct or indirect, have been reported for this process [38, 58–63]. Indeed, while indirect routes are mediated by methanol (CO2-to-CH3OH followed by CH3OH-to-CxHy) or carbon monoxide (CO2-to-CO via RWGS followed by the FT process), direct routes are CO2-based FT (with CO intermediates) or use bifunctional catalysts (with methanol intermediates) [62]. On one hand, in methanol-mediated processes, CO2 and H2 react typically over Cu–Zn-based catalysts to produce methanol (see Section 5.1.4), which is later transformed into hydrocarbons by Methanol-to-CxHy pathways [59]. These Methanol-to-CxHy processes can lead to the production of different types of hydrocarbons, such as olefins, gasoline, branched alkanes, or aromatics. Acidic zeolites are the most commonly used catalysts, and their framework type and composition as well as the chosen operation conditions significantly influence the selectivity of the product [62]. The main features of zeolite catalysts are their shape-selectivity, dimensional structure, stability, and acidic properties, which can be easily tuned through post-synthesis treatments (e.g. dealumination, desilication) or even by varying the nature of the compensating cation. Based on the literature, medium-pore zeolite/microporous materials can generate C5–C11 hydrocarbons, whereas small-pore molecular sieves yield C2–C4 hydrocarbons. In terms of commonly applied zeolites, the desired product will determine which features should be ensured. Indeed, while gasoline is typically obtained when using ZSM-5, SAPO-34 presents high selectivity toward light olefins. On the other hand, for CO-mediated processes, carbon monoxide is typically formed by RWGS reaction (see Section 5.1.2), followed by FT synthesis. The main catalysts used in FT synthesis are cobalt-based, characterized by their high performance/cost ratio, and iron oxides, which usually lead to highly olefinic products and are also active for WGS and RWGS reactions. Iron oxide-based catalysts usually present some promoters such as potassium, which acts as an electronic promoter to iron and can form species able to act as reversible H2 reservoirs, suppressing the hydrogenation of the products; manganese, which acts as both electronic and structural promoter inhibiting the

57

58

5  Sustainable Technologies in CO2 Utilization

formation of CH4 and favoring the dispersion of Fe2O3 as well as the surface basicity; copper, which presents a similar performance than manganese favoring the reducibility of the catalyst and providing new additional active sites for H2 dissociation; or others such as CeO2 (known to be highly active in the WGS reaction at low temperature), zirconium, zinc, magnesium, ruthenium, and lanthanum [64]. In terms of supports, iron-based catalysts tend to act purely as stabilizers to avoid sintering processes during the reaction that might lead to a decrease of the catalyst activity [65]. The most studied supports have been alumina, silica, and titania [38]. In addition, zeolites have also been reported as suitable for use as this kind of catalyst due to their surface characteristics in terms of porosity and inner electric fields. The effects of some zeolite properties (e.g. acidity, type of structure) on catalytic performances have been the focus of studies in the literature [66]. To summarize, and besides the effects of the operating conditions (e.g. reaction temperature and pretreatment pressure), the type of support or iron precursors significantly affect the selectivity of the hydrocarbons [64]. These properties are improved by adding alkali metal or transition metal promoters, as previously mentioned [64].

5.1.4  CO2 to CH3OH Methanol is currently used in the chemical industry as a solvent, alternative fuel, and starting material. Thus, the utilization of CO2 instead of CO as a precursor is considered as an effective method of carbon dioxide utilization. This process (Reaction 5.3) produces CO, hydrocarbons, and higher alcohols as the main by-products and these are favored by high pressures and temperatures [38].

CO2 �( g ) + 3 H� 2 ( g ) � → � CH3OH(v) +� H 2O(v)

(5.3)

In terms of the composition of catalysts, Cu remains as the main active metal, typically combined with ZnO, responsible for enhancing Cu dispersion and CO2 activation, even though several noble metals (e.g. Ag, and Au) have been studied [67–69]. Thus, Cu/ZnO catalysts, prepared using different methods (e.g. impregnation, co-precipitation), Cu/Zn ratios, or even ­calcination/reduction temperatures, are the most commonly used for methanol synthesis ­[70–72]. Even if Al2O3 incorporation improves thermal and chemical stability, adding ZrO2 instead was found to be more promising due to its higher H2O tolerance. Additionally, dopants such as Mo, Fe, Ti, V, La, or Pd were found to promote CO2 adsorption/activation and/or H2 dissociation [70]. Overall, controlling Cu species oxidation state during the reaction, identifying the optimal metal dispersion, maximizing the metal-oxide interface, understanding the role of oxygen vacancies on the adsorption properties, analyzing the cooperative effect among neighboring sites, or building kinetic models for a precise understanding of reaction mechanisms are current challenges [70, 72, 73].

5.1.5  CO2 to CH3OCH3 Dimethyl ether (DME; the simplest ether in nature, gaseous at ambient conditions and liquid at ≈5 bar) can be used in the household, for transportation, for power generation, and even as chemical feedstock. As DME exhibits a high cetane number and combustion efficiency, is sulfur-free, and leads to minor NOx, CO, or soot formation, it constitutes a promising diesel substitute [73, 74]. There are two routes for the production of DME from CO2 hydrogenation [38]: a two-step process (methanol synthesis followed by a step of dehydration on an acid catalyst such as γ-Al2O3, H-ZSM-5, or NaH-ZSM-5) and a single-step process in which multifunctional catalysts are used to perform the two steps simultaneously.

5.1  CO2 Valorization Strategies

With respect to the two-step strategy, the synthesis of methanol has been already analyzed in Section 5.1.4. However, it is important to note that the activity of the acidic supports used for methanol dehydration (and even for the one-pot strategy) depends on their interaction with the water formed during the reaction [38]. In the case of γ-Al2O3, the activity decreases in presence of water because the acid sites present high water adsorption capacity [75]. As H-ZSM-5 zeolite is not as sensitive to the concentration of water, it could be suitable for bifunctional catalysts to be applied in the one-pot alternative. However, H-ZSM-5 zeolite is also active for the transformation of DME into hydrocarbons, which could partially evolve to coke and block zeolite pores, causing deactivation. In this context, a suitable concentration of Na+ (compensating cation) has been reported to act as a moderator of the number and strength of Brønsted acid sites, preventing the formation of hydrocarbons [76, 77]. Although the two-step strategy allows higher DME purity without any particular issues related to water formation, the use of two separated reactors increases the complexity and cost of the process [71]. Different strategies can be used to integrate the two functionalities of methanol synthesis (mixed-oxides; see Section 5.1.4) and methanol dehydration (solid acids) for obtaining CO2-toDME catalysts [72]. Among them, the physical or mechanical combination by dry powder mixing or grinding presents limitations such as the heterogeneous distribution of active sites or mass transfer constraints. Thus, strategies in which the active components are chemically generated during preparation, promoting the contact through spatial proximity for an easier activation of H2 and CO2, become more interesting. Recent studies have focused on optimizing the interactions between both types of active sites; extending the lifetime of catalysts and resistance to metal site oxidation, sintering, and coke deposition; or maximizing carbon utilization efficiency [71, 72, 78– 84]. Indeed, catalysts deactivation is one of the main issues to overcome as there are few publications devoted to this topic [72, 73].

5.1.6  CO2 to R-OH The production of higher alcohols such as ethanol, 1-butanol, 2-butanol, or 1-octanol leads, when compared with methanol, to advantages in terms of transportation, toxicity, and compatibility with gasoline. The synthesis process is, in many cases, the combination of the RWGS reaction (CO2-to-CO) and the subsequent formation of higher alcohols from CO and H2. The main catalysts used for this reactions are Fe-based and Rh-based, although Cu, Co, or Mo-containing systems have been also reported [36, 38, 85–87]. As expected, the desired product will define the suitability of the chosen catalytic system. For instance, in the case of ethanol production, Fe-based FT catalysts mixed with Cu-containing systems are primarily used because of their partial reduction of CO2 into CO. Other metals and compounds such as Pd, Ga, or potassium carbonate have been reported useful. Furthermore, Rh-Se/TiO2 catalytic systems were reported as able to obtain selectivity to ethanol above 80% at low pressures [38]. With respect to butanol synthesis, aldol condensation and oxo synthesis pathways have primarily been reported. No suitable synthetic pathway is known currently for 1-­octanol production [87]. Moreover, the synthesis process for higher alcohols synthesis leads to a mixture of products that needs to be separated for downstream processing, increasing the operation costs. Thus, the development of synthesis of a single higher alcohol with high selectivity from CO2 hydrogenation is a future direction for research [86]. In addition, further research studies must be devoted to developing new bifunctional active sites able to improve carbon chain propagation and alcohol formation; tailoring the active sites at an atomic scale, as atomically dispersing metals have better performances; and understanding the

59

60

5  Sustainable Technologies in CO2 Utilization

effect of the solvent in a slurry bed reactor, the selectivity to higher alcohols is generally higher when these reactions are performed in this type of reactor rather than a fixed bed one [86].

5.1.7  CO2 to HCOOH, R-COOH, and R-CONH2 Formic acid (HCOOH) is currently used in the leather and rubber industries and as feedstock for the production of chemicals as fiber or sweetener. Furthermore, this compound has been considered as a hydrogen storage material through the combination of CO2 hydrogenation with selective formic acid decomposition. Formic acid and formate (R-COOH) synthesis proceed mainly with organometallic complexes at low temperatures with rhodium, ruthenium, iridium, or palladium being the main active metals used [31, 38, 88]. On the other hand, a step toward green formylation of amines is the use of CO2 and H2 instead of toxic compounds such as CO and phosgene as formylation agents. As in the case of formic acid and formates, mainly homogeneous systems, based on ruthenium complexes, have been developed so far for formamides (R-CONH2) synthesis from CO2, H2, and secondary amines [38].

5.1.8  Target Products Analysis Based on Thermodynamics Among all of the alternatives for CO2 hydrogenation, methanation has been considered as the most thermodynamically favorable. Indeed, Jia et al. [89] reported a complete study regarding the thermodynamics of different CO2 hydrogenation reactions producing carbon monoxide, carboxylic acids, aldehydes, alcohols, and hydrocarbons based on the principle of Gibbs free energy minimization and the results were validated with experimental data. The Gibbs free energy and standard equilibrium constants for the most common CO2 hydrogenation reactions can be found in Table 5.1. Table 5.1  Gibbs free energy and standard equilibrium constants in CO2 hydrogenation reactions. Adapted from [89]. Reaction

Product

ΔGº (kJ/mol)

Kº

CO2 → CO

Methane (CH4)

−113.5

7.79·1019

Carbon monoxide (CO)

28.6

9.67·10–6

Formic acid (HCOOH)

43.5

2.43·10–8

CO2 → CH4

CO2 → Carboxylic acids CO2 → Aldehydes CO2 → Alcohols

CO2 → Hydrocarbons

Acetic acid (CH3COOH)

−21.6

6.11·103

Propionic acid (C2H5COOH)

−32.6

5.17·105

Butyric acid (C3H7COOH)

−38.5

5.47·106

Formaldehyde (HCHO) Acetaldehyde (CH3CHO) Methanol (CH3OH) Dimethyl ether (CH3OCH3) Ethanol (C2H5OH) Acetylene (C2H2)

55.9 −12.9 3.5 −4.9 −32.4 41.8

1.63·10–10 1.86·102 2.45·10–1 7.15 4.70·105 4.69·10–8

Ethylene (C2H4)

−28.7

1.07·105

Propylene (C3H6)

−42.1

2.34·107

Propane (C3H8)

−70.9

2.64·1012

Ethane (C2H6)

−78.7

6.26·1013

5.2  Power-to-Gas: Sabatier Reaction Suitability for Renewable Energy Storage

Overall, authors have concluded that the conversion of CO2 into CO requires elevated temperatures and low H2:CO2 ratios to obtain high catalytic performances. Additionally, methanol synthesis was found to be favored by relatively high pressures to minimize the occurrence of RWGS reaction. Also, formic acid or formaldehyde are thermodynamically limited. Furthermore, complex reactions leading to several carbon-based products were reported and, among the different carboxylic acids, propionic acid was found to be the most favorable (selectivity above 90%). In terms of CO2 hydrogenation into aldehydes and alcohols, the major products were propionaldehyde and butanol, respectively [89]. On the contrary, the conversion of CO2 into CH4 is the thermodynamically easiest reaction with nearly 100% CH4 yield under moderate conditions [89]. The effects of some conditions such as the temperature, pressure, and H2:CO2 ratio on the selectivity to CH4 instead of CO have been reported in the literature [89–91]. Indeed, it was found that the equilibrium CO yield increases with temperature for any given feed composition due to the endothermic characteristics of the RWGS reaction. Additionally, higher pressures (up to 100 bar) favored CO2 methanation. Moreover, CO2 conversion into CH4 is favored by higher H2:CO2 ratios, both at lower (1 bar) and higher (30 bar) pressures. Gao et al. [90] also studied the effect of steam incorporation in the reactor feed and found that, both at 1 and 30 bar, CO2 conversion was reduced without a significant impact in CH4 selectivity and carbon deposition was hindered. The authors, who verified the same effects under CO methanation conditions, attributed these results to the inhibitory effect of H2O in the reaction.

5.2  Power-to-Gas: Sabatier Reaction Suitability for Renewable Energy Storage The composition of natural gas, even if slightly modified by its origin, has CH4 as the main component (~90 vol.%), with ethane and propane as the other two more significant components (~6 and ~2 vol.%, respectively) [92]. The existence of an established and solid network of natural gas allows the incorporation of the methane produced from CO2 hydrogenation with green hydrogen (produced from renewable energy by water electrolysis) to the existing grid (Power-to-Gas). The use of the gas pipelines is an order of magnitude larger than electrical power lines so that, in the end, natural gas network is more suitable for storing renewable energy [93]. Furthermore, if CO2 methanation is implemented in industries with large CO2 emissions (e.g. cement or power plants) that can use natural gas for combustion processes, the produced methane could be re-used, promoting circularity and decreasing CO2 emissions in these sectors. Blanco et al. [94] published a detailed review regarding Power-to-X (X = heat, liquids, chemicals, fuels, and mobility) technologies where the production of H2 through electrolysis and its subsequent conversion to methane with CO2 from different sources (e.g. carbon capture, biogas, air) was included. Additionally, Schaaf et al. [95] reported on the suitability of CO2 methanation as an energy storage strategy, as it presents a high storage capacity combined with high charge/discharge periods (Figure 5.1). The overall Power-to-Gas concept is summarized in Figure 5.2. As shown, this is a three-step process involving: (i) the generation of renewable electricity; (ii) the production of renewable hydrogen by water electrolysis using excess renewable electricity; and (iii) the methanation of carbon dioxide followed by CH4 injection into the natural gas grid [96]. Regarding Power-to-X projects in Europe, Wulf et al. [97] reported in 2020 that, although the maximum number of commissioned plants was reached in 2018, the capacities of the installed electrolyzers for H2 production are still increasing. The authors verified that methanation is the main strategy used, with a few projects focusing on the production of liquid fuels [97]. The need

61

62

5  Sustainable Technologies in CO2 Utilization

Figure 5.1  Storage capacity and discharge time of several energy storage technologies (CAES: compressed air energy storage; PHS: pumped hydro storage; SNG: synthetic natural gas). [95] / Springer Nature / CC BY-4.0.

Figure 5.2  Power-to-gas concept. Reproduced with permission from Ref [96].

for valorizing oxygen, as a co-product of water electrolysis for H2 production, was also highlighted. The commercialization of the technologies was found to be closer, which was considered to be an opportunity for new countries such as Portugal and those in eastern Europe to participate in Power-to-X projects. Although 220 projects in 20 different countries were identified in Europe, most of them were based in Germany and France, which plan to install around 500 MW of capacity by 2025 [97]. In terms of a global analysis, Thema et al. [98] analyzed the status of electrolysis and methanation in 2019, with a special focus in central Europe, as seen in Figure 5.3. The authors agreed with the exponential development of the technology both for methanation and hydrogen production, concluding that market implementation is under way. Most of the existing pilot plants were of mean size and had short lifetimes (1–3 years). However, large-scale and mid/long-term plants were reported as planned in the northwest of Germany, Hauts-de-France, and Normandie. As suggested by previous authors, Thema et al. insisted on the need for oxygen (from water electrolysis) and heat (from methanation reaction) valorization to improve the efficiency of the process.

5.3  CO2 Methanation Catalysts

Figure 5.3  Power-to-Gas projects. [98] / Elsevier / CC BY-4.0.

5.3  CO2 Methanation Catalysts The efficiency of the Sabatier reaction is highly dependent on the catalysts used, whose nature and features were the focus of several reviews in recent years [50, 51, 54, 56, 99–102]. Although nickelbased supported catalysts are the most widely used, their tendency to suffer from sintering and deactivation processes under relatively mild conditions drove a desire to identify suitable promoters/stabilizers and/or alternative active metals. In terms of the modification of Ni-catalysts, primarily rare earth or alkali/alkali-earth metal oxides have been found to be suitable for promoting resistance to deactivation [47]. Regarding the use of alternative metals, Ru has reported interesting features and, despite its higher cost, these catalysts present high resistance to deactivation even in the presence of oxygen or water in the reactor feed [56]. Furthermore, other transition metals such as Fe or Co were applied in this reaction. However, their performance has been lower than the performance given by equivalent Ni systems, with harsher operating conditions (namely, pressures above atmospheric) being required for improving activity and selectivity. Considering that most of the catalysts used are metal-supported ones, the nature of the chosen support significantly influences their properties and performances. In this sense, supports can influence the metallic dispersion and can present active sites for the activation of CO2 or even H2. Among all, Al2O3 and SiO2 are the most used supports due to their lower price, even if they usually suffer from deactivation due to carbon deposition and sintering of metal particles. In this context, the development of mesoporous alumina and silica were the subject of various studies, primarily due to the potential establishment of confinement effects able to limit the growth of metal particles, improving metallic dispersion and even hindering the occurrence of sintering processes. The same types of features were reported by micro- and/or mesoporous zeolites, carbons, or even metal-organic frameworks (MOF), which have also reported to be interesting for this reaction. With respect to zeolites, their easily tunable properties in terms of textural properties, basicity, or even hydrophobicity were crucial for identifying the most favorable properties and optimizing the formulation of catalysts. In addition, CeO2 has been used as a support in many of the catalysts in the literature, primarily due to its ability to activate carbon dioxide (oxygen vacancies) and its

63

64

5  Sustainable Technologies in CO2 Utilization

beneficial effect on the metallic dispersion, reducibility, and metal-support interactions. Other oxides such as ZrO2, TiO2, and MgO have also revealed interesting properties as supports for CO2 methanation catalysts, once again due to their promotion of CO2 activation and enhancement of metal-support interactions. Finally, interesting results were also obtained when using hydrotalcitederived materials, especially due to the well-known basicity of these systems. Overall, the most interesting characteristics that should be guaranteed in CO2 methanation catalysts are related to the properties of the active metals (e.g. dispersion and metal-support interactions), the CO2 activation capacity, and the reduction of the inhibitory effect of water in the reaction [56]. These properties will be summarized in the following paragraphs. Beginning with the properties of the active metals, the formation of small (60% CO conversion). To maximize the use of cobalt metal, most FTS catalysts contain nanoparticles of cobalt anchored onto high surface area refractory supports. Although a number of cobalt phases (carbidic, oxidic, mixed cobalt-promoter, and support phases) can co-exist during realistic FT synthesis conditions, the major phase is metallic cobalt and is widely considered to be the active phase for FTS [10]. The size, uniformity of dispersion, phase, and stability of the metallic cobalt nanoparticles plays an important role in the performance of the catalyst with respect to activity, stability, and selectivity. Recent studies have indicated that CoO in conjunction with reducible oxides such as MnOx and TiOx, can act as active sites for CO2 hydrogenation, a reaction that does not readily occur over metallic cobalt catalysts [19, 20]. Whereas cobalt carbide is mostly considered to be inactive for CO dissociation, its formation is typically associated with lower reaction rates and higher methane formation [21, 22]. Reports have indicated that depending on the morphology [23], and chemical environment of the carbide phases, they can act as catalysts for olefin or oxygenate formation [19]. Iron-based catalysts are employed in three modes, these being low temperature (c. 210–240 °C), medium temperature (c. 270–290 °C) [24], and high temperature (c. 330–350 °C) (see Table 6.1). Depending on the reaction temperature employed, various product distributions can be obtained. The CO2 conversion efficiency and build-up of carbon on the catalyst will also be impacted by reaction temperature. The active phase of Fe-based catalysts for CO hydrogenation is considered to be Fe-carbide [25, 26] as this phase has optimal bond breaking and bond formation strength for FTS. Fe-based catalysts can directly convert CO2 into products via RWGS, and literature consensus is that CO2 conversion to CO occurs on Fe-oxide phases also present in the working catalyst [27]. The preparation of Fe-based catalysts can vary significantly depending on their application. For example, fusion of iron oxide together with relevant structural and chemical promoters, similar to ammonia synthesis catalyst manufacture, is used to produce a catalyst suitable for the large scale Secunda and Mosselbay HTFT plants, whereas precipitation methodology is used for the LTFT Sasolburg site and is also believed to be the preferred route for the preparation of Synfuels China’s slurry phase MTFT iron catalyst [28]. The various reactor types applied in commercial scale conventional FT applications are shown in Figure 6.4. These consist of both stationary and moving bed reactors and a specific catalyst morphology needs to be targeted for each specific reactor. Currently the most widely used reactors are fixed beds, slurry beds and fixed-fluidized beds. These conventional FT reactors are designed to work on a large scale, and traditional FT plants are built to process substantial amounts of synthesis gas from a single large gas or coal reserve. These plants require large capital investment: for example Shell’s Pearl plant, including both FTS production and the upstream natural gas condensates, was built at a cost of around $20 billion [29]. The viability of these large plants is also dependent on factors such as the natural gas price and economy of scale.

77

5

10

20 25

30

Carbon number

15

35 40 45

Typical product Mainly straight chain paraffins profile and with negligible amounts of characteristics oxygenated, branched, and aromatic components; most suitable for diesel fuel, wax, and synthetic lubricant production

0

0

5

10

15 20 25 30 Carbon number

35 40 45

Mainly normal paraffins (saturated, straight chain hydrocarbons) typically produces a heavy product slate (wax) with small amounts of olefins, alcohols, and acids; very high alpha (0.95); most suitable for hard wax production

0

0.005

0.01

0.015

0.02

5

10

15 20 25 30 Carbon number

35

40 45

Mostly n-paraffins with small amounts of branched paraffins, olefins, and oxygenates; wax production

0

0

5

10

20

25

30 Carbon number

15

35

40 45

Mainly olefin and aromatic hydrocarbons with very small amounts of heavier (wax) products and high C1–C3 gas yield and large amounts of oxygenated compounds; gasoline and chemicals applications

0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0

Fixed-fluidized bed and circulating fluid bed (CFB)

Slurry bed

Slurry bed and fixed bed multi-tubular

Slurry bed and fixed bed multi-tubular

Reactor type 0.04 0.035 0.03 0.025 0.02 0.015 0.01 0.005 0

Sasol Secunda Operations South Africa (160 000b/d), PetroSA, South Africa (22 000b/d)

Synfuels China CTL (various locations in China) (>150000 b/d est.)

Sasol Sasolburg operations, South Africa (7500b/d)

ORYX-GTL, Qatar (34 000b/d) Shell Pearl GTL, Qatar (140 000 b/d)

Example of commercial applications

0.25

Coal & Natural gas

Coal

Natural gas

Natural gas

Synthesis gas-feed

0.035 0.03 0.025 0.02 0.015 0.01 0.005 0

Fused and milled

Precipitated and spray-dried

Precipitated, spray-dried or extruded

Supported and precipitated, spray-dried, or extruded

Catalyst preparation route

Carbon number distribution

HTFT (330–350 °C)

MTFT (270–290 °C)

LTFT (210–240 °C)

LTFT (190–240 °C)

Temperature classification

Mass Fraction

Fe-based

Mass Fraction

Co-based

Mass Fraction

Parameter

Table 6.1  Overview of Fischer–Tropsch (FT) catalyst and reactor XTL (X = coal and natural gas or combinations) technology currently in commercial operation, using low temperature FT (LTFT), medium temperature FT (MTFT), and high temperature FT (HTFT) modes of operation.

Mass Fraction

6.1  Introduction

Figure 6.4  Schematic representation of Fischer-Tropsch (FT) reactors and images of catalysts employed for stationary and moving bed reactors.

6.1.3  FT Catalysts for SAF Production Cobalt and iron based FTS catalysts are two contenders for the production of jet fuel products from renewable CO2/syngas and H2, each having their own strengths and weaknesses (Table 6.2). Because cobalt catalysts have negligible activity for CO2 hydrogenation (particularly in the presence of CO), it is necessary to convert the CO2 either via co-electrolysis or RWGS to CO and H2 in an intermediate step. This green syngas is then readily converted to products over the cobalt catalyst. In contrast, iron catalysts possess RWGS activity and can convert CO2 in situ to CO, Table 6.2  Comparison of Co-low temperature Fischer-Tropsch Fischer (LTFT) and Fe-HTFT systems for Fischer-Tropsch synthesis (FTS) to produce sustainable aviation fuels (SAF). Cobalt LTFT catalysts

Iron HTFT catalysts

High activity catalyst with established regeneration protocols enabling materials circularity. However, cobalt and noble metal promoters are relatively expensive, and catalysts are sensitive to poisons. Metallic cobalt cannot convert CO2 and H2 directly, need separate RWGS reactor or co-electrolysis unit to produce synthesis gas Good water tolerance enabling higher per pass conversion, minimizing the need for large recycle streams

Inexpensive catalyst, robust to a variety of feed poisons, but used in large quantities and has a short life-time (limited by carbon formation) Can do tandem catalysis, combining RWGS and FT in one catalyst system Greater sensitivity to water, per pass conversion is limited, requiring recycles. Especially problematic with CO2 hydrogenation which produces an addition mole of water for every mole of C converted

Primary product is alkanes. Can maximize wax selectivity leading to higher overall kerosene yields post hydrocracking.

Complex product spectrum (olefins, oxygenates, aromatics). Needs oligomerization for high kerosene yield, but can provide benefit for chemicals production.

79

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

f­ ollowed by FTS. While the RWGS activity is favorable, it produces additional water. This can be problematic for iron catalysts, which are oxidized to FT inactive oxide phases. On the other hand, cobalt catalysts are much more likely to be able to run at high conversions due to their stability against oxidation. Iron catalysts typically run at around 3 bar water partial pressure [30], whereas cobalt systems are reported to be capable of running at 8–10 bar water partial pressures [31]. The important parameter for stability appears to be the pwater/phydrogen ratio. The stability of nano-sized Co- and Fe-carbides as a function of CO conversion and crystallite size is shown in Figure 6.5. The data indicates that even small cobalt crystallites (8–10 nm) are unlikely to be ­oxidized at high CO conversions (70–80%); conversely Fe carbide crystallites, even if they are fairly large (>20nm), can oxidize at over 60% conversion levels [32]. The product spectrum obtained from typical cobalt catalysts is primarily straight chain, paraffinic wax products with high alpha values. These waxes can then be hydrocracked directly to the desired kerosene range hydrocarbons, ensuring high jet fuel yields. Iron systems, particularly at higher temperatures, have a more complex product spectrum, including more isomerized hydrocarbons, olefins, aromatics, and oxygenates. Although these components are attractive from a chemical value perspective, they require complex downstream product work-ups and jet-fuel yields (per mass catalyst) are significantly lower. Strategies to improve kerosene production, such as oligomerization of C3/4’s, add capital intensity to the process [33]. Iron-based catalysts have a much shorter lifetime than their cobalt counterparts, typically lasting weeks or months whereas cobalt catalysts can last years [34] and are relatively easily regenerated. As iron is a cheaper starting material, it is often discarded and not re-used, potentially causing environmental concerns. Iron-based catalysts do have a higher threshold for poison in the feed-gas and are able to tolerate oxygen and nitrogen containing poisons at higher levels, sometimes incorporating them into the product spectrum [35]. Poison tolerance or utilization of appropriate gasclean-up technologies is especially important when using waste or biomass as a renewable feedstock as these sources possess oxygen, nitrogen, halides, and sulfur-containing contaminants, which poison FTS catalysts [36]. As indicated, cobalt catalysts are preferred for application in FT-based SAF processes. This is mainly due to the propensity of cobalt-based systems to produce straight chain hydrocarbons that can be used to derive synthetic kerosene at high yields. Many of the announced FT plants that will be used to generate SAF are envisaged to use cobalt catalysts [37]. This review will therefore focus

60 40

CoO inactive Co0 active

20 0

0 5 10 diameter of metallic Co, dCo, nm

80 XCO+CO ,% 2

80

100

T = 493.15 K (H2/CO)inlet = 2

CO+CO2-conversion,

100 Co-conversion, XCO, %

80

T = 525 K; p = 21bar Fe3O4, ‘inactive’

60 40

(H2/CO/inert)inlet = 2/1/0; WGS eq. (H2/CO/inert)inlet = 2/1/3; WGS eq.

20 0

χ-Fe5C2 ‘active’

(H2/CO/inert)inlet = 2/1/0; no CO2 (H2/CO/inert)inlet = 2/1/3; no CO2

0 5 10 15 20 25 30 diameter of metallic χ-Fe5C2, dχ-Fe5C2, nm

Figure 6.5  Oxidation propensity of cobalt (left) and iron (right) as a function of crystallite size and CO conversion in the reactor as determined by surface thermodynamic calculations. Reproduced with permission from Ref [32] / Elsevier.

6.1  Introduction

more on cobalt-based catalysts for production of SAF. The key challenge for FTS in a PtL process is to ensure that syngas conversion is as high as possible to minimise recycles, while still achieving the required selectivity to produce high jet fuel yield. Iron-based catalysts can be important for directly producing sustainable chemicals from CO2 and H2 and there are opportunities to make these systems more sustainable and water tolerant. These will be touched on briefly in this chapter.

6.1.4  Reactor Technology for SAF Production Using FTS In contrast to the large-scale FT plants described in the previous section, the first commercial biomass-to-liquid (BtL)/PtL plants are expected to be much smaller, less complex, and decentralized. The small-scale FT concept is not new and was first discussed in a gas-to-liquids (GTL) ­context about a decade ago [38]. By taking advantage of new or adapted reactor technologies, GTL plants can be scaled down and operated on a distributed basis, providing a more cost effective a more cost effective to take advantage of smaller gas resources with smaller plants located near gas resources and potential markets. The lower capital cost also establishes a lower hurdle to market entry. Due to the high costs of green H2 production and localized sources of renewable carbon sources, the initial PtL landscape is also expected to start at a significantly smaller scale (500–2500 b/d) This will allow use of reactor technologies having a smaller footprint, but also needs to be highly efficient to process the relatively expensive renewable resource. Some examples of smaller scale reactors for FT for production of sustainable fuels are shown in Figure 6.6. BP and JM [39] have developed the novel CANs™ reactor design which combines the advantages of fixed bed reactors with a slurry phase system allowing the use of smaller catalyst particle sizes, with lower pressure drops. The new reactor system consists of modular catalyst containers, stacked within a fixed bed reactor, providing modified reactant flow paths down a porous central channel. This configuration delivers improved mass transfer and kinetics, aiding in reactor intensification. It is claimed that this configuration allows reduction of unit cost by half, while at the same time increasing product volumes by 50% and operating at >90% overall CO conversion with a C5+ selectivity around 90–92% [40]. The stainless-steel microstructure reactor concept used by INERATEC [41, 42] consists of a 2  cm3 reactor volume and consists of eight parallel catalyst sections sandwiched between

Figure 6.6  Examples for novel small-scale Fischer-Tropsch synthesis (FTS) reactors. Left: BP/JM/Davy CANS™. Reproduced with permission from Ref [39] / Johnson Matthey Plc; Middle: microstructure reactors as used by INERATEC. Reproduced with permission from Ref [42] / Elsevier. Right: Microchannel reactors as used by Compact GTL and Velocys. Reproduced with permission from Ref [51] / Elsevier.

81

82

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

cross-flow oil channels for heat exchange. Each catalyst section consists of two foils with an etched 400 μm deep pillar structure, hexagonally arranged with 800 μm distance between the pillars. The foils are stacked opposite to each other, giving 800 μm channel height. INERATEC reactors will be used in the first containerized SAF plant in Germany (8b/d) with a Sasol Co catalyst, affording high CO conversions and C5+ selectivity [43]. In the Velocys [44] microchannel set-up, each reactor block has thousands of process channels (~0.1–1.0 mm) filled with active cobalt FT catalyst interleaved with water filled coolant channels. The microscale dimensions of the channels increase the surface area per unit volume and thus increase the overall productivity of the process per unit volume. This allows significant process intensification, whereby the reactor volume to produce a given amount of product is reduced by an order of magnitude or more. The systems are capable of running for more than a year 70% CO conversion, with C5+ selectivity around 88–90% without requiring any regeneration [45]. Velocys plans to use this technology on two commercial SAF projects in the UK and United States, one using municipal solid waste as feed stock (c. 1550 b/d) and one ­utilizing woody biomass (2750 b/d) [46]. More conventional multi-tubular fixed bed reactors can also be used for SAF production and do possess a degree of modularity and scalability, as shown by Shell [47]. These are expected to be a preferred choice for small scale PtL/BtL plants. Sunfire is another company that intends to use tubular fixed bed reactors to produce aviation fuel [48] using renewable feed from water electrolysis, direct air capture (DAC), and waste CO2, with the intention of deploying the technology to produce 420b/d of product by 2026 for the Norsk e-Fuel project [49]. Small scale slurry reactors may also be viable for SAF plants in the 2000–2500 b/d range. Sasol has operated small scale slurry reactors since the 1980s, using commercial slurry bubble columns in Sasolburg with capacities of 100 and 2500b/d. Dynamic modelling simulations [50] indicate that slurry reactors containing cobalt catalysts could also operate well at variable feed loads, which would be a requirement for PtL applications, with good temperature distribution in the reactor and no detrimental influence on FT product selectivity. It is likely that new or adapted reactor technologies, could play a key role in the initial deployment of FT plants for small scale SAF production due to their efficiency, cost and lower footprints. However, smaller versions of conventional technologies such as small scale multi-tubular fixed bed and slurry, are also envisaged to play a role as well. As the SAF market ramps up and reaches maturity, much larger plant capacities will be required that could again favor the larger, more traditional reactor technologies.

6.2  State-of-the-art Cobalt Catalysts In the century since Fischer and Tropsch published their first article [52] and patent [53] on CO hydrogenation [54], more than 13,500 patent families were filed on the subject. Although FT-patent filing numbers showed a slight decline since the oil price crash of 2014, there is renewed interest in the process as it is perceived to be viable for the production of SAF. Developing a commercially viable FT catalyst must consider the targeted commercial process which includes reactor type, gas loop impacts, and the envisaged commercial production cost of the catalyst. The reactor type not only impacts the catalyst and support morphology (size and shape), but also which type of support and modifiers are required. For example, in a slurry bubble reactor the catalyst must survive a high attrition, abrasion, and hydrothermal environment for at least two to three years which can be stretched to 10+ years with regeneration.

6.2  State-of-the-art Cobalt Catalysts

The synthesis gas and recycle loops determine the following: Poison levels: Poisons can be split into permanent poisons (e.g. sulfur, alkali/alkali earth metals, and halogens) and non-permanent poisons (e.g. nitrogen poisons) [55]. Promoters can be added to capture poisons (e.g. Mn/Mo for S) or limit the enhanced hydrothermal impact caused by NH3 (from nitrogen poisons) and water. Cobalt phases (face-centered cubic/hexagonal closed packet; FCC/HCP) could also be impacted by poisons and thereby impact catalyst performance. ● CO2 content: CO2 slippage from synthesis gas production and WGS builds-up in the internal recycle, especially when maximizing overall conversion. Cobalt catalysts have inherently low WGS activity but addition of promoters (such as precious metal reduction promoters, Pt [56], etc.) can increase the WGS activity. Some cobalt FT-catalyst systems were reported to be CO2 sensitive [57]. ● Water partial pressure: Although water is the main product from FT-synthesis and hence dependent on per pass conversion, there is also water in the synthesis gas inlet, depending on the efficiency of water removal from upstream units such as RWGS and/or Autothermal Reformer units. Water impacts will be discussed later. ● H2/CO ratios and gradients: Low H2/CO ratios can enhance polymeric carbon formation and carbon deposition [58]. Reduction promoters usually minimize carbon formation. The literature also indicates that the partial pressure ratio of (H2+CO)/H2O can impact sintering. ●

Due to these factors, it is essential to test and develop commercial FT-catalysts using representative synthesis gas and targeted conditions, which can increase the cost and timeline of FT-process development. Shell invested around $1 billion to develop their GTL process [59] with catalyst cost expected to contribute a significant portion. FT catalyst developments therefore focus on stability and selectivity to drive down the catalyst contribution to production costs. This is evident from published presentations showing how Shell [60] (Figure 6.7) and Sasol [61] have improved their commercial offerings to target C5+ selectivities in excess of 95% and alpha values larger than 0.94. A list of commercial cobalt catalysts and commercial reactor selection for these catalysts can be found in Table 6.3. It should be noted that the actual composition of the commercial catalysts employed may be slightly different to the published information. In general compositions of commercially relevant cobalt catalysts are very similar, usually containing: Nanosized cobalt as the FTS active metal (typically 10–30 wt%). A second metal (usually noble) as a reduction promoter (0.05–1 wt%) to facilitate H2 spillover during reduction. ● A structural oxidic promoter (e.g. Zr, Si, and La) (1–10 wt%), which is used to protect the support and impart stability. ● ●

Figure 6.7  Improvement in Shell catalyst selectivity with time, based on analysis of patent data [62].

83

84

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

Table 6.3  List of commercial cobalt catalysts and reactors. Adapted from [63]. Technology provider

Support

BP/JM [64]

ZnO [65] & TiO2

Exxon Mobil [66]

TiO2

CompactGTL [67]

γ-Al2O3

Conoco-Philips [68]

γ-Al2O3

EFT [69]

γ-Al2O3

ENI/IFP/Axens [70]

Support modifier

Promoter

Reactor type

Mn

FB or Davy CANS™

γ-Al2O3

Re

Bubble bed

Re

Microchannel

B

Ru/Pt/Re

Bubble bed

γ-Al2O3

Si

Pt(?)

Bubble bed

GTLF1 [71]

Ni-Aluminate

α-Al2O3

Re

Bubble bed

INERATEC [72]

γ-Al2O3

Si

Pt

Microchannel

Nippon [73]

Silica

Zr

Ru

Bubble bed

Sasol−Topsoe (a)

γ-Al2O3

Si/C

Pt

Bubble bed

Sasol−Topsoe (b) [74]

SiO2

Ti

Mn/Pt

Bubble bed

Shell [75]

TiO2

Mn/V

Fixed bed

Fixed bed

Syntroleum [76]

γ-Al2O3

Si/La

Ru

Bubble bed

Velocys [77]

Silica

Ti

Re

Microchannel

● ●

A selectivity promoter (e.g. Re and Mn), specifically to target longer chain products. A refractory oxidic support such as titania, alumina, silica, or combinations thereof.

The various considerations that need to be taken into account when designing an optimal cobalt catalyst with required activity, stability, and selectivity are described in Figure 6.8 [51, 78–80]. Aspects of these will be discussed in the next section.

Figure 6.8  Considerations for the design of efficient supported cobalt catalysts with high water tolerance and high selectivity in PtL applications, consolidated from information in Refs [51, 78–80].

6.2  State-of-the-art Cobalt Catalysts

6.2.1  Catalyst Preparation Routes for Cobalt-based Catalysts 6.2.1.1  Precipitation Methodology – a Short Summary

Co-precipitation was initially the preferred FTS-catalyst preparation method (e.g. Co/ThO2/kieselguhr during the second World War) [81]. More recently, co-precipitation has been applied, such as for the BASF (formerly Engelhard) Co/ZnO catalyst [82]. Co-precipitation follows a five-step process that includes precipitation, washing to remove unwanted salts, and drying followed by shaping and calcination [83]. For bubble bed application, the drying and shaping steps are usually replaced by re-slurrying and spray drying. Washing out of the residual alkali metals and/or halogens is important as ppm levels thereof can degrade both selectivity and activity of the catalyst. The active metal (Co) crystallite size and composition of the precipitate can be controlled by the precipitation agent, organic hydrolysis reagents, precursor salts, temperature, and ageing time. As disposal of salts can pose an environmental risk, care must be taken in the choice of precipitation agents so that the resulting salts can be recycled, re-used or decomposed. Although batchwise precipitation is still employed, continuous processes are preferable for large scale production. The choice of precipitation chemicals can also impact side reactions such as the formation of cobalt aluminates and cobalt silicates (e.g. using Na2CO3 or KOH as precipitation agents for Co/SiO2 FTScatalysts) [84]. Cobalt silicate formation can be limited by adding the silica after precipitation. 6.2.1.2  Preparation Methods Using Pre-shaped Supports 6.2.1.2.1  Support Modification

Homogeneous covering of the support surface by the modification promoter is important to ensure optimal hydrothermal protection and/or promotional impacts [85–87]. This can be achieved by either organic based slurry impregnations [88], aqueous phase impregnations or grafting [89]. Some aqueous phase modifications may require extensive washing steps to remove residual sulphates, alkali metals or halogens as they can negatively impact catalyst performance. During organic base modifications it is important to choose a solvent with a sufficiently different boiling point to the modifier precursor to minimize losses. To achieve optimum modifier distribution, precursors with optimum interaction with the support surface are required and/or optimal drying and calcination procedures must be targeted. 6.2.1.2.2  Cobalt Impregnation

While bubble bed and micro channel reactor catalysts usually target homogeneous cobalt distributions, larger fixed bed catalyst particles often target an egg-shell cobalt distribution to limit the impact of diffusion limitations during FTS [90]. When targeting a homogeneous distribution of cobalt and promoter, the desired cobalt crystallite size is another critical parameter. To achieve this, the following aspects should be considered: Pore volume determines the maximum amount of precursor that can be added [91]. Slurry impregnations can usually target 95% of pore volume occupation by precursors, whereas incipient wetness (IW) preparations at times require lower concentration solutions to simplify the precursor addition process. ● To obtain a homogeneous impregnated cobalt distribution throughout the support particle, the particle geometry (diffusion path length, 2 or 3D pore structure, pore geometry, and size), interaction between precursor and support surface (point of zero charge [PZC] of the support, pH of the solution, contact angle, and surface tensions), viscosity of the precursor, suspension, and diffusion coefficients need consideration to determine optimum impregnation and drying [92, 93]. These impacts are demonstrated in Figure 6.9, showing the cobalt distribution for optimized slurry impregnation compared with IW impregnation followed directly by similar calcination procedures. ●

85

86

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

Figure 6.9  Scanning electron microscope/energy-dispersive x-ray spectroscopy (SEM/EDX) line scans show the heterogeneous cobalt oxide distribution for incipient wetness (IW) impregnation followed directly by fast drying (left) compare to the more homogeneous distribution from optimized slurry impregnation (three hours slow drying) followed by fast drying (right) [94, 95]. Reproduced with permission from Ref [51] / Elsevier.

The addition of viscosity enhancers, chelating agents [96, 97], organic metal precursors [98], organic acids [99], and binary acids [100] can further enhance the cobalt crystallite size distribution. When targeting egg-shell type impregnations for fixed bed applications, addition of viscosity enhancers or using cobalt salt melts are preferred during impregnation [101]. For deposition-­ precipitation onto pre-shaped supports, both the precipitation agent and cobalt precursor’s ­diffusion into the pre-shaped support and interaction with the support surface are important and must be controlled to obtain homogeneous distributions and crystallite size distributions [102]. Metal precursors can migrate out of the pores if care is not taken during the drying phase from the point of IW until the metal precursor decomposition (calcination) starts. The same parameters for impregnation and deposition should be considered to prevent the cobalt precursor (if not ­chemically fixed to the support) from migrating out of the particle. In addition, heat transfer ­coefficients, evaporation enthalpy and particle outer surface area need consideration for ­optimizing the drying stage of the cobalt catalyst preparation. Models are available that describe the impacts of convective flows, diffusion and adsorption of metals over the surface of the porous support. From these models, targeted conditions can be used to either obtain homogeneous, ­egg-shell or egg-yolk distributions [103]. Even the drying atmosphere (e.g. using air or N2) can impact the final cobalt distribution [104]. Each preparation method needs to be optimized ­carefully and one cannot assume that the optimum procedure for one type of support and support particle shape will be the same for all other supports and support particle shapes. The interaction between the support ­surface and cobalt precursor may differ from the promoter precursor and surface. If care is not taken during the drying phase as well as any storage time between drying and calcination, cobaltpromoter segregation may occur. 6.2.1.2.3  Calcination

Storage time between drying and calcination must be limited as precursor migration may occur. Most calcinations are performed in air, but organic precursors may require the use of nitrogen or diluted air to control the exothermic decomposition [21]. Alternatively, reactive gases such as NO can be added to assist in the decomposition of the cobalt precursor [105, 106]. To maintain the cobalt distribution achieved by impregnation and drying, cobalt precursor mobility must be prevented. This can be achieved by controlling the heating rate and removal of

6.2  State-of-the-art Cobalt Catalysts

decomposition products in such a manner that the cobalt precursor stays in a highly viscous or solid form [107]. The mobility of the cobalt precursors can be further limited by flash or high heating rate calcinations. Care should be taken regarding the heat flow into the system as both the drying and nitrate decomposition are endothermic [108, 109]. Reductive calcinations that use H2 or CO as decomposition medium can also be used if care is taken regarding explosive limits [110, 111]. Organic additives counteract the overall endothermic nature of cobalt nitrate decompositions as their decomposition is exothermic, thereby accelerating the calcination process. The transmission electron microscopy (TEM) images in Figure 6.10 illustrate how optimal cobalt distributions can be achieved by employing different calcination strategies.

Figure 6.10  Transmission electron microscopy (TEM) images of cobalt crystallite distributions in cobalt alumina catalysts employing different calcination strategies. (a) Cobalt oxide microglobule formation of a 30 g Co/100 g alumina catalysts using a heating rate of 1°C/min and an air space velocity of 1 m3 / kg Co(NO3)2.6H2O / hour. (b) Cobalt oxide distribution of a 30 g Co/100 g alumina catalysts using fast heating rate and air space velocity to ensure optimum calcination (flash calcination). (c) Cobalt oxide distribution of a 30 g Co/100 g alumina catalysts using carbon coated alumina (C as accelerator for calcination), using the same heating rate and air flow rate as in (a). (d) Cobalt oxide distribution of 30 g Co/100 g alumina catalysts using the same heating rate and flow rate as (a), but with 1% NO in He as in the calcination atmosphere. Reproduced with permission from Ref [51] / Elsevier.

87

88

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

6.2.1.2.4  Reduction

Activation of cobalt catalysts requires a reduction step in hydrogen or a diluted hydrogen atmosphere. CO reductions can be performed if carbon formation can be limited. Co3O4 is first reduced to CoO, followed by further reduction to cobalt metal. Both these steps are exothermic; hence care must be taken to optimize heat transfer during commercial reductions as well as minimizing hydrogen diffusion and mass transfer limitations. Efficient removal of reduction products such as water and ammonia (from residual nitrates) is reliant on high hydrogen space velocities and use of low heating rates [112, 113]. The impact of water partial pressure during reduction on the performance of a 30 g Co/100 g Alumina is clearly demonstrated in Figure 6.11. The lowest commercially possible water partial pressure is targeted to maximize performance and to achieve the optimal reduction rate in agreement with the thermodynamics of reduction. Therefore, the hydrogen stream used for reduction must be as dry as possible. In fluidized bed reduction of catalyst powders these criteria are easily met, while for fixed bed reductions care must be taken to overcome the limitations particularly toward the reduction reactor outlet. Reduction promoters (Pt, Ru, Pd, and others) lower the maximum reduction temperature required. A lower reduction temperature is sometimes required to limit side reactions such as silicate formation and sintering. Catalyst performance depends critically on the reduction procedure [114] and sometimes intermediate hold steps are required to obtain optimum promoter interaction with the cobalt [115]. Catalyst performance can further be enhanced by up to 30% with a reduction–oxidation– reduction (ROR) [116, 117] sequence. This improvement from ROR treatment is linked to rougher cobalt crystallites with more steps on the surface, increased degree of reduction, and re-dispersion of the cobalt on the support surface. To secure the more active Co-HCP phase, a reduction-carbiding-reduction (RCR) activation step can be employed [118]. The Co-HCP phase improves both the activity as well as the selectivity toward longer chain products [119].

6.2.2  Challenges for Catalysts Operating with High Carbon Efficiency: Water Tolerance To maximize the carbon efficiency for PtL processes conversions should be high, while at the same time not compromising the selectivity to the desired products. Due to the inherent stoichiometry of the FT reaction, a mole of water is produced for every mole of CO converted. Thus, higher conversions afford high reactor water partial pressures, particularly in the case of cobalt (which has negligible water-gas shift activity).

Figure 6.11  Impact of water partial pressure during reduction of a 30 g Co/0.075 g Pt/100 g alumina catalyst on initial Fischer–Tropsch (FT) synthesis performance (Sasol in-house data, redrawn from Ref [51] / Elsevier).

6.2  State-of-the-art Cobalt Catalysts

The positive and negative impacts of water on FTS catalysts have been widely studied and there are several reviews on the topic [120–122]. Water may affect the activity, selectivity, and deactivation of the catalyst, and its impact is related to the amount of water present (CO conversion level), catalyst structure, and catalyst composition as well as the reactor employed. At realistic syngas conversions (50–70%), several bars of steam will be generated, which could have a negative impact on catalyst performance. In general, this would be more problematic in a slurry bed reactor since there is substantial backmixing compared to fixed-bed reactors. This in turn implies that a comparatively large part of the circulating slurry bed catalyst is exposed to high water partial pressures. In a fixed-bed reactor the particles are stationary and there is a gradient across the bed resulting in water partial pressure increasing from inlet to outlet. Thus, a smaller part of the catalyst is exposed to higher water partial pressure. Therefore, slurry bed catalysts require a greater water tolerance than fixed bed catalysts when targeting the same overall outlet FTS conditions. The positive effects of water include enhancement in long chain hydrocarbon selectivity ­[120–122]. Water is known to enhance C5+ selectivity by increasing the chain propagation α-value. The exact mechanism of this improvement is not conclusively known but seems to be linked to the observation that water increases the coverage of monomeric CHx species that are responsible for chain growth. Water assisted-CO activation may also play a role [123]. The observed impact on selectivity is also related to the level of water in the reactor. Tucker and van Steen [124] showed that after a certain level of water in the system is reached (CO conversion >75%), the selectivity to longer chain products can in fact worsen. At these conditions excess amounts of CO2 can be formed due to water gas shift caused by oxidized cobalt species. Hydrogenation and thus CH4 selectivity is also enhanced due to an increased H2/CO ratio on the surface due to water gas shift. At very high water partial pressures, there is also a possibility of capillary condensation in the ­catalyst pores at low reaction temperatures. For catalysts with very narrow pores, this can become problematic for reactant diffusion to the active metal, impacting coverages of H2 and CO on the cobalt nanoparticles buried in the pores and negatively impacting selectivity. The deleterious effects of high water partial pressures on alumina- [124], silica- [125], and titania- [126] based catalysts have been well documented in literature. Van Berge et al. found that an alumina-supported cobalt FTS catalyst was inherently susceptible to hydrothermal attack under typical FTS conditions [127]. This increased the rate of deactivation and resulted in contamination of the wax product with ultra-fine cobalt-rich particulate matter. The problem was solved by pre-coating the support with a silica structural promoter. High water partial pressures (c. 10 bar) during FTS also contributed to significant amounts of cobalt aluminate formation as indicated by a XANES study conducted by Moodley et al. [128]. It was shown that due to the isostructural nature of Co2+ and Al3+, cobalt ions from the unreduced oxide can diffuse into the support in the presence of a hydrated surface resulting in the production of difficult to remove cobalt aluminate. Higher water partial pressures also promote the sintering rate of nano-sized cobalt particles [129]. Fundamental work studying the impact of process conditions on Co/Pt/Alumina catalysts in an in situ magnetometer indicated that the presence of high water partial pressure in combination with high CO partial pressures can cause enhanced sintering of the cobalt phase [130]. Thus, the literature indicates that although water can have a positive impact on long chain selectivity at certain levels, very high water partial pressures can be detrimental to both activity and selectivity. Robustness toward water is therefore an important aspect to consider when designing catalysts for PtL processes that run at high conversion levels. Various approaches will be discussed in the next section.

89

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

6.2.3  Strategies to Increase Water Tolerance and Selectivity for Cobalt Catalysts 6.2.3.1  Optimizing Physico-chemical Support Properties for Stability at High Water Partial Pressure

1.2

0.8

8

0.6

7

0.4

6

0.2

FTS activity Reactor water partial pressure

0

1

2

3 4 5 6 7 8 Time-on-line (days)

9 10

FTS activity (relative)

9 Pressure (bar)

1

0

1.2

10

1 0.8 0.6 0.4

5

0.2

4

0

FTS activity Reactor water partial pressure

0

5 10 15 20 25 30 35 40 45 50 55 60 Time-on-line (days)

15 14 13 12 11 10 9 8 7 6 5 4

Pressure (bar)

In order to optimize catalyst performance at high syngas conversion, one would need to consider the physico-chemical properties of the support such as surface area, pore volume, pore diameter, and thermal conductivity as well as surface acidity and basicity that can affect the stability at ­elevated water partial pressures [131]. The pore volume has to be matched to the loading of an appreciable quantity of active metal precursor (at least 15–20 wt % Co in order to obtain a catalyst that is active enough for commercial purposes) to limit pore blocking and ensure a homogeneous macroscopic cobalt distribution [132]. Support hydrophobicity can stabilise catalyst performance at higher CO conversion by decreasing the interactions of water with the support and the active metal surface. Post-preparation silylation with 1,1,1,3,3,3-hexamethyldisilazane (HDMS) improved catalyst stability at higher CO conversion conditions relative to an unmodified catalyst. However, hydrocarbon selectivity did not significantly improve [133]. Of the various porous supports (such as alumina, silica, titania, zirconia, SiC, zeolites, and carbon fibers or nanotubes) used for cobalt catalysts, carbon and SiC exhibit superior thermal conductivity that has been linked to improvements in heat removal capacity from the catalyst surface [85, 134, 135]. This is specifically relevant to the FT reaction which is quite exothermic (ΔH = –166 kJ/molCO) and could lead to local hot spots on the catalyst surface, particularly at higher conversion. Shaped carbon (nanotubes and fibers) has been investigated extensively due to the weaker metal-surface interaction and smaller effect on reducibility (making crystallite size effects more facile to evaluate). To date, commercial applications are limited mainly due to sintering and degradation of the support during regeneration [136]. There have been many improvements in the porosity and shaping of SiC supports in recent years which is essential to achieve the required metal loading and mechanical and abrasion resistance that are required for a commercial cobalt catalyst [137]. Compared to an alumina support, the use of SiC in a slurry reactor improved stability at high water partial pressures (c. 10 bar) with only 10% deactivation observed after 50 days on stream (Figure 6.12) [31]. An aggressive acid wash further improved anchoring and dispersion of cobalt oxide, likely due to removal of alkali impurities, surface roughening, and the introduction of additional acid sites. Uniform Ti modification of the SiC surface further improved FT activity. This has been linked with literature citing an improvement in metal dispersion and interaction with the support surface after Ti addition to SiC [138]. FTS activity relative to day 1

90

Figure 6.12  FT evaluation of Co/Al2O3 catalyst (left) and Co/SiC catalyst (right) in a slurry reactor at 230 °C, 30 bar, synthesis gas as feed of composition: 60 vol% H2, 30 vol% CO and 10 vol% inerts. In all other experiments the gas space velocity was adjusted to keep the water partial pressure constant at 9–10 bar (syngas conversion of 50–80%). Redrawn with data from Re [31] / Elsevier.

6.2  State-of-the-art Cobalt Catalysts

6.2.3.2  Stabilizing the Support by Surface Coating

Surface hydroxyls act as anchoring points for cobalt nitrate during impregnation (irrespective of the type of support). By altering the impregnation pH and the subsequent electrostatic interaction with the support surface, the dispersion of cobalt nitrate may be optimized significantly (Figure 6.13) [85, 132]. Surface hydroxyls are vulnerable to secondary reactions with organic acids and water that are formed during the FT process. There is an interplay between ensuring an optimal number of anchoring points for the active metal phase whilst making the surface inert enough toward chemical attack. Surface modifiers such as Al, Ba, Ce, La, Mn, Si, Ti, or Zr have been added to Al2O3, SiO2, or TiO2 (at a loading of 1–10%) to coat the surface hydroxyls thereby improving hydrothermal stability especially at conditions favoring higher syngas conversion and water partial pressures [132]. By carbon-coating alumina supports, the cobalt oxide dispersion over the support surface can be improved (Figure 6.14) [139]. High angle annular dark-field transmission electron microscopy (HAADF-TEM) showed smaller, more evenly spaced crystallites and a narrower cobalt size distribution that could limit the thermodynamic drive toward sintering due to surface migration and Ostwald ripening. There was a marked improvement in activity and sintering stability for the carbon coated catalyst over the conventional system. This is evidenced by the smaller change in % ferromagnetic fraction (FM) for the former which can be used to gauge the sintering extent, with sintering being the predominant deactivation mechanism in the early stages of the reaction. This corresponds with findings of Khodakov et al. showing that optimizing dispersion by the addition of carbon had the additional benefit of stabilizing catalyst performance at higher CO conversion relative to the unmodified catalyst [140]. 6.2.3.3  Impact of Crystallite Size on Selectivity

The support may be viewed as a mechanical scaffold that can be used to direct cobalt crystallite size through the selection of an appropriate pore diameter and other physico-chemical properties. Depending on the FT application, these physical properties of the support may be tailored to produce a catalyst with the desired cobalt crystallite size.

Figure 6.13  Interaction between Co nitrate and the silica gel surface during impregnation based on interfacial coordination chemistry. The progression from solution through formation of hydrosilicates and silicates to formation of CoO and Co3O4 is depicted. Reproduced with permission from Ref [132] / Elsevier.

91

92

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

Figure 6.14  Impact of carbon coating on performance of Co/Alumina catalysts. Carbon coating results in improved dispersion and nanoparticle spacing as well as narrowing the cobalt size distribution (left). These effects impact both the activity and stability toward sintering (right). Reproduced with permission from Ref [139] / Springer Nature.

The cobalt crystallite size distribution influences catalyst activity, stability, and selectivity. The selectivity impact levels out for metal crystallites greater than 8 nm [51, 141]. Although activity relates to metal surface area, larger crystallites are more stable toward oxidation and sintering. Therefore, a narrow crystallite size distribution is preferred to optimally utilize the cobalt. Literature consensus is that cobalt-based FTS is a structure sensitive reaction (Figure 6.15). Metallic cobalt particles that are too small do not contain the necessary B5 step-edge sites required for CO dissociation [142] and subsequent C-C bond formation and thus don’t have optimal activity and selectivity. Salmeron et al. have proposed that the low activity and higher methane selectivity observed for small crystallites is linked to slower hydrogen dissociation on the surface rather than effects such as sintering, deactivation by carbon deposition or oxidation. This is in agreement with the conclusions of van Helden et al. that more highly defective Co sites (associated with smaller Co crystallites) exhibited a larger hydrogen sticking coefficient than the more planar sites found on larger crystallites [143]. Moreover, x-ray absorption spectroscopy (XAS) experiments showed that CO is adsorbed molecularly on very small Co nanoparticles but that the fraction of dissociated CO increases with crystallite size. This was proposed to be an effect of the hydrogen-assisted mechanism of CO dissociation for FT [144]. Smaller crystallites will therefore favor hydrogenation of CHx intermediates to methane but larger cobalt crystallites (≥10 nm) while being less active will be more selective towards longer chain hydrocarbons due to surface coverage effects and the more facile dissociation of CO and H2. 6.2.3.4  Metal Support Interactions with Cobalt Crystallites of Varying Size

Strong metal support interaction (SMSI) can negatively influence reducibility of small cobalt crystallites supported on alumina, silica, or titania due to the formation of cobalt silicates, aluminates

6.2  State-of-the-art Cobalt Catalysts

Activity

(111) facet

Coville (2011) Fischer (2010) Pietro (2009) Borg (2008) Martinez (2007) De Jong 1 bar (2006) De Jong 35 bar (2006) Bian (2003) Barbier (2001) Iglesia (1997)

0

5

10 15 20 Co particle size (nm) (a)

25

30

(100) facet

kink site

B-type step

A-type step

(b)

Figure 6.15  (a) Fischer-Tropsch synthesis (FTS) activity as a function of the cobalt metal particle size, showing structure sensitivity in cobalt FTS. Reproduced with permission from Ref [143] / Elsevier. (b) Atomic model of a ∼4.6 nm Co particle (4603 atoms), terminated predominantly by (1 1 1) and (1 0 0) facets. A significant portion of defects are present, in particular, ad-islands terminated by mono-atomic step edges. Reproduced with permission from Ref [142] / Elsevier.

and titanates [145]. This can shift the required reduction temperatures above 450 °C and reduce the active metal surface area available for FT reaction through sintering (Figure 6.16). Titania has been shown to migrate over the surface of cobalt crystallites after reduction at 450 °C forming partially reduced TiOx which covers active sites and leads to a decrease in hydrocarbon selectivity due to site blocking [146]. This strong-metal support interaction (SMSI) effect becomes more pronounced if high surface area supports are used [147]. It has been postulated that water Figure 6.16  Temperature programmed reduction (TPR) patterns for a spherical Co/SiO2 model catalyst following calcination with (a) large Co crystallites (Co-28 nm); (b) medium Co crystallites (Co-13 nm); and (c) small Co crystallites (Co-4 nm). TPR conditions: 5% H2/N2, 25–800 °C, 10 °C/min. Reproduced with permission from Ref [145] / Elsevier.

93

94

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

can reverse this encapsulation effect and this is the reason that titania supports respond positively to increasing water partial pressure during FTS [148]. 6.2.3.5  The Role of Reduction Promoters and Support Promoters in Optimizing Selectivity

The addition of a reduction promoter such as Pt, Ru, or Re will facilitate reduction to cobalt metal by increasing hydrogen availability on the surface through a spill-over mechanism [149]. Noble metal promotion will also influence selectivity and hydrocarbon productivity for catalysts with similar crystallite size (12 nm using x-ray diffraction [XRD] analysis) (Table 6.4). Mn is an important catalyst promoter patented by companies such as Shell and BP as well as receiving considerable attention in the academic literature. A component such as Mn not only limits migration of Ti during reduction (structural promoter), but also optimizes electronic interaction with cobalt crystallites at the support/crystallite interface [150, 151]. FTS inactive mixed Mn-Ti spinels can be formed on the support as well as mixed Mn1-xCoxO compounds. The latter has been found to retard reduction of cobalt while improving catalyst stability as well as selectivity (more so for large Co crystallites than small ones where Mn coverage of active sites might start playing a role). Similarly, it was shown that Mn-promoted alumina (20wt% Co/xMn/0.05%Pt/Al2O3) can shift the Schultz-Flory carbon distribution of a cobalt catalyst toward lighter hydrocarbons (Figure 6.17) [152]. Careful optimization of the Mn/Co ratio was reported to maximise C5+ selectivity for different TiO2 and SiO2 supports [153–155]. The impact of Mn on the FT rate and selectivity for Sasolprepared Co/Alumina catalysts was also evident and an optimal loading is required for balancing rate and selectivity. The data indicates that Mn can poison catalyst selectivity and activity at higher loadings, with Mn:Co mole ratio of around 0.179 being optimal. Researchers at Shell prepared a series of catalyst with various Mn loadings on Co/TiO2 and subsequently tested them in a fixed bed reactor [156]. The experimental data (Figure 6.18) showed that the presence of MnO significantly increased the catalyst activity by c. 80% and C5+ selectivity from 88 to 91 wt %, mostly by suppressing the methane selectivity at 50% conversion and 20 bar. The olefin/paraffin ratio was enhanced, indicating reduced hydrogenation compared to the unpromoted Co catalyst. The maximum activity was obtained at a Mn loading of 0.48 wt% (Mn:Co mole ratio = 0.034). The authors utilized DFT calculations on a Co (0001) model surface to help understand the electronic impact of the decoration of manganese oxide particles on cobalt. It was found that Mn weakens the C–O bond and reduces the barrier for direct CO dissociation while at the same time destabilizing hydrogen on the surface, thereby enhancing selectivity for olefin and long chain production. Of direct relevance to high conversion operation, Tucker et al. [157], added Mn to Co/Pt/Alumina catalysts in an attempt to improve the water tolerance and selectivity (see Figure 6.19). The addition of Mn: Co in 0.15 mol ratio significantly improved the selectivity of the catalysts compared to unpromoted

Table 6.4  Effect of Pt promotion on activity and selectivity for 20wt%Co/Mn/Ti-SiO2 catalysts containing Mn as a structural promoter. Temperature programmed reduction (TPR) in hydrogen with ramp rate of 1 °C/min to 1,000 °C, Fischer-Tropsch synthesis (FTS) conditions: Slurry reactor, 230 °C, H2/CO = 1.5, 15 bar total pressure evaluated at isoconversion (Sasol in-house data). Co3O4 Pt loading reduction to (wt%) CoO (°C)

CoO reduction to Co (°C)

Normalised Normalised Normalised FT activity at methane hydrocarbon day 5 selectivity at day 5 productivity at day 5

0.05

161

  233

1

1

1

0

211

  289

0.85

1.22

0.83

6.2  State-of-the-art Cobalt Catalysts

Figure 6.17  Commercial type cobalt on Mn-modified alumina (20 wt% Co/xMn-Al2O3) showing that carbon distribution shifts toward lower carbon-number with increasing Mn loading [152].

Figure 6.18  Trends of activity and selectivity (210 °C, 50% CO conversion, 20 bar H2/CO), for 16 wt% Co/ TiO2 catalyst as a function of Mn:Co mole ratios as reported by Shell. Redrawn from data in Ref [156] / American Chemical Society.

catalysts, especially at high conversion conditions (>70% CO conversion). The manganese-promoted catalyst decreased the selectivity toward methane and CO2 at high CO conversion (XCO = 90%) with significant enhancement of fuel yield (C5+) up to 14 C-%. It was reported that Mn addition changed the reducibility and Co crystallite size of Mn-Pt-Co/Al2O3 significantly. Manganese seems to be

95

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

Figure 6.19  The yield of C5+ as a function of CO conversion for standard industrial Pt–Co/Al2O3 and industrial Mn–Pt–Co/Al2O3 Mn: Co = 0.15 mol/mol. Reproduced with permission from Ref [157] / Royal Society of Chemistry.

incorporated in the spinel structure of the calcined catalyst. This study provided convincing evidence that Mn as a promoter can boost selectivity at high conversion conditions. 6.2.3.6  Role of Pore Diameter in Selectivity

Pore diameter has the potential to determine cobalt crystallite size and therefore influence the C5+ selectivity. Diffusion of long-chain hydrocarbon products occurs more easily from larger pores due to the removal of transport restrictions via the formation of water-rich intra-particle liquids [158]. If all the cobalt nitrate is impregnated into the pores during catalyst preparation and migration of nitrate out of the pores during calcination is limited, the cobalt oxide crystallites that are formed inside the support pores should not exceed the pore diameter [159]. In alumina, a linear relationship between pore diameter and C5+ selectivity was found for largepore supports (large crystallites) vs narrow-pore supports (small crystallites) (Figure 6.20) [160]. The amount of crystallite contraction after reduction increased for large pore aluminas, as evidenced by the ratio of crystallite diameter analysed with XRD (Co3O4) to TEM diameter (Co). 1.4 1.3

1.3

1.2 1.2

1.1

1.1

1 0.9

1 0.9

Normalised FT rate

0.8

Normalised methane selectivity

0.7

0.8

Normalised CH4 selectivity

1.4

Normalised FT rate

96

0.6 10

15 20 25 Cobalt oxide crystallite size (XRD; nm)

30

Figure 6.20  Effect of varying cobalt oxide crystallite size (XRD) on selectivity and Fischer-Tropsch (FT) rate at a constant optimal pore diameter.

6.2  State-of-the-art Cobalt Catalysts

C5+ selectivity (%)

86 85

Narrow pore alumina

84

Large pore alumina

83 82 81 80 79 78 1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2

d(Co3O3)/d(Co0)

Figure 6.21  Effect of cobalt particle break-up on the C5+ selectivity at 483 K, 20 bar, H2/CO = 2.1, and 50% CO conversion for 20% Co/0.5%Re/Al2O3 tested in a fixed bed reactor. Adapted from Ref [160].

This was linked to an increase in C5+ selectivity, most probably due to more steps, edges and kinks present on the larger crystallites. A similar link between crystallite size and pore diameter for a series of Co catalysts supported on Ti/Mn-modified silica at a constant pore diameter (Figure 6.21) was observed. The addition of a dispersant could decouple crystallite size from the pore diameter to give small crystallites inside a larger pore due to the formation of smaller cobalt crystallites after reduction, in agreement with results from Borg et al. [160], which showed improved selectivity. Process conditions (such as temperature, H2/CO ratio, and CO conversion) will influence selectivity as is clear from the plots in Figure 6.22 which depicts C5+ selectivity for various Co/SiO2 catalyst systems as a function of increasing pore diameter. Even though conditions for these runs were dissimilar, the general trend of increasing hydrocarbon chain length with 95

C5+ selectivity (%)

90 85

80 Storsaeter et al (40–45% CO conversion) de Jongh et al (30–35% CO conversion) Saib et al (30–44% CO conversion)

75

70

Sasol (50% conversion)

65 0

2

4

6

8

10

12

14

16

18

Average pore diameter (nm)

Figure 6.22  Link between pore diameter and C5+ selectivity for different silica-based catalysts evaluated in fixed-bed reactors as reconstructed. Adapted from Refs [104, 161, 162].

97

98

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

larger pores was the same. Notable from the work of de Jong et al. was the focus on improvements in dispersion during catalyst preparation which also benefitted selectivity toward longer chain hydrocarbons. Appropriate surface modification with the same pore diameter, allowed for an improvement in selectivity irrespective of higher temperature and CO conversion conditions. 6.2.3.7  Effect of Activation Conditions on Selectivity

The activation of the catalyst prior to FT synthesis can radically affect activity and selectivity. Metallic cobalt can assume either the face-centered cubic (FCC) crystallographic phase or hexagonal closed packed (HCP) phase. The HCP phase is more stable at a low temperature, whereas the FCC phase becomes more stable when the crystalline size of the cobalt is less than 20 nm and at temperatures above 450 °C [163]. Wulf constructions and DFT calculations suggest that the extra steps and kinks on HCP-Co crystallites would lead to an improvement in FT activity for HCP-enriched catalysts [164]. An ab initio study of the thermodynamics of the relative interaction between CO and the HCP vs FCC cobalt crystallites indicated that as the CO pressure (and chemical potential) on the surface of the crystallite increases, a morphology change is induced which will change the exposed facets at high ­coverage [165]. Work done at Sasol indicated that for the same corresponding size of cobalt nanoparticle there are some noticeable differences in morphology for FCC and HCP particles (Figure 6.23). Most noticeable is that HCP structures expose a larger variety of site arrangements than FCC. There are four additional unique sites on the HCP structure, with the notable absence of square sites that are thought to be FT inactive. If these additional sites are FT active, it could be expected that the FT rate and selectivities could differ between FCC and HCP nanocrystals, in agreement with the findings of Liu et al. [164]. Experimentally, the most reliable method of producing HCP-rich cobalt catalysts is by a reduction-carburisation-reduction activation protocol, forming an intermediate cobalt carbide phase

Figure 6.23  A 4 nm face-centered cubic (FCC) (left) and hexagonal closed packet (HCP) (right) Co particle showing the defined sites and their locations on one particle. Left figure reproduced from Ref [143] with permission from the American Chemical Society. Copyright 2012. Right figure courtesy Pieter van Helden (Sasol).

6.2  State-of-the-art Cobalt Catalysts

95

3.5 hcp-rich

3

C5+ selectivity (%)

Relative intrinsic activity

which preferentially decomposes to HCP-Co [21]. This has been observed experimentally over various types of supports such as alumina, silica (with Ru-promotion being mentioned to form more HCP-phase after activation), and even ZnO [166]. Alumina supports retained some FCC-Co after the RCR activation cycle whereas the defective surface of silica paired with Ru-promotion was postulated to be responsible for preferential formation of pure HCP-Co after completion of the RCR cycle [167]. The addition of Re as promoter to Co/TiO2 catalysts (0.5 and 1 wt%) was also found to increase the fraction of HCP-Co after a standard reduction in hydrogen. However, the influence on turnover frequency (TOF) relative to the FCC-Co catalyst was insignificant [168]. Experimentally, the performance impact can be seen in Sasol fixed-bed data on Co/Alumina catalysts activated in H2 only (to produce an FCC-rich particle) and via RCR activation (to produce a HCP-rich particle) (Figure 6.24). The performance improvement noted for the HCP phase was also observed experimentally on other catalyst systems regardless of support, including unsupported cobalt [169], alumina [170], silica [171] and titania [172] supported cobalt catalysts. To summarize, support properties can influence the stability of the final catalyst through direction of cobalt oxide crystallite size and how those crystallites interact with the support surface, the reactants (H2 and CO), and the products that are formed (water and hydrocarbons). Through careful selection of the support material and final catalyst properties, stability at high water partial pressures can be achieved and the selectivity toward longer chain hydrocarbons may be optimized.

fcc-rich

2.5 2 1.5 1 0.5 0

94

hcp-rich

93

fcc-rich

92 91 90 89 88 87 86

0

1

2

3

4

5

6

85

0

1

Time-on-line (days) 8 7 6 5 hcp-rich

3

fcc-rich

2 1 0 0

1

2

3

4

Time-on-line (days)

5

6

Relative abunduncae of cobalt phases (XRD)

CH4 selectivity (%)

9

4

2

3

4

5

6

Time-on-line (days) 100 90 80 70 60 50 40 30 20 10 0

Co(fcc) Co(hcp) CoO fcc-rich

hcp-rich

Figure 6.24  Activity and selectivity improvement of hexagonal closed packet (HCP)-rich (●) and facecentered cubic (FCC)-rich (○) Sasol lab -prepared 20 wt%Co/Alumina catalysts in a fixed bed reactor (16 bar, 230 °C, H2/CO = 2, c. 55% CO conversion). The HCP-rich catalyst was produced in a similar manner to Adapted from [21]. The composition of the cobalt phases for the fresh catalysts as determined by XRD is shown on the bottom right.

99

100

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

6.2.4  Regeneration of Cobalt PtL Catalysts- Moving Toward Materials Circularity The typical life cycle of a cobalt slurry phase catalyst used in a FT plant can be described as follows: Raw materials such as cobalt salts, noble metal salts and support are used to manufacture the catalyst at a manufacturing facility. The reduced and wax-embedded catalyst is then transported to the GTL facility where it is used in the FTS section, to convert synthesis gas primarily to waxy hydrocarbons. ● After the catalyst has reached the end of its useful life, it is then transported to a metals reclamation [173] facility to remove the hydrocarbon wax and recover cobalt and promoter metals for return to the manufacturing facility where they are reworked into the process. Drawbacks of this approach include that the catalyst support is currently not re-used and that a portion of the expensive metals cannot be reclaimed. There is also an additional carbon footprint associated with the logistics and re-preparation routes. ●

Circularity of materials is a topic that will become increasingly important for sustainability in catalysis. Regeneration and re-use of the spent catalyst is therefore a critical focus area for FTS. On-site regeneration provides logistical benefits as it reduces the amount of spent catalyst sent for metals reclamation, simplifying the catalyst supply chain. Regeneration further allows re-use of the support and a significant reduction in the solid waste footprint. For regeneration to be successful, an understanding of the deactivation mechanisms at play is important [174] so that they can be reversed, and catalyst properties can be returned as close as possible to that of the fresh catalyst. The regeneration concept has been around since the early days of FT research and is well documented in a number of reviews [175–177]. Shell has used regeneration as a tool to manage the lifetime of their fixed bed cobalt HPS (heavy paraffin synthesis) catalyst. They previously reported that in situ regeneration (required every 9 to 12 months) is routinely applied with great efficiency [178]. Their initial regeneration patents involve the use of solvent washing or hydrogen treatment, oxidation, and reduction [179, 180]. Regeneration in a fixed bed reactor can be performed in situ without removal of spent catalyst. However, the reactor needs to be taken offline to perform the regeneration, which results in some down time and may require additional capital cost to ensure sufficient reactor capacity to achieve production targets. Later regeneration patents of Shell involve a process in which the spent FT catalyst particles are first oxidized, then treated with a solvent containing ammonium carbonate and ammonium hydroxide in water [181]. This seems to be significantly better than the standard procedure and may result in dissolving and redispersion of cobalt in sintered spent catalyst. BP also reported that regeneration was crucial for their catalyst to achieve a lifetime of four years. [182]. The main purpose for regeneration was to remove the extremely low levels of carbon deposited on the cobalt crystallites, which caused catalyst deactivation. Regeneration of a slurry phase catalyst requires that the catalyst be removed from the reactor and regenerated externally. An operational advantage is that with online catalyst removal and addition, the slurry bed reactor does not need to be taken off-line for catalyst replacement and allows continuous running of the reactor. Regeneration of spent slurry phase catalysts has also been reported by both Exxon-Mobil [183] and Syntroleum [184]. Exxon’s initial regeneration involved a low temperature oxidation step ­followed by reduction. According to a review by Bartholomew [175], Syntroleum uses an oxidation−reduction regeneration process. Syntroleum’s patents and publications indicate that the first step involves concentrating the wax/catalyst slurry. The concentrated slurry is then dewaxed to convert it into a free-flowing powder. Oxidation is then initiated to combust the remaining hydrocarbons on the surface and pores of the catalyst. The final step is the hydrogen reduction step at high temperatures after which the catalyst is re-slurried before being returned to the FT reactor.

6.3  An Overview of Fe Catalysts: Direct Route for CO2 Conversion

Figure 6.25  Sasol’s oxidative multiple regeneration process, that restores Fischer-Tropsch (FT) activity by redispersing sintered cobalt and removing deleterious carbon. Reproduced with permission from Ref [185] Springer Nature.

On the basis of the fundamental understanding of the deactivation mechanisms of their proprietary Co FT catalyst, Sasol opted for an oxidative regeneration process (Figure 6.25) which has the potential of reversing carbon deposition and sintering, increasing the effective catalyst lifetime and thus potentially reducing catalyst cost. Fundamental focus was put into optimizing the parameters during both oxidation and reduction steps, such that carbon removal and cobalt re-dispersion are maximized [185]. This has resulted in successful multiple regeneration cycles of the catalyst during pilot scale demonstration without significant loss in catalyst integrity. These technologies will assist in making the FT segment of PtL processes more sustainable.

6.3  An Overview of Fe Catalysts: Direct Route for CO2 Conversion 6.3.1  Introduction A direct path for CO2 conversion to hydrocarbons typically requires an iron-based FT catalyst to enable tandem catalysis combining the endothermic RWGS reaction with the highly exothermic FT reaction (Figure 6.26). Iron-based FT catalysts are preferred due to their intrinsic activity for the RWGS reaction which results in near equilibrium conversion of CO2 and H2 over a fairly wide range of reaction temperatures. The FT reaction, which is considered irreversible for all practical purposes, lies at the core of the overall process. It is instrumental in setting the required in situ syngas composition as well as the resulting final syncrude composition and subsequent work-up requirements. It is generally accepted that the hydrogenation of CO2 is less efficient for production of higher hydrocarbons compared to that of CO, producing more methane [186–188]. There is evidence that CO2 hydrogenation to hydrocarbons proceeds via two distinct steps in series using Group VIII metals as catalysts [189]. First, carbon dioxide is partially reduced to carbon monoxide via the RWGS reaction [190]. In the second step, the CO that is produced reacts in the FT reaction, producing mainly hydrocarbons. Notably, H2O is a common by-product of both reactions. The H2O that is produced in the FT reaction inevitably contributes to the equilibrium position of the water

101

102

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

Figure 6.26  A schematic of “tandem” catalysis combining the endothermic reverse water gas shift (RWGS) reaction with the highly exothermic Fischer-Tropsch (FT) reaction over an iron catalyst.

gas shift reaction and ultimately to the maximum CO partial pressure that can be achieved in a reactor. However, due to this facile reversibility, maximum overall conversion of CO2 and FT selectivities are typically governed by this equilibrium. Interestingly, using laser generated Fe carbide catalysts, Fiato et al. [191] proposed that there is also some contribution possible from the direct hydrogenation of CO2. Equation 6.1 is proposed to proceed via dissociative adsorption of CO2 followed by hydrogenation of the adsorbed carbon species. However, the extent to which this reaction contributes to the overall formation of organic products remains uncertain at present.

CO2 +  3H2 →  “CH2 ” + 2H2O

(6.1)

The role of CO2 in the initiation of chain growth and alcohol formation during the FTS was investigated by Xu et al. [192]. The authors found that the conversion of CO2 with an iron catalyst is much slower than the conversion of CO. It has been shown that when CO2 is present at low concentrations (0.2 mol% of CO), it can act to initiate chain growth but does not contribute significantly to chain propagation. However, when CO2 is present in a large amount relative to CO (CO2/CO = 3); ratio of H2/COx = 1.5, the WGS reaction was rapid relative to that of the FTS. According to Krishnamoorthy et al. [193] the addition of 13CO2 to H2/12CO reactants over iron catalysts showed that the hydrocarbon products had a negligible 13C content, indicating that CO2 is much less reactive than CO toward chain initiation and growth. Their studies also showed that except at RWGS reaction equilibrium, where CO and CO2 become kinetically indistinguishable from each other, CO2 does not appear to compete with CO toward chain initiation and growth reactions.

6.3.2  Effect of Temperature

Figure 6.27  Equilibrium stoichiometry as a function of temperature during reverse water gas shift (RWGS) (Ref [194] / U.S. Department of Energy / Public Domain).

For the direct conversion route, the choice of reaction temperature is probably the most important factor to consider apart from actual catalyst composition. The temperature range of 200–400  °C will result in sufficient RWGS activity [194] (Figure 6.27) while also representing a workable window for iron-based FT catalysts.

6.3  An Overview of Fe Catalysts: Direct Route for CO2 Conversion

When selecting operating temperature, the following should be considered. Higher temperatures will be advantageous for overall CO2 conversion and enable increased FT-reaction rates, but due to increased rate of desorption and hydrogenation, mainly short chain hydrocarbons will form. At high temperatures, iron also has one other notable drawback due to formation of substantial amounts of olefins and oxygenates which is highly un-desirable for final jet fuel properties [195]. It is worthwhile to note previous attempts to maximize kerosene yield utilizing iron catalysts in CO hydrogenation. When Kumabe et al. [196] investigated various iron-based FT catalysts, they found that an unpromoted iron catalyst operating at 280 °C gave a maximum kerosene yield at approximately 9%. It is clear that with conventional iron FT catalysts, the primary yield to kerosene is unsatisfactorily low over a wide range of temperatures, and that substantial upgrading of the syncrude is required to deliver an attractive commercial process. In general, the operating temperature should rather be high because of the equilibrium constraints for the reverse RWGS reaction. This puts considerable constraints on the achievable FTS conversion. Lower overall reaction rates will lead to larger reactor volumes and/or higher required gas recycle rates while the lower exothermicity of the overall reaction of CO2 as compared to CO makes temperature control in a catalytic reactor much easier [197]. A further constraint is that optimum ­conditions for each of the reactions are different (high temperature for RWGS, low temperature for FTS hydrogenation of CO). Riedel et al. [198] noted that at temperatures below 360 °C, organic products almost identical to those found with the traditional FT reaction were found (paraffins and ­olefins). At 400 °C, formation of carbon deposited on the catalyst became a major reaction; almost 60 mol% of volatile hydrocarbons were methane compared to about 30–40 mol % for the lower temperatures at similar CO2 conversion of 46–53 %. Accordingly, this result together with high tendency for carbon formation indicates that at 400 °C the FT regime of long chain hydrocarbon formation does not ­prevail anymore.

6.3.3  Effect of Pressure It is not expected that overall system pressure should affect the WGS equilibrium position too much (Table 6.5). However, the positive impact on the subsequent FT rate and selectivity to heavier hydrocarbon products cannot be ignored [199]. Fiato et al. [191] demonstrated the effect of reaction pressure for CO2 hydrogenation on a Fe-carbided catalyst containing 2% K as basic promoter at 270 °C in excess hydrogen (H2:CO~7) (Figure 6.5). The results demonstrated the applicability of the process over a range of pressures but also show very good selectivity at higher pressures. Table 6.5  Effect of pressure on the CO2 hydrogenation using an Fe carbided catalyst. Pressure (bar)

% CO2 converted

5.2

6.9

36.6

38.1

Selectivity (based on C1+ products–%wt.) % CH4

16.5

10.8

%C2+

83.5

89.2

% olefins in C2-C4

80.0

95.5

103

104

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

6.3.4  Effect of H2:CO Ratio Another factor to consider is the H2:CO2 feed ratio, with the stoichiometric ratio of 3:1 being the ideal target. Higher ratios will impact the equilibrium conversion of CO2 to CO in a positive way but results in a shorter hydrocarbon chain product spectrum. A good demonstration of the effect of H2:CO2 ratio can be found in the work of Fiato et al. [191] who systematically varied the molar feed ratio over two types of iron catalysts at 270 °C. The two catalytic systems that were investigated are a conventional precipitated iron catalyst (Fe/Cu/K/Si) and a carbided iron catalyst (Fe5C2 / 2% K; Table 6.6). The results clearly demonstrate the positive effect on overall conversion when operating under excess hydrogen while the penalty in doing so is most clearly seen in the CH4 selectivity.

6.3.5  Catalyst Development The ideal iron catalyst for RWGS-FT tandem processes will need to adhere to the following minimum requirements for successful operation: Given the higher water reactor partial pressures experienced by the catalyst in a one-step process, the active iron carbide phase will need to be stabilized against oxidation. ● Depending on the target reactor temperature, the iron catalyst will need sufficient surface area to sustain a commercially acceptable reaction rate. ● The presence of two catalytically active Fe phases is required that operate in tandem. The first phase corresponds to Fe3O4 which catalyzes the RWGS reaction, while the second Fe phase (χ-Fe5C2) catalyzes the FT process. ● Depending on the H2:CO2 inlet ratio and reactor temperature, the catalyst will need adequate alkali promotion to enable sufficient chain growth for distillate range hydrocarbons. ●

6.3.6  Stability to Oxidation by Water Water is known to inhibit the FT reaction over iron-based catalysts and contribute significantly to the oxidation of the FT-active iron carbide phase with time-on-line causing deactivation [200]. This deactivation is more pronounced at lower temperatures (230 °C) compared to higher reaction temperatures (270  °C). Under CO2 hydrogenation conditions, Hägg carbide may be oxidized to magnetite depending on the chemical potential of oxygen, µO, which can be related to the partial pressure of the oxidizing agents, H2O or CO2. In a systematic study on the oxidation behavior of Table 6.6  Effect of H2/CO2 molar ratios for CO2 hydrogenation using an Fe carbided catalyst. Catalyst

Feed H2/CO2 % CO2 conversion

Fe/Cu/K/Si

7.0

Fe5C2/2%K

7.0

3.0 23

1.7

21

37

13

CH4

64

16.5

6.2

4.2

C2+

36

83.5

93.8

95.8

% olefin in C2-C4

28

80

95

99

Selectivity (based on C1+)

6.3  An Overview of Fe Catalysts: Direct Route for CO2 Conversion

both H2O and CO2 on Hägg carbide under realistic high temperature FT process conditions, an in situ magnetometer was used to follow the oxidation behavior [201]. The results indicated that the oxidation of Hägg carbide and concurrent catalyst deactivation at these conditions are strongly dependent on the H2O levels present in the reactor. No oxidation was observed at CO2 levels up to 8 bar, whereas H2O induced oxidation was observed at a level of 4 bar during 3 to 20 hours of exposure, in agreement with thermodynamic calculations. It was speculated by De Smit et al. that the porosity of the final activated iron catalyst as well as the degree of amorphous carbides could play a key role in stability to oxidation by water [202]. By example, it was shown that an iron catalyst carbided in pure CO at 280 °C contained mainly Hägg carbide, and was shown to be very active and selective for conventional FT but susceptible to oxidation. The oxidation of the catalyst was shown to be more pronounced on the local coordination scale (EXAFS) compared to the long-range crystalline order (XRD), indicating the possible preferential oxidation of amorphous carbide phases. The porous nature of the catalyst material, induced by the high carburization rates at low temperatures (∼280 °C), is likely to contribute to the high extent of oxidation. In contrast, an iron catalyst pretreated in 1% CO/H2 at 350  °C contained approximately 50% crystalline θ-Fe3C, and 50% Hägg carbide, along with a high fraction of amorphous FexC species. The catalyst proved to be much more susceptible to the buildup of surface graphitic carbon species during FTS. It was speculated that the lower porosity of the catalyst, induced by the carburization at higher temperatures (∼350 °C), led to a lower susceptibility to oxidation.

6.3.7  Sufficient Surface Area The specific preparation method and addition of structural promoters are key levers in determining the final active metal surface area available for CO2 adsorption and FTS using iron ­catalysts. In general, conducting the CO2 hydrogenation reaction at relatively low temperatures (220–250 °C) would definitely require some form of supported iron catalyst to ensure adequate dispersion and surface area, while for operation at higher reaction temperatures (> 280 °C) an unsupported iron catalyst prepared via precipitation or fusion would suffice.

6.3.8  Availability of Two Distinct Catalytically Active Sites/phases Schulz et al. [203] studied the temporal changes of product composition together with changes of the catalyst in composition and structure by investigating the FT synthesis with an alkalised precipitated Fe catalyst at 250 °C, 10 bar, H2/CO2 = 3 in a fixed bed reactor. In principle, it was found that the same “episodes” of self organisation are observed as seen with H2/CO as the feed gas. In principle, it was found that the same episodes of self organization are observed as seen with H2/CO as the feed gas. Only in episode III, (after about three hours) do the first FT products appear (Figure 6.28). The RWGS-activity increases further. This implies fast reversible hydrogen chemisorption and hydrogen spillover to the FT sites which might enhance FT

Figure 6.28  Yields of carbiding (Ycarb), reverse water gas shift reaction (YRWGS) and Fischer-Tropsch synthesis (FTS; YFT) as a function of time. Reproduced with permission from Ref [203] / Springer Nature.

105

106

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

reaction. Interestingly, Fiato et al. [191] also reported that the amount of iron oxide, whether it be included in the phase or found on the catalyst surface, should be kept to a minimum. The iron catalyst is less selective for olefinic hydrocarbons and less active in hydrogenating CO2 when the oxide is present.

6.3.9  Sufficient Alkalinity for Adsorption and Chain Growth For synthesis using H2/CO2 mixtures, relatively high amounts of alkali are applied in order to enhance CO2 conversion in the FTS [204]. Alkali is an essential promoter of the oxidic catalysts for the WGS and RWGS reactions [205], and alkali present in the iron catalyst promotes carbon formation through CO dissociation, a pre-requisite for the essential carbiding of the iron catalyst even at low CO partial pressure. It is also known that the addition of alkali metals enhances adsorption of carbon oxides on reduced iron species but adsorption of hydrogen is rather weakened. It was postulated that potassium enhances both the strength and coverage of the CO2 adsorption on the catalyst surface that leads to an increase in CO2 conversion. Enhanced CO2 conversion could in turn, provide more CO species that are strongly adsorbed onto the catalyst surface for subsequent FT synthesis [206]. In a systematic study on the reactivity of CO2 over K-promoted 100Fe/10Zn/1Cu catalysts, Martinelli et al. [207] found that substantial amounts of potassium promoter were needed to achieve acceptable selectivity to middle distillates. In recent articles by Li et al. [208] and Yao et al. [209], a good summary is given of different catalytic systems used in CO2 hydrogenation. The review indicated that although iron-based catalysts are the dominant choice for most research groups, there are certainly novel catalytic systems based on other metals to take note of. These include CeO2–Pt@mSiO2–Co, ZnGa2O4/SAPO-34, ZnZrO/SAPO-34 [210], and In–Zr/SAPO-34 [211]. In general, the preferred reactor temperatures were in the range of 300–350  °C. In Yao’s work, iron-based catalysts were prepared using the organic combustion method (OCM) and subsequently evaluated for the direct and efficient conversion of CO2 to jet fuel range hydrocarbons. In summary, a Fe–Mn–K catalyst showed a CO2 conversion of 38.2% and selectivity to C8–C16 hydrocarbons of 47.8% while maintaining a low selectivity for CH4. Looking into possible new trends with regards to iron-based catalysts for direct CO2 hydrogenation, the commercialization of this technology on large scale will necessitate some form of catalyst regeneration. For traditional iron-based FT catalysts, carbon deposition is considered as one of the major deactivation mechanisms, via blocking of active sites and/or weakening of the catalyst particles. Studies of the regeneration of spent iron FT catalysts have focused on controlled removal of the carbon deposits with oxidative treatments [212–214]. Although it is accepted that for ironbased catalysts the raw material used is currently relatively cheap compared to that for cobaltbased catalysts, recycling of spent catalyst will become much more important as the global focus on sustainability increases.

6.4  Future Perspectives This chapter has given an overview of the state-of-the art for cobalt and iron FT catalysts that can be used in the production of SAF. Currently cobalt is the catalyst of choice for SAF production due to its efficiency in carbon conversion as well as suitability of the product for application as jet fuel. Critical aspects for competitive cobalt catalysts are high per pass conversion and maximum

6.4  Future Perspectives

selectivity to enable high jet fuel yields. Further developments in catalyst and process design will be required to achieve sustainability targets, as discussed in this section. Small scale, decentralized PtL plants and electrification of reactors for syngas production Coal and natural gas feedstocks for syngas production and FTS are largely fungible with ­biomass, waste, and and other renewable feedstocks, provided potential catalyst poisons are within acceptable ranges. While the FT catalyst is agnostic to the syngas source, the production of renewable syngas at required cost, efficiency, and scale remains a current challenge. Co-electrolysis ­technology to generate syngas from water and unavoidable CO2 will continue to improve in terms of capital and operating costs. Another concept for production of renewable syngas includes electrification of reactors for syngas generation such as Topsoe’s eSMR™ and eRWGS™ emerging technologies [215]. The idea is to drive endothermic reactions such as CO2 conversion in RWGS by electrical heating to afford high conversions and eliminate CO2 emission points, while significantly reducing equipment size (up to 100-fold). Most conventional FT processes currently operate at a large scale (Table 6.1), but it is envisaged that new PtL plants for SAF production, at least initially, will be small scale and decentralized to reduce complexity, capital costs and availability of renewable feedstock [37]. Micro-channel or micro-structured reactors are expected to lead the way here [216]. Due to excellent mass transfer, high space-time yields which are nearly 100 times greater compared to conventional slurry-phase reactors, can be reached. The compact and modular plant design allows building of plants with small footprints and the simplified setup can enable decentralized application such as offshore or remote solutions. Feed and flow flexibility are significantly different compared to conventional FT plants and the catalyst in a microchannel reactor needs to be able to withstand this variable feed flow due the intermittent nature of H2 production from renewable energy sources. The modularity and compact nature of the microchannel reactors allows for handling of intermittent feed by being able to drop load to certain reactors or perform serial operation [217]. To maximise efficiency on this small scale, catalysts need to be robust to the varying conditions, affording enhanced conversion while still producing the desired molecules. A Sasol commercial catalyst has been proven to be highly effective for application in INERATEC’s micro-structured reactors under a range of conditions [218] and further improvements in performance can be expected with newly developed catalysts. Greener and more sustainable design and preparation routes for catalysts Catalysis is recognized as an important principle to enable green chemistry [219]. However, to be truly green production of the catalysts themselves needs to be sustainable. Commercial catalyst preparation typically balances cost and practicality, while seeking to reduce unwanted emissions. An example is the use of nitrate salts in large scale catalyst preparation. Cobalt nitrate has a high density of cobalt, is highly soluble in water and thus allows high filling of the pores during impregnation, reducing the amount of impregnation steps required to produce a certain cobalt loading [51]. However, during calcination and reduction, NOX and NH3 are released as by-products [107] and need to be treated before off-gases can be released into the atmosphere. Going forward, the desire would be to eliminate these harmful byproducts by choosing alternative cobalt sources. Similarly, transitioning away from the organic solvents often used in support modification steps [220] to more benign aqueous routes, is a current goal. In future, it is foreseeable that electrical power for the operation of catalyst plants will be generated via renewable energy, while H2 for catalyst reduction could be produced via green electrolysis routes. Due to its cost and unsustainable mining practices [221], it is desirable to optimize cobalt utilization by dispersing it as efficiently as possible on the support. Reducing the level of cobalt and

107

108

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

promoter, while still maintaining performance characteristics is another element for the design of greener catalysts. Examples of this include incorporating core-shell design concepts to produce cobalt catalysts [222] and replacing some cobalt with less expensive Ni in bi-metallic alloy catalysts as recently shown by Sasol [15]. The latter example is also important as it shows the use of theoretical modelling to predict or assist in design of catalysts as well as advances in situ characterization to study the behavior of the alloy systems under relevant reaction conditions. Undoubtedly, advancement in high resolution in operando techniques will assist in the understanding of catalyst performance under representative PtL conditions as well as catalyst and process optimization. Continuous investment in researching catalyst fundamentals yields significant rewards [223, 224]. Ultimately, an integrated approach using experimental data, theory, modeling and digital platforms is expected to take the understanding of transition metal catalysis to the next level [225]. Incorporation of digitalization and artificial intelligence in catalyst design [226] could also assist in the formulation of new catalysts, potentially reducing the intensity of lab and pilot scale experiments. Beyond SAF toward value added green chemicals SAF production employing FT technology with renewable feedstocks, is seen as the first step in the greening of FT [227]. The non-selective nature of FT can be a boon as it provides a route to sustainable chemicals which have a higher margin than jet fuel. The alpha values of cobalt catalysts can be tuned to produce green hard wax which has wide applications in hot-melt adhesives, inks, coatings and cosmetics. These catalysts can also produce green base oils and lubricants in a highly efficient manner [61]. For example, cobalt FT catalysts can produce Group 3+base oils with a low sulfur, high aliphatic content, and a high viscosity index [228]. Although, it is more difficult to alter the product slate of cobalt catalysts than iron catalysts with promotion, there have been advances in the tuning of cobalt catalysts to be more selective to other molecules besides paraffinic hydrocarbons. Doping of ceria onto cobalt catalysts could increase the formation of olefins and oxygenates to around 15% of the total product [229], while other reports indicate that Co-Mn catalysts can enable production of significant amounts of lower alcohols [230]. Going forward it is expected that more research will be conducted on improving and diversifying the chemical selectivity of cobalt catalysts. Because the product spectrum of HTFT catalysts is mostly olefinic in nature, the possibility to produce green olefins as a feedstock for green polymers and other chemicals also exists. Various research groups have reported on FT to olefins (FTO) catalysts [231, 232], and Sasol has also investigated Fe-based olefin selective catalysts [233] that can be considered for conversion of green hydrogen and CO2 into light olefins. It is thus clear that both iron and cobalt catalysts can provide not only a route for SAF production but also to value-added chemicals.

References 1 Air Transport Action Group. The economic & social benefits of air transport. https://www.icao.int/ Meetings/wrdss2011/Documents/JointWorkshop2005/ATAG_SocialBenefitsAirTransport.pdf (accessed 1 March 2022). 2 Air Transport Action Group: Facts and Figures. https://www.atag.org/facts-figures (accessed 1 March 2022). 3 Oxford Economics. (2020). Aviation benefits beyond borders report. https://aviationbenefits.org/ media/167517/aw-oct-final-atag_abbb-2020-publication-digital.pdf (accessed 1 March 2022).

References

4 International Air Transport Association. (2015). Economic performance of the airline industry. https://www.iata.org/whatwedo/Documents/economics/IATA-Economic-Performance-of-theIndustry-end-year-2014-report.pdf (accessed 1 March 2022). 5 Doliente, S.S., Narayan, A., Tapia, J.F.D. et al. (2020). Front. Energy Res. 8: 110. 6 Dieterich, V., Buttler, A., Hanel, A. et al. (2020). Energy Environ. Sci. 13 (10): 3207–3252. 7 Gray, N., McDonagh, S., O’Shea, R. et al. 2021). Adv. Appl. Energy 1: 100008. 8 Edwards, J.T. (2017). Reference jet fuels for combustion testing. AIAA 2017–0146. 55th AIAA Aerospace Sciences Meeting. doi: 10.2514/6.2017-0146. 9 Holladay, J., Abdullah, Z., and Heyne, J. (2020). Sustainable aviation fuel: review of technical pathways. U.S. Department of Energy Office of Energy Efficiency & Renwable Energy technical report. 10 Davis, B.H. (2007). Ind. Eng. Chem. Res. 46 (26): 8938–8945. 11 Dry, M.E. (2001). Stud. Surf. Sci. Catal. 136: 63–68. 12 Agnelli, M., Kolb, M., and Mirodatos, C. (1994). J. Catal. 148 (1): 9–21. 13 Vannice, M.A. (1975). J. Catal. 37 (3): 462–473. 14 Xiao, C.X., Cai, Z.P., Wang, T. et al. (2008). Angewandte Chemie 120 (4): 758–761. 15 van Helden, P., Prinsloo, F., Van den Berg, J.A. et al. (2020). Catal. Today 342: 88–98. 16 De la pena O’Shea, V.A., Álvarez-galván, M.C. et al. (2007). Appl. Catal. A Gen. 326 (1): 65–73. 17 Voss, J.M., Xiang, Y., Collinge, G. et al. (2018). Top. Catal. 61 (9): 1016–1023. 18 Calderone, V.R., Shiju, N.R., Ferré, D.C., and Rothenberg, G. (2011). Green Chem. 13 (8): 1950–1959. 19 Iris, C. and Weckhuysen, B.M. (2021). Chem. Catal. 1 (2): 339–363. 20 Have, I.C.T., Kromwijk, J.J., and Monai, M. et al. (2022). Nat. Commun. 13 (1): 1–11. 21 Claeys, M., Dry, M.E., Van Steen, E. et al. (2014). J. Catal. 318: 193–202. 22 Moodley, D.J., Van De Loosdrecht, J., Saib, A.M., and Niemantsverdriet, J.W. (2010). The formation and influence of carbon on cobalt-based Fischer-Tropsch synthesis catalysts: an integrated review. In: Advances in Fischer-Tropsch Synthesis, Catalysts and Catalysis, (eds. B.H. Davis, M.L. Occelli, and J.G. Speight), 49–81. Boca Raton: CRC Press. 23 Claeys, M. (2016). Nature 538 (7623): 44–45. 24 Xu, J., Yang, Y., and Li, Y.W. (2013). Curr. Opin. Chem. Eng. 2 (3): 354–362. 25 Dry, M. and Steynberg, A. (eds.) (2004). Fischer-Tropsch Technology. Amsterdam: Elsevier. 26 Chang, Q., Zhang, C., Liu, C. et al. (2018). ACS Catal. 8 (4): 3304–3316. 27 Wei, J., Ge, Q., Yao, R. et al. (2017). Nat. Commun. 8 (1): 1–9. 28 Wu, B., Yang, Y., Li, Y., and Xiang, H. (2011). Synfuels China Technology Co Ltd. Fischer-Tropsch synthesis Fe-based catalyst, process of preparation and application thereof. US Patent Application 13/148,209. 29 Brown, A. (2011). Pearl GTL Presentation, XTL Summit. 30 Satterfield, C.N., Hanlon, R.T., Tung, S.E. et al. (1986). Ind. Eng. Chem. Prod. Res. Dev. 25: 407–414. 31 Labuschagne, J., Meyer, R., Chonco, Z.H. et al. (2016). Catal. Today 275: 2–10. 32 Van Steen, E., Claeys, M., Möller, K.P., and Nabaho, D. (2018). Appl. Catal. A Gen. 549: 51–59. 33 Nicholas, C.P. (2017). Appl. Catal. A Gen. 543: 82–97. 34 Font Freide, J.J.H.M., Gamlin, T.D., Hensman, J.R. et al. (2004). J. Nat. Gas Chem. 13: 1–9. 35 Sango, T., Fischer, N., Henkel, R. et al. (2015). Appl. Catal. A Gen. 502: 150–156. 36 Boerrigter, H. and Van der Drift, A. (2004). Large-scale production of Fischer-Tropsch diesel from biomass. Optimal gasification and gas cleaning systems (No. ECN-RX–04–119). Energy research Centre of the Netherlands ECN. 37 Meurer, A. and Kern, J. (2021). Energies 14 (7): 1836. 38 Baxter, I. (2012). Small-scale GTL: back on the agenda. World 85. https://compactgtl.com/ wp-content/documents/pe_worldgas_2012.pdf.

109

110

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

39 Pearson, R., Coe, A. , and Paterson, J. (2021). Johnson Matthey Technol. Rev. 65 (3): 395–403. 40 Ticehurst, P. (2020). FT CANS™ Technology enabling waste to jet fuels. Global Syngas Technologies Council 2020 Conference, 27 Oct 2020. 41 Dittmeyer, R., Boeltken, T., Piermartini, P., Selinsek, M., Loewert, M., Dallmann, F., Kreuder, H., Cholewa, M., Wunsch, A., Belimov, M., and Farsi, S. (2017. Curr. Opin. Chem. Eng. 17: 108–125. 42 Myrstad, R., Eri, S., Pfeifer, P., Rytter, E. and Holmen, A. (2009). Catal. Today 147: S301–S304. 43 Pfeifer, P., Schmidt, S., Betzner et al. (2022). Curr. Opin. Chem. Eng. 36: 100776. 44 LeViness, S., Tonkovich, A.L., Jarosch, K. et al. (2011). Improved Fischer-Tropsch economics enabled by microchannel technology. Velocys white paper. 45 LeViness, S., Schubert, P., McDaniel, J., and Dritz, T. (2012). Velocys Fischer-Tropsch technologyeconomical smaller scale GTL enabled by microchannel reactor technology and superactive catalyst. Presentation at Syngas Convention, Cape Town, South Africa, 2 April 2012. 46 Velocys (2023). Sustainable aviation fuel is critical to achieving net zero carbon aviation. https:// www.velocys.com/projects (accessed 1 March 2022). 47 Bezemer, G.L. (2019). Advances in gas-to-liquids technology at Shell. Presentation at 12th Natural Gas Conversion Symposium, San Antonio, 3 June, 2019. 48 Posdziech, O., Stäber, R., and von Olshausen, C. (2012). Synfuels from Electrolysis, Presentation at European Summer School in Fuel Cells and Hydrogen Technologies. Iraklion, Greece. 49 https://www.norsk-e-fuel.com/en (accessed 28 March 2022). 50 Seyednejadian, S., Rauch, R., Bensaid, S. et al. (2018). Appl. Sci. 8 (4): 514. 51 Van de Loosdrecht, J., Botes, F.G., Ciobica, I.M. et al. (2013). Fischer-Tropsch synthesis: catalysts and chemistry. In: Comprehensive Inorganic Chemistry II: From Elements to Applications (eds. J. Reedijk and K. Poeppelmeier), 525–557. Amsterdam: Elsevier. 52 Fischer, F. and Tropsch, H. (1923). Brennst. Chem. 4: 276–285. 53 Fischer, F., and Tropsch, H. [Hydrocarbons.] German Patent 484,337, Oct. 16, 1929. 54 (a) P. Sabatier, J.B. Senderens, C. R. Acad. Sci. Paris 134 (1902) 514 and (b) BASF, German Patent DRP 293,787 (1913). 55 Inga, J., Kennedy, P., and LeViness, S., S. Corp (2005). Fischer-Tropsch process in the presence of nitrogen contaminants. US Patent Application 10/755, 942. 56 Jermwongratanachai, T., Jacobs, G., Ma, W., et al. (2013). Appl. Catal. A Gen. 464: 165–180. 57 Riedel, T. and Schaub, G. (2003). Top. Catal. 26 (1): 145–156. 58 Moodley, D.J., Van de Loosdrecht, J., Saib, A.M. et al. (2009). Appl. Catal. A Gen. 354 (1–2): 102–110. 59 Shell Global Gas to liquids. https://www.shell.com/energy-and-innovation/natural-gas/gas-toliquids.html (accessed 31 March 2022). 60 Calis, H.P., The Shell GTL process: from lab to world-scale plant presented at EFT Gas to Market & Energy Conversion Forum, Boston. 12–14 October 2010. 61 Moodley, D., Botha, T., Potgieter, J. et al. (2019). Development of highly selective, water tolerant cobalt catalysts for Fischer-Tropsch Synthesis, Presentation at 12th Natural Gas Conversion Symposium, San Antonio, USA, 4 June 2019. 62 Data taken from Shell Patents: EP 428223, EP 398420, EP 455307, EP 510770, GB 2258826, WO 97/00231, WO 98/11037, WO 98/25870, WO 99/34917, WO 01/76736, WO 02/02489, WO 02/05958, WO 02/07882, WO06/067177, WO 06/056610, WO 08/006885, WO 08/061970, WO 08/071640, WO 08/090105; 1989 to 2007. 63 Rytter, E. and Holmen, A. (2015). Catalysts 5 (2): 478–499. 64 BP (2018). BP and Johnson Matthey license innovative waste-to-fuels technology to biofuels producer Fulcrum BioEnergy. https://www.bp.com/en/global/corporate/news-and-insights/ press-releases/bp-and-johnson-matthey-license-innovative-waste-to-fuels-technology-to-biofuelsproducer-fulcrum-bioenergy.html (accessed 1 March 2022).

References

65 Ferguson, E., Krawiec, P., Wells, M.J., and BP PLC. (2020). Fischer Tropch process using supported reduced cobalt catalyst. US Patent 10,717,075. 66 Soled, S.L., Fiato, R.A., and Iglesia, E. (1989). Cobalt-ruthenium catalyst for Fischer-Tropsch synthesis. Eur. Pat. Appl. 87310896. 67 Compact GTL (2023). The modular gas solution. http://www.compactgtl.com (accessed 3 March 2023). 68 Srinivasan, N., Espinoza, R.L., Coy, K.L., Jothimurugesan, K., and ConocoPhillips Co. (2004). Fischer-Tropsch catalysts using multiple precursors. US Patent 6,822,008. 69 Espinoza, R. and Agee, K. (2016). Patent US 9 358 526 B2, 7 June 2016. 70 Bellussi, G., Carluccio, L.C., Zennaro, R., and Del Piero, G. (2007). Process for the preparation of Fischer-Tropsch catalysts with a high mechanical, thermal and chemical stability. Patent WO2007009680 A1, January 2007. 71 Rytter, E., Skagseth, T.H., Wigum, H., and Sincadu, N. (2005). Enhanced strength of alumina-based Co Fischer-Tropsch catalyst. Patent WO, 2005072866. 72 Moodley, D., Botha, T., Potgieter, J. et al. (2002). Optimal supported cobalt Power-to-Liquids (PtL) catalysts for Sustainable Aviation Fuel. 27th North American Meeting (NAM27) of the North American Catalysis Society, New York. 73 Ikeda, M., Waku, T., and Aoki, N. (2005). Catalyst for Fisher-Tropsch synthesis and method for producing hydrocarbon. Patent WO2005099897 A, 1. 74 Botha, J.M., Moodley, D.J., Potgieter, J.H., Van Rensburg, H., van de Loosdrecht, J., Moodley, P., and Sasol South Africa Pty Ltd. (2021). A cobalt-containing catalyst composition. US Patent Application 16/319,674. 75 Dogterom, R.J., Mesters, C.M.A.M., Reynhout, M.J., and Shell Oil Company (2011). Process for preparing a hydrocarbon synthesis catalyst. US Patent 8,003,564. 76 Inga, J., Kennedy, P., LeViness, S., and Syntroleum Corp. (2005). Fischer-Tropsch process in the presence of nitrogen contaminants. US Patent Application 10/755,942. 77 Velocys. (2022). Think Smaller. http://www.velocys.com/resources/Velocys_Booklet.pdf (accessed 1 March 2022). 78 Munnik, P., de Jongh, P.E., and de Jong, K.P. (2015). Chem. Rev. 115 (14): 6687–6718. 79 Jahangiri, H., Bennett, J., Mahjoubi, P., Wilson, K., and Gu, S. (2014). Catal. Sci. Technol. 4 (8): 2210–2229. 80 Bartholomew, C.H. and Farrauto, R.J. (2011). Fundamentals of Industrial Catalytic Processes. Hoboken: John Wiley & Sons. 81 Stranges, A. N. (2003). Germany’s synthetic fuel industry 1927–1945. Presented at theAIChE 2003, New Orleans, 2003. http://www.fischertropsch.org (accessed 1 April 2023). 82 Baijense, C.R. and Rekker, T., and BASF Corporation. (2011). Fischer-Tropsch catalyst. US Patent 7,884,139. 83 Van Berge, P.J., Visagie, J.L., Van De Loosdrecht, J., Van Der Walt, T.J., Sollie, J.C., Veltman, H.M., Engelhard de Meern, B.V., and Sasol Technology Pty Ltd. (2009). Process for activating cobalt catalysts. US Patent 7,592,289. 84 Puskas, I., Fleisch, T.H., Full, P.R. et al. (2006). Appl. Catal. A Gen. 311: 146–154. 85 Munirathinam, R., Pham Minh, D., and Nzihou, A. (2018). Ind. Eng. Chem. Res. 57 (48): 16137–16161. 86 Van Berge, P.J., Van De Loosdrecht, J., Caricato, E.A., Barradas, S., and Sasol Technology Pty Ltd. (2003). Process for producing hydrocarbons from a synthesis gas, and catalysts therefore. US Patent 6,638,889. 87 Van Berge, P.J., van de Loosdrecht, J., Visagie, J.L., and Sasol Technology Pty Ltd. (2004). Cobalt catalysts. US Patent 6,835,690.

111

112

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

88 Botha, J.M., Visagie, J.L., Cullen, A., Taljaard, J.H., Meyer, R., and Sasol Technology Pty Ltd. (2020). Hydrocarbon synthesis catalyst, its preparation process and its use. US Patent 10,569,255. 89 Richard, L., Jarosch, K., Robota, H.J., Leonarduzzi, D., Roberts, D., and Velocys Technologies Ltd. (2017). Cobalt-containing Fischer-Tropsch catalysts, methods of making, and methods of conducting Fischer-Tropsch synthesis. US Patent 9,707,546. 90 Cho, E.H., Koo, K.Y., Lee, H.W. et al. (2017). Int. J. Hydrogen Energy 42 (29): 18350–18357. 91 Van Berge, P.J., van de Loosdrecht, J., Caricato, E.A., Barradas, S., Sigwebela, B.H., and Sasol Technology Pty Ltd. (2002). Impregnation process for catalysts. US Patent 6,455,462. 92 Neimark, A.V., Kheifets, L.I., and Fenelonov, V.B. (1981). Ind. Eng. Chem. Prod. Res. Dev. 20 (3): 439–450. 93 de Jong, K.P. (ed.) (2009). Synthesis of Solid Catalysts. Hoboken: John Wiley & Sons. 94 Van Berge, P. J. (2000). In: Scaling up of an alumina supported cobalt slurry phase Fischer–tropsch catalyst preparation, CatCon – World Wide Catalyst Industry Conference, Houston. 95 Van Berge, P.J., van de Loosdrecht, J., Caricato, E.A., Barradas, S., Sigwebela, B.H., and Sasol Technology Pty Ltd. (2002). Impregnation process for catalysts. US Patent 6,455,462. 96 Mochizuki, T., Koizumi, N., Hamabe, Y. et al. (2007). J. Jpn. Pet. Inst. 50 (5): 262–271. 97 Van Dillen, A.J., Terörde, R.J., Lensveld, D.J. et al. (2003). J. Catal. 216 (1–2): 257–264. 98 Terorde, R.J.A.M., Kramer, L.L., and Sasol Technology Pty Ltd. (2015). Process for the preparation of fischer-tropsch catalysts and their use. US Patent 8,841,229. 99 Daly, F., Richard, L., Rugmini, S., and Velocys Technologies Lit. (2016). Catalysts. US Patent 9,381,501. 100 van de Loosdrecht, J., Datt, M.S., Botha, J.M., and Sasol Technology Pty Ltd. (2013). Catalysts. US Patent 8,394,864. 101 Poncelet, G., Martens, J.A., Delmon, B. et al. (eds.) (1995). Preparation of Catalysts VI: Scientific Bases for the Preparation of Heterogeneous Catalysts. Amsterdam: Elsevier. 102 De Jong, K.P. (1991). Deposition precipitation onto pre-shaped carrier bodies. Possibilities and limitations. In: Studies in Surface Science and Catalysis, vol. 63 (eds. G. Poncelet, P.A. Jacobs, P. Grange, and B. Delmon), 19–36. Amsterdam: Elsevier. 103 Lekhal, A., Glasser, B.J., and Khinast, J.G. (2001). Chem. Eng. Sci. 56 (15): 4473–4487. 104 Munnik, P., Krans, N.A., De Jongh, P.E., and De Jong, K.P. (2014). ACS Catal. 4 (9): 3219–3226. 105 Sietsma, J.R.A., Van Dillen, A.J., De Jongh, P.E., De Jong, K.P., and Johnson Matthey PLC. (2012). Metal nitrate conversion method. US Patent 8,263,522. 106 Wolters, M., Munnik, P., Bitter, J.H., De Jongh, P.E., De Jong, K.P., and Johnson Matthey PLC. (2012). Method for producing a supported metal nitrate. US Patent Application 13/258,846. 107 Van de Loosdrecht, J., Barradas, S., Caricato, E.A. et al. (2003). Top. Catal. 26 (1): 121–127. 108 Gauché, J.L., Pienaar, C., Swart, J.C.W., Botha, J.M., Moodley, D.J., Potgieter, J.H., Davel, J.Z., and Sasol South Africa Limited. (2022). Process for preparing a cobalt-containing catalyst precursor and process for hydrocarbon synthesis. US Patent Application 17/312,376. 109 Barradas, S., Eloff, C.C., Visagie, J.L., and Sasol Technology Pty Ltd. (2017). Process for preparing a cobalt-containing hydrocarbon synthesis catalyst precursor. US Patent 9,687,822. 110 Soled, S.L., Baumgartner, J.E., Reyes, S.C. et al. (1995). Synthesis of eggshell cobalt catalysts by molten salt impregnation techniques. In: Studies in Surface Science and Catalysis, vol. 91. (eds. G. Poncelet, J. Martens, B. Delmon et al.), 989–997. Amsterdam: Elsevier. 111 Van Rensburg, H., and Sasol Technology Pty Ltd. (2016). Process for preparing a cobalt-containing Fischer-Tropsch catalyst. US Patent 9,248,435. 112 Hoek, A., Moors, J.H., and Shell Oil Company. (2002). Catalyst activation process. US Patent 6,475,943. 113 Visagie, J.L., Veltman, H.M., Engelhard de Meern, B.V., and Sasol Technology Pty Ltd. (2009). Producing supported cobalt catalysts for the Fisher-Tropsch synthesis. US Patent 7,524,787.

References

114 Behrmann, W.C., Davis, S.M., Mauldin, C.H., and Exxon Research and Engineering Company. (1992). Method for preparing cobalt-containing hydrocarbon synthesis catalyst. Patent WO1992006784A1. 115 Visagie, J.L., Botha, J.M., Koortzen, J.G., Datt, M.S., Bohmer, A., van de Loosdrecht, J., Saib, A.M., and Sasol Technology Pty Ltd. (2011). Catalysts. US Patent 8,067,333. 116 Oosterbeek, H. (2007). Phys. Chem. 9 (27): 3570–3576. 117 Hauman, M.M., Saib, A., Moodley, D.J. et al. (2012). ChemCatChem 4 (9): 1411–1419. 118 Patanou, E., Tsakoumis, N.E., Myrstad, R., and Blekkan, E.A. (2018). Appl. Catal. A Gen. 549: 280–288. 119 Van Rensburg, H., and Sasol Technology Pty Ltd. (2016). Process for preparing a Fischer-Tropsch catalyst. US Patent 9,387,463. 120 Rytter, E. and Holmen, A. (2017). ACS Catal. 7 (8): 5321–5328. 121 Okoye-Chine, C.G., Moyo, M., Liu, X., and Hildebrandt, D. (2019). Fuel Process. Technol. 192: 105–129. 122 Dalai, A.K. and Davis, B.H. (2008). Appl. Catal. A Gen. 348 (1): 1–15. 123 Bertole, C.J., Kiss, G., and Mims, C.A. (2004). J. Catal. 223 (2): 309–318. 124 Tucker, C.L. and van Steen, E. (2020). Catal. Today 342: 115–123. 125 Kiss, G., Kliewer, C.E., DeMartin, G.J. et al. (2003). J. Catal. 217 (1): 127–140. 126 Kliewer, C.E., Soled, S.L., and Kiss, G. (2019). Catal. Today 323: 233–256. 127 van Berge, P.J., van de Loosdrecht, J., Caricato, E.A, Barradas, S. (1999). Patent PCT/GB 99/00527 to Sasol Technology. 128 Moodley, D.J., Saib, A.M., van de Loosdrecht, J. et al. (2011). Catal. Today 171 (1): 192–200. 129 Moodley, D., Claeys, M., van Steen, E. et al. (2020). Catal. Today 342: 59–70. 130 Claeys, M., Dry, M.E., van Steen, E. et al. (2015). ACS Catal. 5 (2): 841–852. 131 Gholami, Z., Tišler, Z., and Rubáš, V. (2021). Catal. Rev. 63 (3): 512–595. 132 Bartholomew, C.H., Rahmati, M., and Reynolds, M.A. (2020). Appl. Catal. A Gen. 602: 117609. 133 Ojeda, M., Pérez-Alonso, F.J., Terreros, P. et al. (2006). Langmuir 22 (7): 3131–3137. 134 Valero-Romero, M.J., Rodríguez-Cano, M.Á., Palomo, J. et al. (2021). Front. Mater. 7: 455. 135 De la Osa, A.R., De Lucas, A., Diaz-Maroto, J. et al. (2012). Catal. Today 187 (1): 173–182. 136 Bezemer, G.L., Radstake, P.B., Koot, V. et al. (2006). J. Catal. 237 (2): 291–302. 137 Lillebø, A., Håvik, S., Blekkan, E.A., and Holmen, A. (2013). Top. Catal. 56 (9): 730–736. 138 Liu, Y., Florea, I., Ersen, O. et al. (2015). Chem. Commun. 51 (1): 145–148. 139 Van de Loosdrecht, J., Datt, M., and Visagie, J.L. (2014). Top. Catal. 57 (6): 430–436. 140 Cheng, K., Subramanian, V., Carvalho, A. et al. (2016). J. Catal. 337: 260–271. 141 Bezemer, G.L., Bitter, J.H., Kuipers, H.P. et al. (2006). J. Am. Chem. Soc. 128 (12): 3956–3964. 142 Weststrate, C.J., Van Helden, P., and Niemantsverdriet, J.W. (2016). Catal. Today 275: 100–110. 143 van Helden, P., Ciobîcă, I.M., and Coetzer, R.L. (2016). The size-dependent site composition of FCC cobalt nanocrystals. Catal. Today 261: 48–59. 144 Tuxen, A., Carenco, S., Chintapalli, M. et al. (2013). J. Am. Chem. Soc. 135 (6): 2273–2278. 145 Saib, A.M., Borgna, A., Van de Loosdrecht, J. et al. (2006). J. Catal. 239 (2): 326–339. 146 Víctor, A., Galván, M.C.Á., Prats, A.E.P. et al. (2011). Chem. Commun. 47 (25): 7131–7133. 147 Bertella, F., Concepción, P., and Martínez, A. (2017). Catal. Today 296: 170–180. 148 Dalai, A.K. and Davis, B.H. (2008). Appl. Catal. A Gen. 348 (1): 1–15. 149 Jacobs, G., Ribeiro, M.C., Ma, W. et al. (2009). Appl. Catal. A Gen. 361 (1–2): 137–151. 150 Morales, F., Grandjean, D., Mens, A. et al. (2006). J. Phys. Chem. B 110 (17): 8626–8639. 151 Morales, F., de Smit, E., de Groot, F.M. et al. (2007). J. Catal. 246 (1): 91–99. 152 Potgieter, J. (2018). The effect of support pore diameter and Mn loading on catalyst selectivity for Co/Mn-Ti/alumina catalysts. Presentation at Catalysis Society of South Africa, Waterberg.

113

114

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

153 den Breejen, J.P., Frey, A.M., Yang, J. et al. (2011). Top. Catal. 54 (13): 768–777. 154 Dinse, A., Aigner, M., Ulbrich, M. et al. (2012). J. Catal. 288: 104–114. 155 Feltes, T.E., Espinosa-Alonso, L., Smit, E. et al. (2010). J. Catal. 270 (1): 95–102. 156 Gupta, S.S., Shenai, P.M., Meeuwissen, J. et al. (2021). J. Phys. Chem. C 125 (39): 21390–21401. 157 Tucker, C.L., Bordoloi, A., and van Steen, E. (2021). Sustain. Energy Fuels 5 (22): 5717–5732. 158 Dalai, A.K. and Davis, B.H. (2008). Appl. Catal. A Gen. 348 (1): 1–15. 159 Khodakov, A.Y. (2009). Catal. Today 144 (3–4): 251–257. 160 Borg, Ø., Eri, S., Blekkan, E.A. et al. (2007). J. Catal. 248 (1): 89–100. 161 Saib, A.M., Claeys,M., and van Steen, E. (2002). Catal. Today 71: 395–402. 162 Storsæter, S., Borg, Ø., Blekkan, E.A., and Holmen, A. (2005). J. Catal. 231: 405–419. 163 Jacobs, G., Ma, W., and Davis, B. (2014). Catalysts 4 (1): 49–76. 164 Liu, J.-X., Su, H.-Y., Sun, D.-P., Zhang, B.-Y., and Li, W.-X. (2013). J. Am. Chem. Soc. 135: 16284–16287. 165 Lin, H., Liu, J.X., Fan, H.J., and Li, W.X. (2020). J. Phys. Chem. C 124 (42): 23200–23209. 166 Nay, B., Smith, M.R., and Telford, C.D. (1998). Catalyst treatment. US Patent 5,728,918. 167 Braconnier, L., Landrivon, E., Clémençon, I. et al. (2013). Catal. Today 215: 18–23. 168 Mauldin, C. and Varnado, D. (2001). Rhenium as a promoter of titania-supported cobalt FischerTropsch catalysis. In: Studies in Surface Science and Catalysis 136. (eds. E. Iglesia, J.J Spivey, and T.H. Fleisch), 417–422. Amsterdam: Elsevier Science. 169 Lyu, S., Wang, L., Zhang, J. et al. (2018). ACS Catal. 8 (9): 7787–7798. 170 Sadeqzadeh, M., Karaca, H., Safonova, O.V. et al. (2011). Catal. Today 164 (1): 62–67. 171 Gnanamani, M.K., Jacobs, G., Shafer, W.D., and Davis, B.H. (2013). Catal. Today 215: 13–17. 172 Jalama, K., Kabuba, J., Xiong, H., and Jewell, L.L. (2012). Catal. Commun. 17: 154–159. 173 Brumby, A., Verhelst, M., Cheret, C. (2005). Catal. Today 106 (1–4): 166–169. 174 Saib, A.M., Moodley, D.J., Ciobîcă, I.M. et al. (2010). Catal. Today 154 (3–4): 271–282. 175 Bartholomew, C. H. (2013). Recent Progress in the Science and Technology of Deactivation and Regeneration of Cobalt Fischer−Tropsch Catalysts in the Catalyst Review. 26. 7. Spring House: The Catalyst Group Resources, Inc. 176 Rytter, E. and Holmen, A. (2015). Catalysts 5 (2): 478–499. 177 Argyle, M.D. and Bartholomew, C.H. (2015). Catalysts 5 (1): 145–269. 178 Overtoom, R., Fabricius, N., Leenhouts, W. et al. (2009). From bench scale to world scale. In: Proceedings of the 1st Annual Gas Processing Symposium (eds. H.E. Alfadala, G.V.R. Reklaitis, and M.M. El-Halwagi), 378–386. Amsterdam: Elsevier. 179 van Burgt, J. M., Ansorge, J. (Shell International Research Maatschappij B.V., NL). Regeneration process for a Fischer−Tropsch catalyst. Great Britain Patent 2222531 A, 30 June 1988. 180 Eilers, J., Tijm, P.J.A. (Shell International Research Maatschappij B.V., NL). (1991). Process for the activation of a catalyst. Great Britain Patent 2258826 A, 20 August 1991. 181 Bezemer, G.L., Nkrumah, S., Oosterbeek, H., Stobbe, E.R., and Shell Oil Company. (2010). Process for regenerating a catalyst. US Patent Application 12/639,707. 182 Font Freide, J.J.H.M., Gamlin, T.D., Hensman, J.R. et al. (2004). J. Nat. Gas Chem. 13 (1): 1–9. 183 Clavenna, L.R., Woo, H.S., Mauldin, C. H., and Wachter, W.A. (ExxonMobil Research and Engineering Company, USA). (2000). Cobalt catalyst compositions useful for conducting carbon monoxide hydrogenation reactions. US Patent 6521565 B1, 1 August 2000. 184 Huang, J.R., Arcuri, K., Agee, K., and Schubert, P.F. (Syntroleum Corporation, USA). (2002). Process for regenerating a slurry Fischer−Tropsch catalyst. US Patent 6989403 B2, 12 April, 2002. 185 Saib, A.M., Gauché, J.L., Weststrate, C.J. et al. (2014). Ind. Eng. Chem. Res. 53 (5): 1816–1824. 186 Hwang, J.S., Jun, K.W., and Lee, K.W. (2001). Appl. Catal. A Gen. 208 (1–2): 217–222.

References

187 Fiato, R.A., Soled, S.L., Rice, G.W., Miseo, S., and Exxon Research and Engineering Co. (1992). Method for producing olefins from H2 and CO2 using an iron carbide-based catalyst. US Patent 5,140,049. 188 Herranz, T., Rojas, S., Pérez-Alonso, F.J. et al. (2006). Appl. Catal. A Gen. 311: 66–75. 189 Lee, J.F., Chern, W.S., Lee, M.D., and Dong, T.Y. (1992). Can. J. Chem. Eng. 70 (3): 511–515. 190 VanderWiel, D.P., Zilka-Marco, J.L. et al. (2000). Carbon Dioxide Conversions in Microreactors. American Institute of Chemical Engineers. 191 Fiato, R.A., Iglesia, E., Rice, G.W. et al. (1998). Iron catalyzed CO2 hydrogenation to liquid hydrocarbons. In: Studies in Surface Science and Catalysis 114. (eds. T. Inui, M. Anpo, K. Izui et al.), 339–344. Amsterdam: Elsevier. 192 Xu, L., Bao, S., Houpt, D.J. et al. (1997). Catal. Today 36 (3): 347–355. 193 Krishnamoorthy, S., Li, A., and Iglesia, E. (2002). Catal. Lett. 80 (1): 77–86. 194 Miller, J.E. (2007). Initial case for splitting carbon dioxide to carbon monoxide and oxygen (No. SAND2007-8012). Sandia National Laboratories (SNL), Albuquerque, NM, and Livermore, CA (United States). 195 de Klerk, A. (2011). Energy Environ. Sci. 4 (4): 1177–1205. 196 Kumabe, K., Sato, T., Matsumoto, K., Ishida, Y., and Hasegawa, T. (2010). Fuel 89 (8): 2088–2095. 197 Riedel, T., Claeys, M., Schulz, H. et al. (1999). Appl. Catal. A Gen. 186 (1–2): 201–213. 198 Riedel, T., Schaub, G., Jun, K.W., and Lee, K.W. (2001). Ind. Eng. Chem. Res. 40 (5): 1355–1363. 199 Botes, F.G., Niemantsverdriet, J.W., and Van De Loosdrecht, J. (2013). Catal. Today 215: 112–120. 200 Pendyala, V.R.R., Jacobs, G., Mohandas, J.C. et al. (2010). Catal. Lett. 140 (3): 98–105. 201 Claeys, M., van Steen, E., Botha, T. et al. (2021). ACS Catal. 11 (22): 13866–13879. 202 de Smit, E., Cinquini, F., Beale, A.M. et al. (2010). J. Am. Chem. Soc. 132 (42): 14928–14941. 203 Schulz, H., Riedel, T., and Schaub, G. (2005). Top. Catal. 32 (3): 117–124. 204 Lee, J.F., Chern, W.S., Lee, M.D., and Dong, T.Y. (1992). Can. J. Chem. Eng. 70 (3): 511–515. 205 Ando, H., Matsumura, Y., and Souma, Y. (2000). J. Mol. Catal. A Chem. 154 (1–2): 23–29. 206 Yan, S.R.,Jun, K.W., Hong, J.S. et al. (2000). Appl. Catal. A Gen. 194: 63–70. 207 Martinelli, M., Visconti, C.G., Lietti, L. et al. (2014). Catal. Today 228: 77–88. 208 Li, W., Wang, H., Jiang, X. et al. (2018). RSC Adv. 8 (14): 7651–7669. 209 Yao, B., Xiao, T., Makgae, O.A. et al. (2020). Nat. Commun. 11 (1): 1–12. 210 Li, Z., Wang, J., Qu, Y. et al. (2017). ACS Catal. 7 (12): 8544–8548. 211 Gao, P., Dang, S., Li, S. et al. (2017). ACS Catal. 8: 571–578. 212 Martin, H.Z., Ivan, M., Tyson, C.W., and Standard Oil Development Co. (1951). Regeneration of an iron catalyst with controlled CO2: CO ratios. US Patent 2,562,804. 213 Schnobel, M. (2009). Reclamation of iron from equilibrium HTFT catalyst. Presentation at International Symposium of Catalyst Deactivation, Delft, The Netherlands. 214 McAdams, D.R., Segura, M.A., and Standard Oil Development Co. (1949). Regeneration of Iron Type Hydrocarbon Synthesis Catalyst. US Patent 2,487,159. 215 Male, P. and Aasberg-Peterson, K. (2021). Fischer-Tropsch synthesis & refining renewable synthetic fuels – challenges and opportunities. European Refinery Technology Conference, Madrid, 16 November 2021. 216 Martinelli, M., Gnanamani, M.K., Demirel, B. et al. (2020). Appl. Catal. A Gen. 608: 117740. 217 Loewert, M. and Pfeifer, P. (2020). ChemEngineering 4 (2): 21. 218 Pfeifer, P., Zimmermann, P., Loewert, M. et al. (2021). Open Research Europe. 219 Anastas, P.T., and Warner, J.C. (1998). Green Chemistry Theory and Practice. New York City: Oxford University Press. 220 Rytter, E., Tsakoumis, N.E., Myrstad, R. et al. (2018). Catal. Today 299: 20–27.

115

116

6  Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis

221 Nkulu, C.B.L., Casas, L., Haufroid, V. et al. (2018). Nat. Sustain. 1 (9): 495–504. 222 Calderone, V.R., Shiju, N.R., Ferré, D.C. et al. (2014). Top. Catal. 57 (17–20): 1419–1424. 223 van Helvoort, T. and Senden, V.M. (2014). Gas to Liquids: Historical Development of GTL Technology in Shell. Shell International. 224 van de Loosdrecht, J., Ciobica, I.M., Gibson, P. et al. (2016). ACS Catal. 6 (6): 3840–3855. 225 Sharapa, D.I., Doronkin, D.E., Studt, F. et al. (2019). Adv. Mater. 31 (26): 1807381. 226 Goldsmith, B.R., Esterhuizen, J., Liu, J.X. et al. (2018). AIChE J. 64 (9): 3553–3553. 227 Sasol EcoFT (2023). FT Synthesis. https://www.sasol.com/our-businesses/sasol-ecoft (accessed 29 June 2023). 228 Sasol (2023). Base oils. https://www.sasol.com/innovation/gas-liquids/products/base-oils (accessed 1 March 2022). 229 Gnanamani, M.K., Jacobs, G., Graham, U.M. et al. (2016). Catal. Today 261: 40–47. 230 Paterson, J., Peacock, M., Purves, R. et al. (2018). ChemCatChem. 10 (22): 5154–5163. 231 Torres Galvis, H.M. and de Jong, K.P. (2013). ACS Catal. 3 (9): 2130–2149. 232 Zhong, L., Yu, F., An, Y. et al. (2016). Nature 538 (7623): 84–87. 233 Botes, G.F., Bromfield, T.C., Coetzer, R.L. et al. (2016). Catal. Today 275: 40–48.

117

7 Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives Maurizio Peruzzini, Fabrizio Mani, and Francesco Barzagli Istituto di Chimica dei Composti Organometallici, Consiglio Nazionale delle Ricerche (ICCOM–CNR), via Madonna del Piano 10, Sesto Fiorentino (Florence), Italy

7.1  Introduction The world population increased from 3 billion to 8 billion between 1950 and early 2022. Correspondingly, global demand for food and energy increased to support the economic growth and the quality of life of populations in rapidly developing countries. About 80% of the energy ­consumed in the world is produced by the combustion of the fossil fuels which accounts for the massive emission into the atmosphere of anthropogenic carbon dioxide (CO2) and can significantly contribute to the acceleration of natural global warming and of the related climate changes [1]. Although the global focus on CO2 is mainly due to its contribution to climate change, it is ­interesting to note how it can represent a commercial input for a range of products and services [2]. Therefore, in recent years the development of technologies specifically addressing the problems related to CO2 capture, utilization, and storage (CCUS) from exhaust gases and, whenever ­possible, directly from the air, has gained increasing interest, with the double aim of reducing CO2 in the atmosphere while making it available for use in industrial processes [3, 4]. Around 250 ­million tons (Mt) of CO2 were used in 2020, and its annual consumption is expected to grow steadily over the coming years. The largest consumer was the fertilizer industry, where over 140 MtCO2 were used for urea production [5]. Urea synthesis is probably the most important industrial process where a potentially harmful substance (CO2 is the most abundant anthropogenic greenhouse gas in the atmosphere) is ­transformed into a widespread, largely used, product. Urea is currently the most widely used nitrogen fertilizer and accounts for about 70% of the total fertilizers used worldwide [6]. More than 200 ­million tons of urea are industrially produced every year, mainly destined for use in the agricultural sector as fertilizer (about 90%), but also in the industrial manufacture of resins, polymers, cosmetics, and, more recently, for the abatement of nitrogen oxides (NOx) in exhaust gases from diesel engines [7–9]. The synthesis of urea has been known since 1828, when it was first obtained by the German chemist Friedrich Wöhler by reacting silver cyanate and ammonium chloride, according to the following reaction [10]:

AgNCO +  NH4Cl → CO(NH2 )2 + AgCl

(7.1)

Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 1, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

118

7  Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives

Wöhler’s discovery was of great impact, because for the first time an organic substance (urea) was artificially synthesized starting from inorganic compounds, without the involvement of living organisms [11]. As a result, the Berzelius’s theory of vitalism was progressively abandoned marking one of the most important conceptual advancements in chemistry during the XIX century. About 40 years later, in 1870, the Russian chemist Aleksandr Bazarov discovered a process to synthesize urea by heating solid ammonium carbamate (NH2CO2NH4), obtained in turn directly from the reaction of CO2 with NH3, at high temperature (130–140 °C) and under pressure [12], as reported in reaction (7.2):

2NH3 +  CO2 → NH2CO2NH4 → CO(NH2 )2 + H2O

(7.2)

The Bazarov reaction has been further investigated for many years [12–14], and even today most of the patented processes for large-scale commercial urea production are substantially based on this reaction [15]. Simplifying the mechanism, urea synthesis mainly involves two consecutive steps: the formation of ammonium carbamate from the direct combination of ammonia and carbon dioxide (Reaction 7.3), followed by its dehydration to form urea and water (Reaction 7.4) [11].

2NH3 +  CO2 → NH2CO2NH4

(7.3)



NH2CO2NH4  CO(NH2 )2 +  H2O

(7.4)

Reaction 7.3 is fast and strongly exothermic (ΔH° = −151 kJ mol–1) and, if the pressure is high enough to force NH3 and CO2 into the liquid phase, the conversion to ammonium carbamate is complete [11, 15]. On the contrary, Reaction 7.4 proceeds slower and it is slightly endothermic (ΔH° = 32 kJ mol–1); it only runs in liquid phase and requires high temperature to afford urea in appreciable yields, while keeping reasonable the reaction rate [11, 15]. As a consequence, all commercial processes for the production of urea starting from gaseous CO2 and NH3 require severe and energy-intensive consuming operating conditions, with high temperatures (170–220 °C), and very high pressures (125–250 bar) [11]. The difference between the diverse processes developed and patented mainly concerns how the unconverted material is recycled, since the reactants cannot be completely converted into urea [11]. In most cases, the synthesis of urea is strictly connected to the ammonia production, where also CO2 is largely available, coming from the H2 production and purification. Indeed, NH3 is synthetized from the Haber-Bosch process (i.e. the catalytic reaction between N2 and H2, in which H2 is commonly obtained by steam reforming of CH4 and N2 from air) [16]. Mutatis mutandis, another option of using CO2 as a raw material for the generation of high value compounds concerns the production of the so-called urea derivatives (or simply ureas), generally indicated as R2N(CO)NR’2. Although they do not have the same impressive annual production of urea, urea derivatives are valuable products too, with a wide range of applications in several areas of industrial production, including pharmaceuticals, agrochemicals and dye chemicals, as well as in different biological and polymer studies. Recent applications of this class of substances also entail their use as solvents and precursors of isocyanates and of raw materials for the synthesis of polyurethanes [17, 18]. Thus, most of those chemicals can be safely considered as high valuable commodities with a relatively large market. As an example, ethylene urea (Figure 7.1), a cyclic urea used in paints and as precursor for agrochemicals, was produced in the order of 12,000 tons in 2018, with a market price of 10 USD/kg [19].

7.2  Catalytic Synthesis of Urea O O

O

HN H 2N

NH2

urea

N H

NH

N H

N,N’-dimethyl urea

ethylene urea

Figure 7.1  Chemical structure of urea and of two widely used urea derivatives, N,N’-dimethyl urea and ethylene urea.

Despite this wide range of applications, the synthesis of urea derivatives is traditionally achieved using the highly toxic reagent phosgene (COCl2) or its analogs, or other unsafe reagents, such as carbon monoxide (CO) [17, 18]. Other methods involved the use of isocyanates, generated in situ by Hofmann rearrangements of amides or by Curtius rearrangements of acyl azides, which, ­however, raise some environmental problems [18]. Alternative syntheses have also been devised, which usually encompass the reactions of amines with urea, ethylene carbonate, or diethyl carbonate, but the high costs of these reagents (obtained from the reaction between CO2 and ammonia, ethylene oxide, or ethanol) make these processes particularly expensive [20]. However, environmental constraints and safety regulations have recently raised further interest to develop safer and cheaper synthetic routes, moving the focus of the research in this field toward the direct synthesis of urea derivatives using the non-toxic and cheap carbon dioxide, in the presence of different catalysts and dehydrating agents in order to increase the reaction yield [17, 19]. In fact, the most important hindrance to the CO2 reactivity is its great thermodynamic stability (∆G° = −394 kJ mol–1) and its kinetic inertness, due to the strong C=O bonds, to the lack of external couples of non-bonded electrons on carbon atoms and to the overall apolarity of the CO2 molecule [21, 22]. Consequently, high temperature and pressure are mandatory to attain CO2 activation and to recover appreciable yields of the products. To reduce those severe and energy intensive operating conditions, some catalyzed processes have been investigated, which represent reliable and potentially useful methodologies to overcome the thermodynamic and kinetic burdens for any reductive transformation of CO2. Indeed, the coordination of CO2 to a metal center weakens the strong C=O double bond, thus lowering the activation energy for the attack of nucleophilic reactants to the CO2 carbon atom and enhancing the overall reaction rate. In this context, the design of efficient catalysts for the synthesis of both urea and urea derivatives would represent a paradigmatic case for the development of innovative industrial processes carried out under milder operating conditions, using safer chemicals while still maintaining high efficiency and productivity, thus allowing the definition of more sustainable synthetic routes from both energy and environmental point of views.

7.2  Catalytic Synthesis of Urea Despite the massive world production of urea, and the urgency to reduce the energy cost of its synthesis, few studies concerning the use of catalysts are reported in the literature, regardless of the starting reagent (gaseous mixture of H2, N2, and CO2, or solid ammonium carbamate). To date, no technology has shown sufficient maturity to become a sound candidate for replacing the Bazarov conventional process.

119

120

7  Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives

7.2.1  Urea from CO2 Reductive Processes Different strategies have been developed and have been experimentally exploited to obtain urea starting from CO2 and NH3, or from the direct gas-phase reaction of CO2, N2, and hydrogen carrying species under milder operating conditions than those used in the conventional industrial process. Because the urea molecule consists of a carbonyl group linked to two amino groups (NH2), the direct activation of CO with ammonia represents an appealing pathway for its synthesis under mild conditions. The strategy devised to achieve this ambitious goal foresees the combination of simultaneously activated CO and ammonia species. Notably, these active species can be obtained with the simultaneous reduction of CO2 and N2 or other simple nitrogen compounds (e.g. nitrate and nitrite ions or even nitrogen oxides as NO) in the presence of hydrogen containing species [23]. 7.2.1.1  Electrocatalysis

Recently, the catalytic conversion of N2 and CO2 to urea through electrochemical reactions carried out at ambient conditions, has emerged as a greener viable alternative strategy to the energy-­ intensive industrial route [24]. The process features a multi-step proton-electron coupling process, as summarized in Reaction (7.5):

N2 +  CO2 + 6 H+ + 6  e− → CO(NH2 )2 +  H2O

(7.5)

Most of the research activity in this exciting field is aimed at improving some of the most critical aspects of the process related to the scarce proclivity of gaseous CO2 and N2 to be strongly adsorbed over the surface of most heterogeneous catalysts and to the high dissociation energy of the thermodynamically high stable double and triple bonds of CO2 and N2, respectively. As a consequence, the electrocatalytic C–N coupling reaction illustrated in Reaction (7.5) generally results in low urea yields and poor selectivity which have so far hampered his general application [25–28]. One of the first example reporting on this approach was published in 1995 by Shibata et al. [29] who synthesized urea by the simultaneous electrochemical reduction of NO3− (or NO2−) and CO2 by using a Cu-loaded gas diffusion electrode. In this work, KNO3 (or KNO2) was added to a KHCO3 solution, used as an electrolyte, and CO2 was continuously provided in the cathode gas-chamber. Urea was produced following the simultaneous electroreduction of both nitrate (or nitrite) and carbon dioxide on the catalyst surface followed by the contemporaneous combination of the formed ammonia and carbon monoxide, to give urea as the final target molecule (Reaction 7.6).

− − 2NO− 2 + CO2 + 9 H2O + 12e → CO(NH2 )2 + 14 OH

(7.6)

The use of nitrate or nitrite as a nitrogen source is particularly interesting for urea electrosynthesis because of its much greater solubility in H2O, compared to the gaseous N2, and therefore its greater reduction potential. Additionally, this approach is particularly appealing because it allows the reuse of a large amount of nitrogen-containing contaminants that would be otherwise discharged into surface waters, threatening the environment and human health [30]. Inspired by this approach, Meng et al. [31] have recently developed an efficient electrocatalyst featuring oxygen vacancy-rich ZnO porous nanosheets for accomplishing the electrosynthesis of aqueous urea by using CO2 and nitrite rich contaminant streams as feedstocks. The mechanism of the reaction was also investigated by a combination of in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) and online differential electrochemical mass spectrometry (DEMS), which allowed the authors to point out that the formation of urea molecules ensues from the coupling of surface adsorbed NH2 and CO2H intermediates.

7.2  Catalytic Synthesis of Urea

In 2020, Chen et al. [27] devised a process for the synthesis of urea based on the direct ­electrochemical coupling under ambient conditions of N2 and CO2 in H2O in the presence of an electrocatalyst consisting of Pd-Cu alloy nanoparticles on TiO2 nanosheets. As a result, the authors achieved a urea formation rate of 3.36 mmol g–1 h–1 (quantified via isotope-labelling experiments) and a corresponding Faradic efficiency of 8.92% at the optimized potential of –0.4 V vs a reversible hydrogen electrode (RHE) in a flow cell. Moreover, the authors proposed a possible mechanism, backed up by robust density functional theory (DFT) arguments, suggesting that this coupling reaction occurred through the formation of C–N bonds via the thermodynamically spontaneous reaction between the surface activated species *N=N* (the asterisks indicate the side-on sorption of N2) and CO, from which urea is easily generated by immediate hydrogenation. High performance for the direct electrochemical coupling of N2 and CO2 was recently (2021) obtained by Yuan et al. [32] by using an innovative strategy involving Mott–Schottky heterostructures as catalysts. These hybrid substances can be obtained by integrating a metal and a semiconductor to form a Mott–Schottky heterojunction. The authors verified that the space-charge region in the heterointerfaces of the catalyst promoted the chemisorption and the subsequent activation of both CO2 and N2 on locally generated electrophilic and nucleophilic regions. Moreover, the local charge redistribution also contributed to fully exposing the active sites and accelerated electrocatalytic kinetics, which improved the formation of the C–N bonds and the generation of the ­chemisorbed *NCON* intermediate urea precursor. As a best result, by using heterostructured Bi-BiVO4 hybrids as electrocatalysts, a maximum urea yield rate of 5.91 mmol g–1 h–1 and Faradaic efficiency of 12.55% in 0.1 M KHCO3 at −0.4 V vs. RHE, under ambient conditions, was achieved. Soon thereafter, using a similar approach, the same research group described another efficient urea electrochemical synthesis by using perovskite hetero-structured hybrids [28]. In detail, BiFeO3/BiVO4 hybrids gave a maximum urea yield rate of 4.94 mmol g–1 h–1 with a Faradaic efficiency of 17.18% in 0.1 M KHCO3 at −0.4 V vs RHE. As a further improvement, Yuan et al. [24] in 2021 demonstrated that nickel borate [Ni3(BO3)2] nanocrystals enriched with frustrated Lewis pairs (FLPs) were capable of accomplishing the simultaneous electrocatalytic reduction of CO2 and N2 to urea under ambient conditions. In particular, the prepared Ni3(BO3)2-150 nanocrystal, where the surface hydroxyl and neighboring Ni site serve as a Lewis base and acid, respectively, showed the best performance, with a maximum urea yield rate of 9.70 mmol g–1 h–1 and a Faradaic efficiency of 20.36% at −0.5 V vs RHE. It was demonstrated that the Lewis basic and acidic sites in FLPs acted synergistically, favoring the capture adsorption of both CO2 and N2 by attractive orbital interaction and promoting the activation of the reactant molecules by the unique “donation–acceptance” process between gas molecules and catalytic active sites. Afterward, a smooth downhill process takes place, easily combining the intermediate activated molecules by bringing about the electrocatalytic C–N coupling reaction from which *NCON* urea precursors were eventually generated. The same research group further exploited the frustrated Lewis pairs strategy by developing indium-based [InO(OH)] nanocrystals for the ambient urea electrosynthesis: as a best result, InO(OH)-100 performed a maximum urea yield rate of 6.85 mmol g–1 h–1 and a Faradaic efficiency of 20.97% at −0.4 V vs RHE [33]. Aiming at improving the emerging electrochemical urea synthesis pathway with the development of low-cost and efficient electrocatalysts, in 2021 Zhu et al. [8] investigated the direct coupling of N2 and CO2 to produce urea using some specific two-dimensional (2D) transition metal borides (MBenes) as electrocatalysts, under ambient conditions. Corroborated by supportive DFT calculations, the authors demonstrated that three different MBenes electrocatalysts, namely Mo2B2, Ti2B2, and Cr2B2, are able to adsorb and activate N2 and CO2 on their basal planes; then the key intermediate *NCON can be formed via the coupling of *N2 and *CO, which can be further reduced

121

122

7  Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives

to urea via four proton-coupled electron transfer steps. The limiting potentials of urea formation for our three studied MBenes are in the range of −0.49 to −0.65 eV, similar to that of the Pd–Cu alloy catalyst [27]. 7.2.1.2  Photocatalysis

The utilization of semiconductor particles as photocatalysts for the photoreduction of CO2 to useful chemicals represents an alternative and potentially very useful method for bringing about the catalytic synthesis of urea. For this purpose, various studies have been conducted in the last 20 years using photocatalysts based on titania (TiO2) dispersed or immobilized in solutions containing CO2 and nitrogen compounds, inside quartz reactors, transparent to the wavelength TiO2 absorbing region between 300 and 380 nm. An example of the apparatus used for this technique is reported in Figure 7.2b. As early as in 1998, Kuwabata et al. [34] successfully synthesized urea from the simultaneous photoreduction of CO2 in the presence of several nitrogen compounds (NO3–, NO, NH2OH): the photoinduced reaction experiments were carried out at room temperature and atmospheric pressure in a quartz reactor, containing TiO2 nanocrystals immobilized into a polyvinylpyrrolidinone gel film as photocatalyst; a 500-W high-pressure mercury arc lamp was used as a proper light source. In 2005, Ustinovich et al. [35] reported a study aimed at improving the catalytic performance of TiO2 in the simultaneous photoreduction of CO2 and NO3− to urea. They used emulsified perfluorocarbons, and in particular perfluorodecalin (PFD), capable of dissolving large quantities of gaseous substrates, to prepare a titania-stabilized PFD-in-water (PFD:TiO2) emulsion. Using this specific catalyst, the authors found that the urea photoproduction rate was incremented by two to seven times (depending on the nitrate concentration, with a maximum urea photoproduction rate of about 0.55 mm h–1 with 1.5 M of NaNO3) compared to the more conventional TiO2 suspension, mostly due to the higher active concentration of CO2 involved into light-induced reaction at the interface. In the same year, Shchukin and Mohwald [36] reported the innovative photosynthesis of urea from aqueous inorganic precursors (CO2 and NO3–) realized inside polyelectrolyte capsules, by using photocatalysts based on nanosized TiO2. Polyelectrolyte capsules act as a photocatalytic microreactors, and this spatial confinement entails several advantages, such as low overconcentration and overheating of the reaction area, controlled access of the reagents to the reaction area, and protection of the activity of the encapsulated material (catalyst) against oxidation or poisoning.

Figure 7.2  (a) Schematic view of urea formation over TiO2 supported on zeolite photo catalyst under ultraviolet illumination. Reproduced with permission from Ref [37] / John Wiley & Sons. (b) Schematic of the experimental device for photocatalytic co-reduction of N2/CO2 to urea. Reproduced with permission from Ref [38] / American Chemical Society.

7.2  Catalytic Synthesis of Urea

According to their study, the efficiency of urea synthesis from the photoreduction of CO2 and NO3− grows with the decrease of the capsule size. They achieved a maximum 37% yield of urea after eight hours of irradiation with Cu-modified TiO2 nanoparticles encapsulated inside 2.2 µm poly(styrene sulfonate)/poly(allylamine hydrochloride) capsules. In 2012, Srinivas et al. [37] described the photocatalytic synthesis of urea at atmospheric pressure and ambient temperature using as photocatalysts TiO2 or iron titanates (i.e. Fe2TiO5 or Fe2Ti2O7), supported at different catalyst loading over the HZSM-5 zeolite. The simultaneous photoreduction of aqueous nitrate ion in the presence of 2-propanol and oxalic acid (used as source of CO2) in a batch reactor under both solar and UV light, resulted in urea formation. A schematic representation of this intriguing photocatalytic pathway is reported in Figure 7.2a. As a finding, the yield of urea increased when the catalysts were supported on zeolite, with the maximum yield of urea about 19 ppm obtained when the reaction was conducted with 10 wt% Fe-titanate/HZSM-5 as photocatalyst (3 g L–1), a KNO3 solution (1.6*10-3 M) as a source of N2, isopropanol (1%) as a source of CO2, and a high-pressure Hg vapor as a source of light for the six hours of the experiment. More recently, Maimaiti et al. [38] reported on the use of Ti3+-TiO2/Fe-CNTs as innovative photocatalyst for the synthesis of urea from co-reduction of gaseous CO2 and N2. The active catalyst was prepared from crystalline brookite (the orthorhombic modification of titania) as Ti(III) self-doped TiO2 (Ti3+-TiO2) loaded on carbon nanotubes with Fe cores (Fe-CNTs). A simplified sketch of the experimental device used is reported in Figure 7.2b. As a finding, the urea yield obtained using this nanocomposite photocatalyst was almost 5 times higher than that exhibited by Ti3+-doped brookite alone under the same operating conditions, reaching a maximum urea yield of μmol L–1 g–1 in a 4 h experiment, thus confirming that the Fe-CNT support is actively improving the photocatalysis. The authors also verified the maintenance of the catalytic activity, carrying out successive cycles of catalysis. From the investigation of the reaction mechanism, the authors found that the performance of photocatalytic co-reduction of CO2 and N2 to urea in water depends on the arrangement of Ti3+ sites and oxygen vacancies on the surface of Ti3+-TiO2 loaded on Fe-CNT support. From a mechanistic viewpoint it was speculated that the oriented adsorption and activation of N2 and CO2 molecules may be driven by the alternation of Ti3+ sites and oxygen vacancies, which resulted in the formation of a sixmembered ring intermediate, from where the reductive formation of two urea molecules could eventually occur. 7.2.1.3  Magneto-catalysis

A curious and innovative low energy catalytic pathway for the synthesis of green urea was proposed in 2017 by Yahya et al. [39], who investigated the heterogeneous catalytic co-reduction of N2 and CO2 (in presence of H2) under a magnetic field, by using synthesized hematite (α-Fe2O3) nanowires (NWs) as a catalyst. In their experiments, the synthesized NWs were placed in a reactor between the poles of DC (direct current) electromagnet, and a gaseous mixture of H2, N2 and CO2 (variable total flow rate, but gas ratio fixed to 3:1:1) was flowed inside. The reaction was carried out at ambient pressure and room temperature, and under magnetic field ranging from 0.0 to 2.5 T. The authors studied the dependence of the urea yield from the applied magnetic field, gas flows and time. Density functional modelling of the process corroborated the hypothesis that the absorption of the reagent gas mixture on the hematite surface could enhance the net spin density (and hence net magnetic moment) of the magnetic material, thus allowing the catalysis to take place in the presence of an adequate magnetic field (magnetic induction). A claim that this green process could represent a concrete and less-energy intensive alternative to the Haber-Bosch process for the synthesis of artificial fertilizers was advanced by the authors.

123

124

7  Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives

7.2.2  Urea from Ammonium Carbamate As an alternative strategy to circumvent the high temperatures and pressures affecting the conventional industrial production of urea, the possibility of promoting the dehydration of ammonium carbamate (Reaction 7.4) by using a catalyst under milder operating condition was also investigated. Quite surprisingly, very few reports can be found in literature concerning this technique. The most intriguing aspect of this approach is that carbamate can be obtained directly from CO2 capture processes using both aqueous and non-aqueous solutions of ammonia as a sorbent (Reaction 7.3). By this way, the capture of CO2 could be directly combined with its use, through processes with a lower energy impact than conventional technology. Barzagli et al. [15] reported for the first time in 2011 an extensive screening on the use of some simple transition metal salts as catalysts for the catalytic dehydration of ammonium carbamate. Ammonium carbamate was easily obtained at room temperature from the capture of 15% CO2 in N2 (that simulates an exhaust combustion gas) in ethanol-water ammonia solution. In particular, salts of chromium, cobalt, copper, iron, manganese, nickel, ruthenium, and zinc were considered. In a typical experiment, ammonium carbamate was heated in a sealed reactor at temperatures in the range 120–140 °C, without applying any external pressure, in the presence of 1–1.5% (on molar scale) of the catalyst for two to three days. The yield of the conversion of carbamate into urea was easily evaluated by using 13C NMR spectroscopy. As a result, Cu(II) and Zn(II) salts were found as the most efficient catalysts, with urea yields in the range 45–54%. Under the same operating conditions, Ru(II) and Ni(II) produced urea with a lower yield (26–33%), while Mn(II), Fe(III), Cr(III), and Co(II) salts gave always conversions lower than 25%. The catalytic activity of the added metal salts was confirmed by some blank experiments, without any catalyst, carried out under the same operating conditions that gave less than 3% carbamate conversion. The proposed mechanism is reported in Figure 7.3. The metal ion (M2+) can coordinate the NH3 released from the thermal decomposition of the carbamate, to form the corresponding hexakis-amino complex [M(NH3)6]2+. The availability of two adjacent molecules of ammonia could favor the reaction with CO2 and the formation of one urea molecule (with the concomitant release of a water) as a dihapto-coordinated ligand in the transient [M{k2-N,N-(NH2)2CO}(NH3)4]2+ complex (structure II in Figure 7.3). Elimination of urea by excess of ammonia restores the hexa-amino complex. By considering the peculiar coordinating properties of the different metal ions and the stability constants of the related ammonia complexes, the best catalytic performance of Cu(II) and Zn(II) ions could be rationalized and traced back to the presence of four strongly and two labile coordinated ammonia ligands. In contrast, the ligand displacement step is more difficult to be accomplished for other transition metals due to the greater stability of their six-coordinated ammonia complexes. Inspired by the potential catalytic activity of Cu(II) for the synthesis of urea from ammonium carbamate, in 2021 Hanson et al. [40] reported the catalytic behavior of tetraammineaquocopper(II) sulfate complex, [Cu(NH3)4(OH2)]SO4 showing interesting conversion yields up to 18±6% at 120 °C after 15 hours in a high-pressure metal reactor (catalyst loading = 1%) in the presence of water. The authors investigated the reaction mechanism by DFT methods, and found that the [Cu(NH3)4]2+ complex was probably the effective catalyst. The proposed pathway for the reaction between [Cu(NH3)4]2+ and carbamic acid entailed the formation of an adduct intermediate, where an activated NH3 molecule moved to the carbon atom to form urea with simultaneous elimination of water (Figure 7.4). Finally, the catalytic cycle closed by regeneration of the [Cu(NH3)4]2+ catalyst by coordination of a free NH3 molecule originating from the solution ammonium carbamate that equilibrated with carbamic acid and NH3.

7.2  Catalytic Synthesis of Urea

Figure 7.3  Proposed pathway of the M2+ catalyzed reaction between CO2 and two coordinated ammonia molecules (M = Cu, Zn). Urea displacement from the coordination sphere by the stronger ligand NH3 may regenerate easily the active catalytic species. Reproduced with permission from Ref [15] / Royal Society of Chemistry.

Figure 7.4  Proposed pathway for the reaction between [Cu(NH3)4]2+ and carbamic acid. Reproduced with permission from Ref [40] / American Chemical Society.

125

126

7  Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives

The potential of some metal ions to be used as catalysts for the dehydration of carbamates was also investigated by Manaka [41] in 2021. In this work, 14 different salts of Ni, Fe, Sn, Cu, Mn, Bi, and Nd featuring diverse Lewis acidity were tested in DMSO for the conversion of ammonium carbamate at 140 °C for six hours in a sealed reactor (catalyst loading = 10%). The study showed a poor conversion to urea for most of the screened catalysts, and the greater catalytic activity was found for nickel acetate and nickel hydroxide, with yields up to 6%. Corroborated by further in-depth investigation using x-ray absorption fine structure (XAFS) and Fourier-transform extended XAFS (FT-EXAFS) spectroscopies, the author proposed a putative reaction pathway for the urea synthesis catalyzed by nickel hydroxide: at the beginning of reaction, the addition of ammonium carbamate to the Ni(OH)2 catalyst breaks Ni-O-Ni bonds and promotes the coordination of ammonia and carbamate with the catalyst; once the carbonyl group of the carbamate is coordinated, it is activated by Ni, which acts as a Lewis acid, and becomes susceptible to be nucleophilically attacked by the ammonium ion, to finally form urea. In 2020, Manaka et al. also investigated the conversion of ammonium carbamate into urea promoted by some selected organic bases as catalysts and different organic solvents, in mild condition [42]. Remarkably, a conversion of 35% was achieved by using 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) as catalyst, after 72 hours at 100 °C in a sealed reactor and in presence of DMSO (catalyst loading = 10%). Based on the NMR and FT-IR results, the authors proposed the mechanism shown in Figure 7.5. The catalytic cycle begins with a cation exchange between the ammonium carbamate and DBU, which weakens the carbamate C=O bond (and increases its electrophilicity) and releases an ammonia molecule. The free ammonia molecule may subsequently attack the carbonyl group of the carbamate anion to form urea and water, while the DBU catalyst was simultaneously regenerated. Other ammonium salts, like ammonium carbonate and ammonium bicarbonate, were demonstrated to be also active in catalyzing the urea formation.

Figure 7.5  Proposed reaction pathway for the catalytic synthesis of urea using DBU as a catalyst. Reproduced with permission from Ref [42] / Springer Nature.

7.3  Catalytic Synthesis of Urea Derivatives

Although the studies reported so far have demonstrated the feasibility of the production of urea from the catalytic decomposition of carbamates and have allowed to foresee good potentialities for a further improvement of the process, especially from an energy point of view, at present the hardest obstacle to the commercial application of this catalytic process is the catalyst recovering and recycling. Actually, whatever the metal salts are used, at the end of the reaction, both the metal salts and urea are dissolved in the dehydration water, and hence it is laborious to separate the catalyst for its reuse in a continuous cycle. As a further drawback, the urea produced by this catalytic process could be in same cases significantly contaminated by metal ions that may severely hamper its commercialization for end use unless expensive and complicated purification steps were implemented in the process.

7.3  Catalytic Synthesis of Urea Derivatives Most of the recent research activity for the development of safe and cheap processes for the production of urea derivatives focuses on the reaction of CO2 with amines in the presence of either organic or inorganic catalysts [43]. The reaction at room temperature and low CO2 pressure between the weak acid CO2 and primary or secondary amines (i.e. amines with at least one hydrogen atom on the amine functionality, R1R2NH) to yield the corresponding amine carbamate (R1R2NCO2− + R1R2NH2+, i.e. the carbamate anion and the amine protonated cation), as reported in Reaction (7.7), is among the most well established and studied organic reactions. The subsequent dehydration of the amine carbamate eventually generates the urea derivative, (R1R2N)2CO, as in Reaction (7.8) [44].

1 2 + 2R1R 2NH + CO2  R1R 2NCO− 2 + R R NH2

(7.7)



1 2 + 1 2 R1R 2NCO− 2 +  R R NH2  (R R N )2 CO + H2O

(7.8)

Even if the Reaction (7.6) is exothermic, the reactivity rate of a number of primary and secondary amines may be low at room temperature and pressure. More important, the efficient dehydration Reaction (7.8) is endothermic and slow, and therefore requires high temperature and catalytic conditions. Hence, the effective conversion of CO2 and amines into urea derivatives is still challenging. Actually, the use of transition metal compounds (oxides, salts, or metal complexes) to carry out the catalytic synthesis of substituted ureas starting from CO2 and primary or secondary amines, has been known for some decades: the improvements obtained over the years concern the switch from expensive catalysts with low conversion yield, to cheaper and more stable catalysts with higher production yields and asking for milder operating conditions. Moreover, it should be noted that the catalyst can affect both the carbonylation of the amine and the dehydration of the carbamate to urea. The palladium-catalyzed carbonylation of amines has attracted considerable attention as an efficient pathway to obtain N,N’-disubstituted urea derivatives. A pioneering study was reported in 1986 by Morimoto et al. [45] who tested several transition metal catalysts (Ru, Pd, Rh) eventually showing that Pd-compounds, in the presence of PPh3 in CCl4/CH3CN co-solvent, under mild operating conditions, gave the best catalytic efficiency toward the reaction between CO2 and dialkylamines. As a best result, the conversion of diethylamine into tetraethylurea was achieved with a yield of 36%, after three days of reaction at room temperature, by using PdCl2(CH3CN)2 as a

127

128

7  Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives

catalyst. As for the reaction mechanism, the authors hypothesized the addition of HNEt2 to palladium and the formation of a [HPd{N(CH2CH3)2}] intermediate. However, most of the palladium-catalyzed studies developed in the following years to obtain higher conversion yields were performed at high temperature and high pressure using CO and an organic solvent as medium [46, 47]. A significant improvement was described in 2011 by Della Ca’ et al. [48], who reported an efficient synthesis of symmetrical urea derivatives via oxidative carbonylation of primary amines catalyzed by the stable palladium complex K2PdI4 in the presence of liquid carbon dioxide as reaction medium. In a standard experiment, the amine and the catalyst were loaded in a sealed stainless-steel reactor, together with liquid CO2 (as a medium) and a gaseous mixture of CO and air (25–27 bar). The reactor was stirred and heated at temperatures in the range 70–110 °C for 10–72 hours. The reaction did not generate any other by-product but water and theconversion was greater than 80% for most of the tested aliphatic and aromatic amines. Although CO was still used as a reagent, the authors argued that the use of CO2 as a reaction medium made the process greener and safer, since no organic solvents was needed and the excess of carbon dioxide prevented from any risk of explosion or combustion hazards. Ruthenium-based compounds have also been extensively studied for the production of urea derivatives. As earlier as in 1983, Zoeckler and Laine synthetized the symmetric urea derivative diphenylurea, (PhNH)2CO, by reacting CO2 (at 14 bar) with a silazane, PhNHSi(CH3)3, using as catalyst the ruthenium carbonyl Ru3(CO)12 in tetrahydrofuran (THF). The reaction was performed in a sealed reactor heated at 110 °C for 10 hours, with a yield of 48% (based on PhNHSi(CH3)3) [49]. A pioneering study on ruthenium-catalyzed synthesis of urea derivatives was reported a few years later by Dixneuf et al., located in Rennes [50, 51]. In their work, the French researchers found that symmetrical ureas could be obtained with good yields by the reaction of CO2 with primary amines, in the presence of a terminal alkyne and a ruthenium complex, as shown in Scheme 7.1. HC 2 RNH2 + CO2

C R' [Ru]

RHN

NHR C O

[R'C2 H + H2O]

Scheme 7.1  Symmetrical ureas synthesis from CO2 and primary amines, in presence of a terminal alkyne and a ruthenium-based catalyst.

The reactions were carried out in a sealed reactor heated at 140 °C for 20 hours, and with a CO2 pressure of 50 bar. In these operating conditions, by using cyclohexylamine as the reagent, RuCl3·3H2O/ PBu3 as the catalyst and 2-methylbut-3-yn-2-o1 as the terminal-alkyne compound, Dixneuf and Bruneau reported the formation of N,N’-dicyclohexylurea in 61% yield. Under the same conditions, the same reaction was performed with similar yields (in the range 56–61%) by using (p-cymene)RuCl2(PPh3) and (hexamethylbenzene)RuCl2(PMe3) as catalysts. Lower conversion yields were reported by varying temperature or diluents. In addition to the ruthenium catalyst, the authors underlined the crucial role of the presence of a terminal alkyne, and in particular of its stoichiometric amount, corresponding to one mole of alkyne per mole of urea formed. In the proposed mechanism, the ruthenium catalyst reacted with the substituted alkynes to form intermediate π–alkyne (or vinylidene) ruthenium species from which the disubstituted urea can be assembled via a series of steps traversing the formation of intermediate vinyl carbamate species. Nucleophilic attack of a second equivalent of amine to such an intermediate and Ru-catalyzed dehydration of the transiently assembled ruthenium enol afforded the disubstituted urea and restored the active Ru-catalyst.

7.3  Catalytic Synthesis of Urea Derivatives

The catalytic effect of ruthenium compounds in the carbonylation of amines was intensively studied also in the following decade: as a matter of fact, a variety of pincer ruthenium complexes, such as [(PNPPh)Ru(H)Cl(CO)] (PNPPh = bis[2-(diphenylphosphino)ethyl]amine), were used to obtain other urea derivatives (also asymmetrical) with high yields starting from amines and ­carbonyl precursors such as methanol or N,N-dimethylformamide [52, 53]. With a similar approach, fully exploiting the knowledge acquired with ruthenium complexes, processes have been developed by Bernskoetter et al. based on the use of similar pincer iron complexes, such as [(RPNP)Fe(H) (CO)] (RPNP = N{CH2CH2(PR2)2}, R = iPr), that are excellent catalysts for the catalytic dehydrogenative coupling of primary amines and methanol [54]. In 2005, Shi et al. [55] reported on the unprecedented CO2 activation by polymer-immobilized nanogold catalysts in the synthesis of disubstituted ureas. In the specific case, cyclohexylamine and benzylamine were reacted with CO2 (at 50 bar) using 0.05 wt% of the Au/Poly1 catalyst in an autoclave, heated at 180 °C, for 20 hours. As a result, the disubstituted ureas (dicyclohexylurea and dibenzylurea) were obtained in high yields (83–85%) and with excellent turn over frequencies for product (TOFP) of approximatively 3,000 mol/mol/h (mol product per mol gold per hour). However, due to the relatively weak chemical bond between Au nanoparticles and the polymer support, and the extreme reaction conditions required for the heterogeneous reaction, the efficiency of the catalyst decreased over time [56]. The direct synthesis of cyclic urea derivatives from CO2 and diamines was accomplished by using cerium oxide CeO2 as an effective and reusable heterogeneous catalyst [57, 58]. High conversion yields (78–98%), together with good product selectivity, were obtained by treating a 1,2-diamine, or a 1,3-diamine, with CO2 (even at a low CO2 pressure of 5 bar) in a sealed reactor, heated at 160 °C for 12 hours, in the presence of a catalytic amount of CeO2 and 2-propanol as organic solvent. The proposed reaction mechanism, showed in Figure 7.6 for the reaction between

Figure 7.6  Proposed reaction pathway for the synthesis of 2-imidazolidinone (ethylene urea) from CO2 and ethylenediamine, by using CeO2 as the catalyst. Reproduced with permission from Ref [57] / Royal Society of Chemistry.

129

130

7  Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives

ethylenediamine and CO2, entails a stepwise mechanism which involves: (1) the adsorption of both diamine and CO2 over the CeO2 surface to form the carbamate moiety; the (2) decomposition of the carbamic acid into amine and CO2; the (3) nucleophilic attack of the amino group to the carbamate moiety, which affords the cyclic urea and represents the rate-determining step of the catalytic cycle; and, finally, (4) the desorption of the cyclic urea, in this particular case ethylene urea (product 1 in the figure), and the simultaneous regeneration of the CeO2 catalyst [57]. Similar conversion yields for the synthesis of cyclic ureas from diamines were recently reported by More and Srivastava [59] by using synthetic MOF-derived CeO2 compounds as catalysts. In particular, the reaction of CO2 and several diamines (namely ethylenediamine, 1,2-diaminopropane, 1,2-diamino2-methylpropane, and 1,3-diaminopropane) in a sealed reactor heated at 160 °C for 12 hours in the presence of the MOF catalyst M-CeO2-573, led to the formation of the corresponding cyclic urea ­derivatives with high conversion yields (87–95%) and very high product selectivity (95–99%). The weak Lewis acidic sites present on the catalyst surface helped the adsorption of amines, whereas the significant number of surface oxygen vacancies (due to the catalyst morphology) and basic sites of the catalyst favored the activation of CO2. Moreover, the ­MOF-based catalyst was found to be stable over time and largely reusable after several recycles under the reported operating conditions. The use of different metal oxalates as catalysts for the synthesis of N,N′-dialkylureas was reported in 2016 by Sun et al. [60]. In a typical experiment, several aliphatic primary amines were allowed to react with CO2 (20 bar) in the presence of metal oxalates (metal = Ce, Mn, Na, Ni, Y, Zr). N-­methyl-2-pyrrolidinone (NMP) was used as solvent and 4A zeolite as dehydrating reagent. Y2(C2O4)3 provided the best catalytic performance for the carbonylation of amines and for the subsequent formation of the N,N′-dialkylurea. The yttrium oxalate (as well as yttrium borate, carbonate, citrate, or oxide) outperformed all the other tested metal oxalates as potential catalysts for the preparation of N,N′-dialkylureas. The particular catalytic efficiency of Y2(C2O4)3 was traced back to the large ionic radius of the Y3+ cation and to the especially well-suited structure of the coordination complex, which should allow a positive interaction with the substrate molecules. Heating for 24 hours at 150 °C, aliphatic primary amines were converted to the corresponding N,N’-dialkylureas with high yields in the range 71–86%. Noticeably, the catalytic process did not work with both secondary and aromatic amines. In a separate study, another yttrium-compound, the mixed oxide yttria-stabilized zirconia Y0.08Zr0.92O1.96 (YSZ-8), was investigated by the same research group [56]. This mixed oxide ceramic material was shown to be very active and highly selective toward the carbonylation of aliphatic primary amines with CO2, forming the corresponding N,N′-dialkylureas with high conversion yields (up to 80.6%) after 48 hours of heating at T = 160 °C, employing NMP as solvent and 4A ­zeolite as dehydrant. The presence of oxygen vacancies in YSZ-8 was considered crucial for the carbonylation process by creating specifically localized additional reduction potential for the activation of CO2 which henceforth easily promoted the formation of the adsorbed key intermediate species. In 2017, Stephan et al. [61] explored the use of a variety of indium compounds in the catalytic transformations of CO2 and aliphatic or aromatic silylamines into urea derivatives. In particular, it was found that indium amido compounds In{N(SiMe3)2}2Cl·THF and In{N(SiMe3)2}Cl2·(THF)n were able to effectively catalyze the reaction thus allowing the preparation of a wide range of aryl and alkyl symmetric ureas. The reactions were performed at 110 °C, with 3 bar of CO2, in 12–48 hours, with 0.05–5 mol% catalyst loading. The catalytic conversion ranged from 70 to 99%, depending on the silylamine used as substrate. The behavior of some commercially available oxovanadium(V) compounds, in particular of VO(OiPr)3, for the catalytic amination of CO2 with primary amines under ambient pressure to attain gram-scale of different urea derivatives (Figure 7.7), was first demonstrated in 2021 by Moriuchi et al. [62].

7.3  Catalytic Synthesis of Urea Derivatives

Figure 7.7  Oxovanadium(V)-catalyzed amination of carbon dioxide for the synthesis of ureas. Reproduced with permission from Ref [62] / Royal Society of Chemistry.

Typically, the catalytic runs were performed at 130 °C by using several primary amines in the presence of VO(OiPr)3 (8 mol%) as catalyst and N,N-dimethylacetamide (DMA) as solvent under carbon dioxide, for approximatively 15 hours. As a result, most of the investigated amines were converted into the corresponding symmetric disubstituted ureas with high yields, ranging from 60 to 89%. The quantification of the reaction yields was carried out by using 13C NMR spectroscopy which was confirmed to be a powerful and efficient analytic tool [63]. Notably, the authors verified that chiral amines were also transformed into the corresponding chiral urea derivatives without loss of chirality. In the same article the reaction of CO2 with 2-phenylethylamine catalyzed by other transition metal oxides, including Ti, W, Nb, Fe, and Mo, was also investigated, but poorer results were obtained in comparison with most of the tested oxovanadium(V) compounds. In 2016, Barzagli et al. [17] devised a strategy for the integrated process of CO2 capture from a flue gas followed by its catalyzed transformation into urea derivatives. The CO2 capture was efficiently performed at 20 °C in an absorption column containing a solution of a primary amine (3 mol dm-3) in diethyleneglycoldiethyl ether or diethylketone. The flue gas (CO2 15% v/v) was fed from the bottom of the column where CO2 was allowed to react with the amines to form the solid amine carbamates (Reaction 7.7) that were easily separated from the solution. The subsequent conversion of the amine carbamate salt into the corresponding 1,3-disubstituted urea was carried out in a sealed stainless-steel reactor heated at 150 °C for five hours, in the presence of 1% (molar scale) of CuCl or CuCl2·2H2O as catalysts. The reaction mixture and the products obtained were analyzed by 13C NMR spectroscopy. Using the n-butylamine, iso-butylamine and n-octylamine as starting amines, the corresponding 1,3-disubstituted ureas were obtained with conversions in the range 37–44% and very high selectivity. To confirm the catalytic role of copper salts for the dehydration reaction (Reaction 7.8), the same experiments were repeated without catalysts but, ­irrespectively of the used amine, the observed conversion was never higher than 30% even after prolonged heating (16 hours). A similar method was developed in 2021 by Choi et al. for the synthesis of urea derivatives starting from alkyl ammonium carbamates using a titanium(IV) complex as catalyst [19]. Alkyl ammonium carbamates were easily obtained by capturing CO2 from a flue gas or even from air with different primary aliphatic amines (and diamines) in an appropriate solvent, at room temperature. The subsequent conversion of the carbamate to the corresponding urea derivative was ­performed in a stirred autoclave heated at 170 °C for 24 hours, in the presence of the titanium complex Cp2Ti(OTf)2 (Cp = cyclopentadienyl; OTf = trifluoromethanesulfonate) as homogeneous catalyst (2 mol%) and an organic solvent. High carbamates conversion yields (> 80%) were obtained for several amines, when using 1,3-dimethyl-2-imidazolidinone as a solvent. As reported by the authors, this catalytic method can be used for the synthesis of different urea derivatives and has proved particularly suitable for producing cyclic urea derivatives, such as ethylene urea (obtained with yield of 80%, starting from ethylenediamine), which has a high commercial value.

131

132

7  Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives

As we have previously summarized, a variety of catalytic systems based on different transition metals have been demonstrated capable of producing ureas by bringing about the direct combination of amines and carbon dioxide with high efficiency and selectivity. However, it is worth noticing that alternative metal-free and environmentally benign processes have also been proposed and studied. In this respect, Kong et al. [64] developed a simple process in 2010 for the synthesis of symmetrical ureas from amines and CO2 by using polyethylene glycol supported potassium hydroxide (KOH/PEG1000) as an inexpensive and recyclable catalyst, that did not need of any additional dehydrating agent. In the optimized operating condition (heating in a stainless-steel autoclave at 150 °C for 10 hours with 8 MPa of CO2 and 10 mol% of KOH/PEG1000 with respect to the substrate), several primary and secondary aliphatic amines, as well as several diamines, were converted into the corresponding urea derivatives in moderate to high yields (32.9–82.0%, depending on the amine used). Aniline and other aromatic amines were not converted with this procedure, probably due to their lower basicity. The crucial role played by polyethylene glycol (PEG) should be underlined. First, the authors reported that PEG1000 could form complexes through the coordination of the potassium cation (similarly to crown ethers), which resulted in an increase in the basicity of KOH, which in turn could also facilitate the formation of the ammonium carbamate salt during the synthesis of urea derivatives. Additionally, PEG can act as a physical dehydrating agent since it is highly hygroscopic. Finally, KOH/PEG1000 can be easily recovered and reused several times, maintaining a high catalytic activity. With a similar approach, Yang et al. [65] developed an innovative process for the CO2 capture and conversion into urea derivatives. Specifically, they appointed a catalytic binary system consisting of polyethylene glycol and an amidine or guanidine superbase, capable of capturing CO2 as liquid amidinium carbonate salt, which was subsequently reacted with amines to form symmetrical ureas. The catalytic reaction is reported in Figure 7.8 [44]. In their experiment, the absorption of CO2 with the catalyst was performed at 40 °C, whereas the subsequent conversion to ureas was carried out by heating the amidinium carbonate salt at 110 °C for 24 hours in a sealed reactor. With this protocol, by using the DBU/PEG150 catalyst (DBU = diazabicyclo[5.4.0]-undec-7-ene) several primary aliphatic amines, secondary aliphatic amines and diamines were converted into the corresponding urea derivatives in very high yields (up to 99%). As a further development, recently some bicyclic guanidines, featuring a rigid framework and unique electronic and chemical properties, were proved to be effective organocatalysts for the chemical fixation of CO2 into linear and cyclic urea derivatives [43]. In the study reported by Marchegiani et al. [66], several amines were converted into linear and cyclic ureas in moderate to high yields (31–92%, depending on the amine used) after after 24 hours of reaction at 100 °C with CO2 (70 bar) and by using the bicyclic guanidine 1,5,7-triazabicyclo[4.4.0]dec-5-ene (TBD) as the

Figure 7.8  Carbon dioxide capture/activation by the 1,8-diazabicyclo[5,4,0]undec-7ene/polyethylene glycol (DBU/PEG) system and subsequent conversion of n-butylamine into dibutyl urea. Reproduced with permission from Ref [44] / Springer Nature.

7.4  Conclusions and Future Perspectives

catalyst.In particular, the authors stressed the relevance of a new and general bicyclic guanidinecatalyzed synthesis of 5-methyleneimidazolidin-2-ones (ethylene urea derivatives) by reaction of propargylamines, primary amines and CO2 in relative mild condition. Finally, the potential use of ionic liquids (IL) to catalyze the conversion of amines into the corresponding symmetrical urea derivatives deserves to be mentioned as ILs have some characteristics suitable for this process, such as high CO2 absorption, recyclability, and dehydrating properties [67]. In 2003, Shi et al. [68] experimented with the reaction of a series of aliphatic amines with CO2 in the presence of several CsOH/IL catalytic systems, obtaining the corresponding urea derivatives with moderate to high yields (53–98%), after four to six hours of reaction (temperature was not reported). As a best result, the yield of 98% was obtained for the production of ­N,N’-dicyclohexylurea, by using cyclohexylamine as substrate and 1-butyl-3-methylimidazolium ­chloride (BMImCl) ionic liquid in the presence of CsOH as catalyst. Because the utilization of the strong base CsOH may cause many drawbacks, such as corrosion, deactivation, and even uncontrolled degradation of the IL under high temperature, Li et al. [69] developed a similar base-free IL-based system in 2010, identifying Co(acac)3/BMMImCl (acac = acetylacetonate) as the best performing catalyst. In their experiments, carried out at 160 °C for 10 hours with the this catalyst, the reaction of several aliphatic and aromatic amines with CO2 (5 MPa) produced the corresponding symmetrical urea derivatives with high selectivity and with yields in the range of 19–81% depending on the starting amine. The stability of the Co(acac)3/BMMImCl catalyst was confirmed during the reaction by FT-IR spectroscopy monitoring, and by its utilization in subsequent catalytic cycles.

7.4  Conclusions and Future Perspectives Climate change mostly due to the increasing concentration of greenhouse gases in the atmosphere is one of the most urgent global challenges. Among the greenhouse gases, the emissions of CO2 derived from anthropogenic activities are considered the main cause of global warming. To reduce the hurdle of high energy costs of the CO2 capture from large scale industrial plants (such as fossil fuelled power plants) or directly from atmospheric air, the use of anthropogenic CO2 as a source of carbon for the synthesis of commercial compounds and fine chemicals is now attaining a growing attention and is raising a general consensus from both scientists and industrial specialists as well as from the civic society and decision makers. In considering different approaches to turn anthropogenic CO2 into a feedstock for the synthesis of commodity chemicals, the urea synthesis is of paramount importance. Urea is indeed the most used nitrogen fertilizer worldwide and its production should be necessarily increased over time to produce more cereals (the engine of the human civilization) and to grow the intensive farming of biomass for the production of biofuels (120 × 106 metric tons in 2020), in order to replace, at least in part, oil. As a matter of fact, the current cost of oil over $100USD/barrel (April 2022) makes the costs of biofuels already competitive with oil. Moreover, the non-toxic and inexpensive CO2 could replace whenever possible unsafe reactants such as carbon monoxide (CO) and phosgene (COCl2) in the manufacture of useful chemicals and polymers, such as urea derivatives and isocyanates or polyurethanes and polycarbonate plastics, respectively. To overcome the great thermodynamic stability and the kinetic inertness of CO2, a variety of catalytic processes have been devised aimed at reducing the activation energy for the attack of different nucleophiles to the carbon atom of the CO2 molecule, which is the key point of catalytic CO2 reduction chemistry. In this respect, two main synthetic routes have been devised for achieving the urea synthesis which may be classified as single or two-step process, respectively. The two-step process comprises

133

134

7  Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives

the synthesis of ammonium carbamate from CO2 and NH3, that is thermodynamically and kinetically favored, followed by the endothermic and slow carbamate dehydration in the presence of a variety of catalysts. In the single step process, which generally gives lower yields, the reduction of CO2 and nitrogen containing species (N2, NO2–, NO3–) is carried out altogether in the presence of hydrogen or other reducing reagents by means of electrochemical, magnetic, or photo catalytic systems. Urea derivatives can be generally obtained by the one-step catalyzed conversion of amines and CO2 at high temperature and pressure, with very good yield and the selectivity. The catalytic synthesis of ureas represents a milder and greener route to the synthesis of such valuable intermediates with respect to the conventional syntheses that often use unsafe chemicals. A few examples of two-step synthesis through the intermediate amine carbamates were also reported. Although great advancements have been made in recent years to achieve a reliable catalytic process to produce urea and its derivatives, much work has yet to be done for transferring the most promising lab-scale experiments so far reported to commercial large-scale production plants. To provide potential advantages in term of energy saving with respect to the traditional commercial processes, it will be mandatory to replace expensive catalysts with cheaper, efficient, and stable ones, and to reduce as much as possible the temperature and pressure of the process (i.e. the energy costs), while keeping the conversion reaction still working under continuous feeding conditions.

References 1 IPCC (2014). Climate change 2014: synthesis report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Geneva, Switzerland. 2 Samanta, S. and Srivastava, R. (2020). Mater. Adv. 1: 1506–1545. https://doi.org/10.1039/ D0MA00293C. 3 Huppmann, D., Rogelj, J., Kriegler, E. et al. (2018). Nat. Clim. Chang. 8: 1027–1030. https://doi. org/10.1038/s41558-018-0317-4. 4 Barzagli, F., Giorgi, C., Mani, F., and Peruzzini, M. (2017). J. CO2 Util. 22: 346–354. https://doi. org/10.1016/j.jcou.2017.10.016. 5 International Energy Agency (IEA) (2019). Putting CO2 to use: creating value from emissions. www.iea.org/publications/reports/PuttingCO2touse (accessed 10 May 2022). 6 Driver, J.G., Owen, R.E., Makanyire, T. et al. (2019). Front. Energy Res. 7: 1–15. https://doi.org/ 10.3389/fenrg.2019.00088. 7 GlobalData (2020). Global urea capacity and capital expenditure outlook, 2020-2030. https://store. globaldata.com/report/gdch0063mar–global-urea-capacity-and-capital-expenditureoutlook-2020-2030. 8 Zhu, X., Zhou, X., Jing, Y., and Li, Y. (2021). Nat. Commun. 12: 1–9. doi.org/10.1038/ s41467-021-24400-5. 9 Glibert, P.M., Harrison, J., Heil, C., and Seitzinger, S. (2006). Biogeochemistry 77: 441–463. https:// doi.org/10.1007/s10533-005-3070-5. 10 Wöhler, F. (1828). Ann. Phys. 87: 253–256. doi.org/10.1002/andp.18280870206. 11 Meessen, J. (2014). Chemie Ing. Tech. 86: 2180–2189. https://doi.org/10.1002/cite.201400064. 12 Tsipis, C.A. and Karipidis, P.A. (2005). J. Phys. Chem. A 109: 8560–8567. https://doi.org/10.1021/ jp051334j. 13 Bagnell, L.J., Hodges, A.M., Linton, M., and Mau, A.W.H. (1989). Aust. J. Chem. 42: 1819–1829. https://doi.org/10.1071/CH9891819.

References

14 Krase, N.W. and Gaddy, V.L. (1922). Ind. Eng. Chem. 14: 611–615. https://doi.org/10.1021/ ie50151a009. 15 Barzagli, F., Mani, F., and Peruzzini, M. (2011). Green Chem. 13: 1267. https://doi.org/10.1039/ c0gc00674b. 16 Pérez-Fortes, M., Bocin-Dumitriu, A., and Tzimas, E. (2014). Energy Procedia 63: 7968–7975. https://doi.org/10.1016/j.egypro.2014.11.834. 17 Barzagli, F., Mani, F., and Peruzzini, M. (2016). J. CO2 Util. 13: 81–89. https://doi.org/10.1016/j. jcou.2015.12.006. 18 Xu, M., Jupp, A.R., Ong, M.S.E. et al. (2019). Angew. Chemie Int. Ed. 58: 5707–5711. https://doi.org/ 10.1002/anie.201900058. 19 Koizumi, H., Takeuchi, K., Matsumoto, K. et al. (2021). Commun. Chem. 4: 1–7. https://doi. org/10.1038/s42004-021-00505-2. 20 Wu, C., Cheng, H., Liu, R. et al. (2010). Green Chem. 12: 1811–1816. https://doi.org/10.1039/ c0gc00059k. 21 Aresta, M., Dibenedetto, A., and Angelini, A. (2013). J. CO2 Util. 3–4: 65–73. https://doi. org/10.1016/j.jcou.2013.08.001. 22 Nocito, F. and Dibenedetto, A. (2020). Curr. Opin. Green Sustain. Chem. 21: 34–43. https://doi. org/10.1016/j.cogsc.2019.10.002. 23 Kim, J.E., Choi, S., Balamurugan, M. et al. (2020). Trends Chem. 2: 1004–1019. https://doi. org/10.1016/j.trechm.2020.09.003. 24 Yuan, M., Chen, J., Xu, Y. et al. (2021). Energy Environ. Sci. https://doi.org/10.1039/D1EE02485J. 25 Yang, H.B., Hung, S.-F., Liu, S. et al. (2018). Nat. Energy 3: 140–147. https://doi.org/10.1038/ s41560-017-0078-8. 26 Suryanto, B.H.R., Du, H.-L., Wang, D. et al. (2019). Nat. Catal. 2: 290–296. https://doi.org/10.1038/ s41929-019-0252-4. 27 Chen, C., Zhu, X., Wen, X. et al. (2020). Nat. Chem. 12: 717–724. https://doi.org/10.1038/ s41557-020-0481-9. 28 Yuan, M., Chen, J., Bai, Y. et al. (2021). Chem. Sci. 12: 6048–6058. https://doi.org/10.1039/ D1SC01467F. 29 Shibata, M., Yoshida, K., Furuya, N., and Electroanal, J. (1995). Chem. 387: 143–145. https://doi.org/10.1016/0022-0728(95)03992-P. 30 Chen, G.-F., Yuan, Y., Jiang, H. et al. (2020). Nat. Energy 5: 605–613. https://doi.org/10.1038/ s41560-020-0654-1. 31 Meng, N., Huang, Y., Liu, Y. et al. (2021). Cell Reports Phys. Sci. 2: 100378. https://doi.org/10.1016/j. xcrp.2021.100378. 32 Yuan, M., Chen, J., Bai, Y. et al. (2021). Angew. Chemie - Int. Ed. 60: 10910–10918. https://doi.org/ 10.1002/anie.202101275. 33 Yuan, M., Zhang, H., Xu, Y. et al. (2022). Chem Catal. 2: 309–320. https://doi.org/10.1016/j. checat.2021.11.009. 34 Kuwabata, S., Yamauchi, H., and Yoneyama, H. (1998). Langmuir 14: 1899–1904. https://doi.org/ 10.1021/la970478p. 35 Ustinovich, E.A., Shchukin, D.G., and Sviridov, D.V. (2005). J. Photochem. Photobiol. A Chem. 175: 249–252. https://doi.org/10.1016/j.jphotochem.2005.04.037. 36 Shchukin, D.G. and Möhwald, H. (2005). Langmuir 21: 5582–5587. https://doi.org/10.1021/ la050429+. 37 Srinivas, B., Kumari, V.D., Sadanandam, G. et al. (2012). Photochem. Photobiol. 88: 233–241. https://doi.org/10.1111/j.1751-1097.2011.01037.x.

135

136

7  Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives

38 Maimaiti, H., Xu, B., Sun, J., and Feng, L. (2021). ACS Sustain. Chem. Eng. 9: 6991–7002. https://doi.org/10.1021/acssuschemeng.1c00644. 39 Yahya, N., Qureshi, S., Ur Rehman, Z. et al. (2017). J. Magn. Magn. Mater. 428: 469–480. https://doi.org/10.1016/j.jmmm.2016.12.005. 40 Hanson, D.S., Wang, Y., Zhou, X. et al. (2021). Inorg. Chem. 60: 5573–5589. https://doi.org/10.1021/ acs.inorgchem.0c03467. 41 Manaka, Y. (2021). J. Japan Pet. Inst. 64 172–177. https://doi.org/10.1627/jpi.64.172. 42 Manaka, Y., Nagatsuka, Y., and Motokura, K. (2020). Sci. Rep. 10: 1–8. https://doi.org/10.1038/ s41598-020-59795-6. 43 Li, J.-Y., Song, Q.-W., Zhang, K., and Liu, P. (2019). Molecules 24: 182. https://doi.org/10.3390/ molecules24010182. 44 Wang, H., Xin, Z., and Li, Y. (2017). Top. Curr. Chem. 375: 49. https://doi.org/10.1007/ s41061-017-0137-4. 45 Morimoto, Y., Fujiwara, Y., Taniguchi, H. et al. (1986). Tetrahedron Lett. 27: 1809–1810. https://doi.org/10.1016/S0040-4039(00)84381-6. 46 Orito, K., Miyazawa, M., Nakamura, T. et al. (2006). J. Org. Chem. 71: 5951–5958. https://doi.org/ 10.1021/jo060612n. 47 Guan, Z.H., Lei, H., Chen, M. et al. (2012). Adv. Synth. Catal. 354: 489–496. https://doi.org/ 10.1002/adsc.201100545. 48 Della Ca’, N., Bottarelli, P., Dibenedetto, A. et al. (2011). J. Catal. 282: 120–127. https://doi.org/ 10.1016/j.jcat.2011.06.003. 49 Zoeckler, M.T. and Laine, R.M. (1983). J. Org. Chem. 48: 2539–2543. https://doi.org/10.1021/ jo00163a023. 50 Fournier, J., Bruneau, C., Dixneuf, P.H., and Lecolier, S. (1991). J. Org. Chem. 56: 4456–4458. https://doi.org/10.1021/jo00014a024. 51 Bruneau, C. and Dixneuf, P.H. (1992). J. Mol. Catal. 74: 97–107. https://doi.org/ 10.1016/0304-5102(92)80227-8. 52 Kim, S.H. and Hong, S.H. (2016). Org. Lett. 18: 212–215. https://doi.org/10.1021/acs. orglett.5b03328. 53 Krishnakumar, V., Chatterjee, B., and Gunanathan, C. (2017). Inorg. Chem. 56: 7278–7284. https://doi.org/10.1021/acs.inorgchem.7b00962. 54 Lane, E.M., Hazari, N., and Bernskoetter, W.H. (2018). Chem. Sci. 9: 4003–4008. https://doi. org/10.1039/c8sc00775f. 55 Shi, F., Zhang, Q., Ma, Y. et al. (2005). Chem. Soc. 127: 4182–4183. https://doi.org/10.1021/ja042207o. 56 Sun, D., Xie, K., Fang, Y., and Yang, X. (2018). Catalysts 8: 188. https://doi.org/10.3390/ catal8050188. 57 Tamura, M., Noro, K., Honda, M. et al. (2013). Green Chem. 15: 1567. https://doi.org/10.1039/ c3gc40495a. 58 Primo, A., Aguado, E., and Garcia, H. (2013). ChemCatChem 5: 1020–1023. https://doi.org/ 10.1002/cctc.201200329. 59 More, G.S. and Srivastava, R. (2021). Ind. Eng. Chem. Res. 60: 12492–12504. https://doi.org/ 10.1021/acs.iecr.1c01759. 60 Sun, D.-L., Ye, J.-H., Fang, Y.-X., and Chao, Z.-S. (2016). Ind. Eng. Chem. Res. 55: 64–70. https://doi.org/10.1021/acs.iecr.5b02936. 61 Xu, M., Jupp, A.R., and Stephan, D.W. (2017). Angew. Chemie - Int. Ed. 56: 14277–14281. https://doi.org/10.1002/anie.201708921.

References

62 Moriuchi, T., Sakuramoto, T., Matsutani, T. et al. (2021). RSC Adv. 11: 27121–27125. https://doi.org/ 10.1039/D1RA04125H. 63 Hu, X.E., Yu, Q., Barzagli, F. et al. (2020). ACS Sustain. Chem. Eng. 8: 6173–6193. https://doi.org/ 10.1021/acssuschemeng.9b07823. 64 Kong, D.-L., He, L.-N., and Wang, J.-Q. (2010). Synlett 2010: 1276–1280. https://doi.org/ 10.1055/s-0029-1219799. 65 Yang, -Z.-Z., He, L.-N., Zhao, Y.-N. et al. (2011). Energy Environ. Sci. 4: 3971. https://doi.org/ 10.1039/c1ee02156g. 66 Marchegiani, M., Nodari, M., Tansini, F. et al. (2017). J. CO2 Util. 21: 553–561. https://doi.org/ 10.1016/j.jcou.2017.08.017. 67 Xia, S.-M., Chen, K.-H., Fu, H.-C., and He, L.-N. (2018). Front. Chem. 6: 1–7. https://doi.org/ 10.3389/fchem.2018.00462. 68 Shi, F., Deng, Y., SiMa, T. et al. (2003). Angew. Chemie Int. Ed. 42: 3257–3260. https://doi.org/ 10.1002/anie.200351098. 69 Li, J., Guo, X., Wang, L. et al. (2010). Sci. China Chem. 53: 1534–1540. https://doi.org/10.1007/ s11426-010-4026-8.

137

139

Part II Transformation of Volatile Organic Compounds (VOCs)

141

8 Catalysis Abatement of NOx/VOCs Assisted by Ozone Zhihua Wang1 and Fawei Lin2 1 2

State Key Laboratory of Clean Energy Utilization, Zhejiang University, Hangzhou P.R. China School of Environmental Science and Engineering, Tianjin University, Tianjin P.R. China

8.1  NOx/VOC Emission and Treatment Technologies Ozone pollution in the summer and haze in winter have become typical air pollution in many countries. Therefore, simultaneous control of ozone and particle matter (PM) has become the critical target for air pollution control. NOx and VOCs are critical precursors of ozone and PM under atmospheric photochemical environment. NOx includes N2O, NO, NO2, and N2O5, whereas VOCs are mixture of serious organic pollutants with boiling points ranging 50~260  oC at ambient condition. Finding ways to eliminate NOx and VOCs effectively is the most important research topic at present. Taking these factors into account, this chapter introduces newly developed emission abatement technologies for NOx and VOCs.

8.1.1  NOx/VOC Emissions NOx primarily originates from the combustion of fossil fuels via routes of thermal NOx, fuel NOx, and prompt NOx through power plants, automobiles, chemical production processes, and so on. Many governments, including those of the USA, Europe, Japan, and China have legislated strict emission standards for NOx. At present, the most strict emission standard in the world for coalfired power plants is NOx95.0%

>95.0%

T:dichloroethane>chlorobenzene [Figure 8.7c] [57]. 8.3.3.3  Dioxins and Furans

Polychlorinated dibenzo-dioxins and polychlorinated dibenzo-furans (PCDD/Fs), often known as dioxins and furans, are persistent and toxic byproducts of combustion and industrial production that cause considerable environmental and human health harm. Generally, they are not VOCs due to low volatility. Nevertheless, they are the most concerning Cl containing organic pollutants in flue gas and therefore this chapter also discusses recent studies investigating the catalytic ozonation of dioxins. Ozone assistance can promote the destruction of PCDD/Fs. Lu et al. reported that the destruction efficiency of PCDD/Fs achieved 45% over V–Mn/Ti–CNTs at 150 oC without ozone, while this value improved to 91% with 50 ppm ozone input [75]. Efficient catalytic ozonation of PCDD/Fs exhibits apparent activation energy, thus always requiring much higher temperature (c. 150~220  oC) [72]. Higher temperature could accelerate ozone decomposition to generate more highly active oxygen species and consequently improve catalytic performance. However, too rapid decomposition of O3 into O2 could suppress ozone utilization, thus weakening catalytic efficiency. Therefore, reaction temperature is a critical factor for catalytic ozonation that should be optimized with different organic pollutants and reaction conditions.

References

8.3.4  Sulfur-containing VOCs Sulfur-containing VOCs are odorous emissions that cause unpleasant odors at very low concentrations and are produced primarily by animal production plants, wastewater treatment, and pulp and paper. Methyl mercaptan (CH3SH) and dimethyl sulfide (CH3SCH3) are two sulfur-containing VOCs that have high electron density carbon-sulfur linkages and are quickly oxidized by electrophilic molecules like ozone. Xia et al. [84] investigated the catalytic ozonation of methyl mercaptan by Ag/MnO2. The elimination rate of 70 ppm CH3SH reached 95 percent after Ag was deposited on three-dimensional MnO2 porous hollow microspheres (PHMSs), which was much greater than that of pure MnO2 PHMSs (79%) and ozone breakdown alone (28%). Because Mn has a low oxidation state, adding Ag raises the lattice oxygen content, allowing more oxygen to be absorbed and further excited into reactive oxygen species. The catalytic ozonation of dimethyl sulfide (DMS) by V2O5/TiO2 was explored by Sahle Demessie and Devulapelli [85]. The oxidation of DMS can be aided by V2O5/TiO2, with a full conversion rate of 99 percent, a molar yield of CO2 and SO2 of 80–90 percent, and DMSO2 as the major oxidation product.

8.4  Conclusions The use of low temperature ozone oxidation techniques has the potential to allow the simultaneous removal of multiple pollutants, typically including SO2, NOx, and VOCs, if combined with intrinsic desulfurization. These initial emissions, NO and VOCs, will be oxidized into compounds such as NO2, N2O5, CO2, and H2O, with different ozone injection and reaction conditions. However, N2O5 formation and VOC oxidation exhibit high activation energy and homogeneous oxidation by ozone cannot attain desirable performance within a short time and at a low temperature. To facilitate oxidation efficiency, catalytic ozonation is conducted for NO deep oxidation into N2O5 and VOC oxidation. This chapter describes the principles of NO homogenous oxidation by ozone, heterogeneous catalytic ozonation of NO into N2O5, and catalytic ozonation of multiple VOCs. Generally, catalytic ozonation should be conducted at 90~130 oC for NO oxidation into N2O5, which is suitable for low temperature flue gas treatment. The optimal temperature for the catalytic ozonation of VOCs varies among different VOC molecules, while most VOCs can be oxidized from ambient temperature to 120  oC. The conversion efficiency is also affected by ozone input, which depends on the initial concentration of NO and VOCs. Catalytic decomposition of O3 and oxidation of NO, NO2, and VOC molecules can occur simultaneously. The active oxygen species and variations on catalyst surface derived from ozone contribute to further oxidation. Therefore, catalyst design should consider the activity for ozone decomposition. Catalytic ozonation exhibits obvious advantages in low temperature, high efficiency, high stability, and thoroughness without many byproducts.

References 1 Zhang, Z.M., Lin, F.W., Xiang, L. et al. (2022). Chem. Eng. J. 427. 2 Lin, F.W., Zhang, Z.M., Li, N. et al. (2021). Chem. Eng. J. 404. 3 Asghar, U., Rafiq, S., Anwar, A. et al. (2021). J. Environ. Chem. Eng. 9 (5). 4 Javed, M.T., Irfan, N., and Gibbs, B.M. (2007). J. Environ. Manage. 83 (3): 251–289. 5 Busca, G., Lietti, L., Ramis, G., and Berti, F. (1998). Appl. Catal. B-Environ. 18 (1–2): 1–36. 6 Forzatti, P. (2001). Appl. Catal. A-General 222 (1–2): 221–236.

157

158

8  Catalysis Abatement of NOx/VOCs Assisted by Ozone

7 Yang, C.T., Miao, G., Pi, Y.H. et al. (2019). Chem. Eng. J. 370: 1128–1153. 8 Liao, C.M., Chen, J.W., Chen, J.S., and Liang, H.M. (2001). Appl. Math. Model. 25 (12): 1071–1087. 9 Mok, Y., Shin, D., Koh, D., and Kim, K. (2005). Appl. Chem. 9 (1): 217–220. 10 Mok, Y.S. (2004). J. Chem. Eng. Jpn. 37 (11): 1337–1344. 11 Lin, F., Wang, Z., Ma, Q. et al. (2016). Energ Fuel 30 (6): 5101–5107. 12 Sun, C., Zhao, N., Zhuang, Z. et al. (2014). J. Hazard. Mater. 274: 376–383. 13 Ji, R., Wang, J., Xu, W. et al. (2018). Ind. Eng. Chem. Res. 57 (43): 14440–14447. 14 Skalska, K., Miller, J.S., and Ledakowicz, S. (2011). Chem. Eng. Sci. 66 (14): 3386–3391. 15 Stamate, E., Chen, W., Jørgensen, L. et al. (2010). Fuel 89 (5): 978–985. 16 Skalska, K., Miller, J., and Ledakowicz, S. (2011) Chem. Papers 65 (2): 193–197. 17 Skalska, K., Miller, J.S., and Ledakowicz, S. (2012). Chem. Eng. Process. Process Intensification 61: 69–74. 18 Asif, M. and Kim, W.-S. (2014). Ozone Sci. Eng. 36 (5): 472–484. 19 Yoon, H.J., Park, H.-W., and Park, D.-W. (2016). Energ. Fuel 30 (4): 3289–3297. 20 Li, B., Zhao, J., and Lu, J. (2015). Int. J. Environ. Res. Public Health 12 (2): 1595–1611. 21 Janssen, C., Simone, D., and Guinet, M. (2011). Rev. Sci. Instrum. 82 (3): 034102. 22 Lin, F., Wang, Z., Zhang, Z. et al. (2020). Chem. Eng. J. 382: 123030. 23 Stamate, E., Irimiea, C., and Salewski, M. (2013). Jpn. J. Appl. Phys. 52 (5S2). 24 Jakubiak, M. and Kordylewski, W. (2011). Chem. Process Eng. 32 (3): 229–239. 25 Wang, Z.H., Li, B., Ehn, A. et al. (2010). Fuel 89 (9): 2346–2352. 26 Jõgi, I., Levoll, E., and Raud, J. (2016). Catal. Lett. 147 (2): 566–571. 27 Cui, S., Hao, R., and Fu, D. (2019). Fuel 246: 365–374. 28 Jõgi, I., Erme, K., Levoll, E. et al. (2018). Plasma Sources Sci. Technol. 27 (3): 035001. 29 Lin, F., Wang, Z., Ma, Q. et al. (2016). Appl. Catal. B 198: 100–111. 30 Lin, F., Wang, Z., Shao, J. et al. (2017). Chin. J. Catal. 38: 1270–1280. 31 Jõgi, I., Erme, K., Raud, J., and Laan, M. (2016). Fuel 173: 45–51. 32 Erme, K., Raud, J., and Jogi, I. (2018). Langmuir 34 (22): 6338–6345. 33 Lin, F., Wang, Z., Shao, J. et al. (2017). RSC Adv. 7 (40): 25132–25143. 34 Ding, J., Zhong, Q., and Zhang, S. (2014). Fuel Process Technol. 128: 449–455. 35 Guo, L., Zhong, Q., Ding, J. et al. (2016). Ozone Sci. Eng. 38 (5): 382–394. 36 Guo, L., Zhong, Q., Ding, J. et al. (2016). RSC Adv. 6 (91): 87869–87877. 37 Zhao, W., Zhang, S., Ding, J. et al. (2016). J. Mol. Catal. A: Chem. 424: 153–161. 38 Ding, J., Zhong, Q., Zhang, S., and New, A. (2015). Ind. Eng. Chem. Res. 54 (7): 2012–2022. 39 Ding, J., Lin, J., Xiao, J. et al. (2016). J. Alloys. Compd. 665: 411–417. 40 Ding, J., Zhong, Q., Cai, H., and Zhang, S. (2016). Chem. Eng. J. 286: 549–559. 41 Han, C., Zhang, S., Guo, L. et al. (2018). Chem. Eng. Res. Des. 136: 219–229. 42 Liu, B., Xu, X., Liu, L. et al. (2019). Ind. Eng. Chem. Res. 58 (4): 1525–1534. 43 Liu, Y.X. and Zhang, J. (2011). Ind. Eng. Chem. Res. 50 (7): 3836–3841. 44 Ma, J., Wang, C., and He, H. (2017). Appl. Catal. B 201: 503–510. 45 Lian, Z.H., Ma, J.Z., and He, H. (2015). Catal. Commun. 59: 156–160. 46 Zhang, Z., Jiang, Z., and Shangguan, W. (2016). Catal. Today 264: 270–278. 47 Yan, Y., Yang, C., Peng, L. et al. (2016). Atmos. Environ. 143: 261–269. 48 Einaga, H. and Ogata, A. (2009). J. Hazard. Mater. 164 (2–3): 1236–1241. 49 Einaga, H., Teraoka, Y., and Ogat, A. (2011). Catal. Today 164 (1): 571–574. 50 Jin, M., Kim, J.W., Kim, J.M. et al. (2011). Powder Technol. 214 (3): 458–462. 51 Park, J.-H., Jurng, J., Bae, G.-N. et al. (2012). J. Nanosci. Nanotechnol. 12 (7): 5942–5946. 52 Jin, M., Kim, J.H., Kim, J.M. et al. (2013). Catal. Today 204: 108–113.

References

53 Park, J.H., Kim, J.M., Jurng, J. et al. (2013). J. Nanosci. Nanotechnol. 13 (1): 423–426. 54 Huang, H., Huang, W., Xu, Y. et al. (2015). Catal. Today 258: 627–633. 55 Einaga, H., Maeda, N., Yamamoto, S., and Teraoka, Y. (2015). Catal. Today 245: 22–27. 56 Fang, R., Huang, W., Huang, H. et al. (2019). Appl. Surf. Sci. 470: 439–447. 57 Chen, G.Y., Wang, Z., Lin, F.W. et al. (2020). J. Hazard. Mater. 391. 58 Li, M., Hui, K.N., Hui, K.S. et al. (2011). Appl. Catal. B 107 (3–4): 245–252. 59 Rezaei, E. and Soltan, J. (2012). Chem. Eng. J. 198-199: 482–490. 60 Rezaei, E., Soltan, J., and Chen, N. (2013). Appl. Catal. B 239-247: 239–247. 61 Hu, M., Hui, K.S., and Hui, K.N. (2014). Chem. Eng. J. 254: 237–244. 62 Li, J., Na, H., Zeng, X. et al. (2014). Appl. Surf. Sci. 311: 690–696. 63 Rezaei, F., Moussavi, G., Bakhtiari, A.R., and Yamini, Y. (2016). J. Hazard. Mater. 306: 348–358. 64 Xiao, H., Wu, J., Wang, X. et al. (2018). Mol. Catal. 460: 7–15. 65 Shao, J.M., Lin, F.W., Wang, Z.H. et al. (2020). Appl. Catal. B-Environ. 266. 66 Wang, H.C., Liang, H.S., and Chang, M.B. (2011). J. Hazard. Mater. 186 (2–3): 1781–1787. 67 Liang, H.S., Wang, H.C., and Chang, M.B. (2011). Ind. Eng. Chem. Res. 50 (23): 13322–13329. 68 Jin, D., Ren, Z., Ma, Z. et al. (2015). RSC Adv. 5 (20): 15103–15109. 69 Zhang, Z., Xiang, L., Lin, F. et al. (2021). Chem. Eng. J. 426. 70 Lin, F., Zhang, Z., Xiang, L. et al. (2022). Chem. Eng. J. 435: 134807. 71 Zhao, R., Wang, Q., and Lu, S. (2015). Chem. Lett. 44 (12): 1676–1678. 72 Wang, H.C., Chang, S.H., Hung, P.C. et al. (2009). J. Hazard. Mater. 164 (2–3): 1452–1459. 73 Ji, S.S., Li, X.D., Ren, Y. et al. (2013). Chemosphere 92 (3): 265–272. 74 Yu, M.F., Lin, X.Q., Yan, M. et al. (2016). Environ. Sci. Pollut. Res. Int. 23 (17): 17563–17570. 75 Wang, Q., Tang, M., Peng, Y. et al. (2018). Chemosphere 199: 502–509. 76 Zhan, M.X., Yu, M.F., Zhang, G. et al. (2018). Waste Manag. 76: 555–565. 77 Yuan, M.-H., Chang, -C.-C., Chang, C.-Y. et al. (2015). J. Taiwan Inst. Chem. Eng. 53: 52–57. 78 Einaga, H. and Futamura, S. (2005). Appl. Catal. B 60 (1–2): 49–55. 79 Zhao, D.Z., Shi, C., Li, X.S. et al. (2012). J. Hazard. Mater. 239-240: 362–369. 80 Wang, H.C., Huang, Z.W., Jiang, Z. et al. (2018). ACS Catal. 8 (4): 3164–3180. 81 Xi, Y., Reed, C., Lee, Y.-K., and Ted Oyama, S. (2005). J. Phys. Chem. B 109: 17587–17596. 82 Ghavami, M., Aghbolaghy, M., Soltan, J., and Chen, N. (2020). Front. Chem. Sci. Eng. 14 (6): 937–947. 83 Tian, M.Z., Liu, S.J., Wang, L.L. et al. (2020). Environ. Sci. Technol. 54 (3): 1938–1945. 84 Xia, D.H., Xu, W.J., Wang, Y.C. et al. (2018). Environ. Sci. Technol. 52 (22): 13399–13409. 85 Sahle-Demessie, E. and Devulapelli, V.G. (2008). Appl. Catal. B-Environ. 84 (3–4): 408–419. 86 Hicks, J., Vasko, P., Goicoechea, J.M., and Aldridge, S. (2019). J. Am. Chem. Soc. 141 (28): 11000–11003. 87 Wen, Z.C., Wang, Z.H., Zhou, J.H. et al. (2009). Ozone-Sci. Eng. 31 (5): 393–401. 88 Goldan, P.D., Kuster, W.C., Williams, E. et al. (2004). J. Geophys. Res.-Atmosph. 109: D21309. 89 Li, K., Ji, J., He, M., and Huang, H.B. (2020). Catal. Sci. Technol. 10 (18): 6257–6265. 90 Yildiz, Y., Kuzu, S., Sen, B. et al. (2017). Int. J. Hydrogen. Energy 42 (18): 13061–13069. 91 Zhang, Z., Xiang, L., Lin, F. et al. (2021). Chem. Eng. J. 426: 130814.

159

161

9 Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions Elisabete C.B.A. Alegria1,*, Manas Sutradhar2, and Tannistha R. Barman3 1 Departamento de Engenharia Química, ISEL, Instituto Politécnico de Lisboa, Portugal, and Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal 2 Faculdade de Engenharia, Universidade Lusófona - Centro Universitário de Lisboa, Campo Grande 376, Lisboa, 1749–024, Portugal, and Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal 3 Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal * Corresponding author

9.1  Introduction Volatile organic compounds (VOCs) are a class of chemicals that can evaporate at room temperature and are currently one of the main environmental concerns as they are major contributors to atmospheric pollution [1–3]. VOCs may be divided into families of pollutants, namely the aldehydes (e.g. formaldehyde and acrolein), the chlorinated compounds (e.g. trichloroethene and vinyl chloride), and a series of low-weight aromatic hydrocarbons well-known as BTEX that are among the most hazardous VOCs to human health and the environment [4, 5]. VOCs are mainly used as solvents, frequently encountered as industrial contaminants, and are emitted from a large variety of sources that include vehicle emissions, printing, the leather industry, underground gasoline storage tanks, landfills and waste treatment plants, and domestic heating [6]. At the industrial level, prior policy is the total or partial substitution of VOCs by alternative sources, or their treatment in specialized equipment before being released into the atmosphere [1]. In the last case, the combustion of such compounds occurs in specific thermal oxidation equipment, forming CO2 and H2O as the final products and thus reducing environmental contamination [7–9]. However, although efficient, this method is very expensive in terms of energy and is not suitable for low concentrations of VOCs. Therefore, there is a need for the development of innovative and cost-­effective techniques for VOC abatement. In recent years, several technologies have been applied to eliminate VOCs [10], with special emphasis on flame combustion, adsorption [11, 12], catalytic oxidation [13], biological degradation, non-thermal plasma (NTP) oxidation, and photocatalytic decomposition, among others. Lately, innovative approaches have been proposed that combine two VOC treatment techniques to increase the efficiency of the process. These include adsorption-catalysis and adsorption-­ photocatalysis coupling technologies, and the combination of heterogeneous catalysis with a NTP. In the combined adsorption-catalysis process [14, 15], the VOC molecules in low concentration are adsorbed onto an adsorbent and the oxidation process occurs in the desorption step by the action of a catalyst or, even more innovatively, hybrid systems with the same material playing the dual Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 1, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

162

9  Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions

role of adsorbent and catalyst (at higher temperatures) in successive steps [16–18]. The adsorptionphotocatalysis coupling technique consists in the adsorption of VOC molecules to the photocatalyst surface being oxidized by the photo-generated holes. The environmentally friendly conditions, high efficiency, and low energy consumption make this method very attractive [19–21]. Finally, NTP technology is based on a treatment performed at room temperature and atmospheric pressure and has therefore received increasing attention in the removal of VOCs from industrial and/or indoor air. Nevertheless, the formation of unwanted products and low efficiency have hampered the expansion of this technique [22, 23]. The combination of heterogeneous catalysis with a NTP has thus emerged as an attempt to increase the potentiality of this technology. Partial and selective catalytic oxidation has been considered a highly promising pathway due to the high conversion of VOCs to oxidized products. However, it remains a challenge to develop catalysts and synthetic pathways that can promote catalytic oxidation under more sustainable conditions, with high efficiency and selectivity [2, 13]. Among the possible oxygenated products are benzyl alcohol, benzaldehyde, acetophenone, tolualdehyde, toluic acid, terephthalic acid, benzoic acid, and others, which are important intermediates for many chemicals, agrochemicals, fragrances, pharmaceuticals, and polymers. This chapter presents the complete or partial oxidation of several VOCs, namely BTEX, under homogeneous and heterogeneous conditions, and various VOC removal strategies [24].

9.2  Benzene Benzene is an important chemical feedstock and a constituent of gasoline. However, due to its harmful effect on health, its elimination from various gas streams is mandatory, including from coke ovens, distillation towers, petrochemical plants, and others. The complete oxidation of benzene in air to CO2 was first reported by Anderson et al. at temperatures above 100 ºC using a 0.1 wt% Pt-loaded TiO2 photocatalyst [25]. Recently, Liu et al. [26] prepared Sn-modified ɑ-MnO2 composites with different tin contents, to improve the catalytic performance in benzene oxidation. The conversion was tested with pure octahedral-type cryptomelane-molecular sieve tunnels (ɑ-MnO2) and the Sn-modified ɑ-MnO2 composite, and the best conversion was achieved with a moderate Sn content, reaching 90 and 100% conversions at 200 and 250 ◦C, respectively, under a gas hourly space velocity (GHSV) of 120 L g–1 h–1, with CO2 as the core product and maleic acid as the main intermediate. Easily accessible acidic sites, an abundant supply of reactive oxygen, and weakened Mn-O bonds were critical for the diffusion of benzene molecules and surface oxygen transport [26]. The presence of a metallic additive was equally important in the work presented by Lee et al. [27], in which the addition of Cu improved the catalytic activity of a series of mesoporous manganese oxide catalysts in the catalytic oxidation of benzene. The new sea-urchin-like coppermanganese oxide catalysts reached a maximum benzene conversion of 90% at 219 °C, which was 23 °C lower than that required for the crude in the MnO2-based catalyst. The improvement in the catalytic performance could be due to the increase in the specific area and to the desorption of oxygen from the lattice after the addition of Cu, thus increasing the oxygen vacancies on the catalyst surface. The oxidation of benzene should occur through the Mars and Van Krevelen mechanism, which is very common in the oxidation of VOCs on metal oxide. According to this mechanism, gaseous O2 molecules are initially adsorbed on the surface of the catalyst occupying the oxygen vacancies and converted into reactive oxygen species (ROS) (O2−, O− or O2−) through an electron transfer process between metal centers, which react with benzene molecules that lie on the surface

9.2  Benzene

of the catalyst. The final products of benzene degradation, CO2, and H2O, will be desorbed, returning electrons to the catalyst and recovering the oxygen vacancies on its surface [27]. Porous λ-MnO2 was prepared through Zn2+ ions extraction of the spinel-type zinc manganese oxide (ZnMn2O4) by acid leaching (3 M HNO3) and applied as a catalyst for benzene oxidation [28]. Combustion reactions were carried out in a continuous flow micro-reactor with a total flow rate of the reactant mixture (benzene + O2 + N2) and a gas hourly space velocity (GHSV) of 60 L g−1 h−1. Under these conditions, selectivity to CO2 was always above 99% with λ-MnO2 exhibiting a remarkably boosted catalytic activity in comparison to the precursor ZnMn2O4 with a conversion of 90% at 170 ºC and 280 ºC, respectively. Studies performed using in situ Diffuse Reflectance Infrared Fourier Transform (DRIFT) analyses suggest that the oxidation reaction of benzene should occur via phenolate formation via active oxygen species, subsequently producing benzoquinone; finally, the aromatic rings are transformed into carboxylate and maleate, which are then oxidized to the final products (CO2 and H2O) (Figure 9.1) [28]. A series of porous manganese oxides (MnOx) were prepared by Guo et al. using the solution combustion synthesis (SCS) method and varying the citric acid/manganese nitrate ratios [29]. These compounds were applied to the oxidation of benzene, and it was found that the presence of citric acid in the synthesis process improved the qualities of the catalyst, with the sample prepared with a ratio of citric acid to manganese nitrate of 2:1 showing the best results with a conversion of 90% occurring at 212 ºC under a GHSV of 60 L g−1 h−1. In situ FTIR analysis revealed the following product pathway: benzene → phenolate → o-benzoquinone → maleates+acetates+vinyl → CO2+H2O [29]. Pt–MOx/Al2O3 catalysts were prepared with bimetallic PtM (M = W or Mo) NPs with uniform morphologies and regular sizes and tested for benzene oxidation [30]. The Pt-WOx/Al2O3 sample revealed the best catalytic performance reaching 90% conversion at 140 °C under a GHSV of 40 L g−1 h−1 and the TOFPt was 19 times higher than with Pt/Al2O3. The high catalytic efficiency could be explained by the good ability to adsorb and activate benzene at MOx points in close contact with Pt. The interaction of the highly dispersed metallic Pt species with the small MOx ensembles was stronger in this case and probably responsible for the higher tolerance to chlorine poisoning leading to higher resistance of the Pt species to deactivation under the oxygen-rich and Cl-containing conditions, which is very common in the oxidative removal of VOCs (Figure 9.2) [30]. Interestingly, Co3O4 nanoparticles were supported on porous material derived from eggshell waste, by the impregnation method, allowing the production of a series of catalysts of Co3O4/eggshell type with different Co3O4 loadings [31]. The Co3O4/eggshell with a 16.7% Co3O4 loading catalyst was demonstrated in the catalytic oxidation of benzene, reaching 90% benzene conversion at 256 °C. The catalytic effect of these composite materials is remarkable compared to commercially pure Co3O4. The characterization and analysis of FTIR in situ shed some light on the possible mechanism of the reaction, in which active oxygen species (Olatt and Oads) appear to play an important role (Figure 9.3) [31]. Figure 9.1  The proposed mechanism for benzene oxidation on the λ-MnO2 catalyst. Reproduced with permission from Ref [28].

163

164

9  Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions

O2

e–

[O] (O–,O2–,O22–......)

Pt

C

MOx(M=W, Mo)

O

Al2O3

H

Figure 9.2  The proposed mechanism of benzene oxidation over the PtM/Al2O3 (M=W, Mo) samples. Reproduced with permission from Ref [30]. Figure 9.3  Proposed reaction steps of benzene oxidation over the Ag/eggshell catalyst.

Complete oxidation was also achieved with other mixed oxides, namely Co3O4–CeO2 oxides (20, 30, and 40 wt.%), prepared by the mechanochemical treatment of a mixture of cerium hydroxide and cobalt hydroxycarbonate precursors at relatively low temperatures (200–250 °C) [32]. This method, besides being environmentally more tolerable, proved to be more sustainable due to having both lower waste production and lower energy consumption. The Co3O4-30 Wt.% CeO2 catalyst allows the complete conversion of benzene to be achieved at 200 °C, a temperature significantly lower than that required for other transition metal oxides, and with high selectivity, as no products resulting from incomplete oxidation were detected. In another study, a simple, one-step solvothermal method was used to support Ag on Co3O4 (Ag/ Co3O4) [33]. The same catalyst was prepared by the impregnation method (Ag/Co3O4-I) and the activity of both were compared for the same silver content. The activity of the various catalysts followed the following order: 2%Ag/Co3O4>1%Ag/Co3O4>2%Ag/Co3O4-I>Co3O4, demonstrating a higher activity for the catalysts obtained by the solvothermal method, which may be due to the fact

9.2  Benzene

that the former has more surface-active oxygen species and more abundant lattice defects. It was also possible to confirm the promoting effect of the support on the various catalysts, with no effect on their morphology after the addition of Ag, as shown by scanning electron microscopy (SEM) and Brunauer–Emmett–Teller theory (BET), but improving their pore structure [33]. In the same year, novel three-dimensional mesoporous vanadia/meso-Co3O4 materials were reported as cheap, active, and selective catalysts for the catalytic oxidation of aromatic VOCs (benzene, toluene, and/or xylene) [34]. A low vanadium (V2O5) doping (1 wt%) on the Co3O4 surface could significantly improve the catalytic performance, conceivably promoting oxygen mobility, allowing the complete oxidation of either individual aromatic VOCs (benzene, toluene, and/or xylene) and/or their mixture to carbon dioxide, at low temperature [34]. Recently, Mn-based mullite-type oxides (e.g. SmMn2O5, GdMn2O5) were also designed to be possible substitutes for noble metal catalysts due to their low cost, high thermal stability, and high efficiency [35]. With this in mind, Liu et al. [36] synthesized porous Mn-based mullite SmMn2O5 by partially removing Sm3+ and using in situ dismutation reaction of Mn3+ from SmMnO3 perovskite. The catalytic activity of SmMn2O5 obtained by the partial removal of Sm and Mn was compared with that of SmMnO3 and crude SmMn2O5, with the former demonstrating a significant improvement in its benzene oxidation capacity. This feature is certainly related to the higher number of active manganese ions on the surface as well as other excellent physicochemical properties acquired after crude SmMn2O5 treatment [36]. The importance of support materials with large active surface area and specific physicochemical properties that allow a significant reduction in the amount of supported noble metals and metal oxides, as well as a greater dispersion of metal centers on their surface, was also recognized. The ZSM-5 material used by Todorova et al. [37] belongs to the zeolite family and emerges as a strong candidate for robust support for catalysts in the hydrocarbon oxidation reaction due to its microporous crystalline structure, high surface to pore volume ratio, and thermal resistance. Before metal impregnation, the ZSM-5 zeolite was treated with HF and NH4F buffer solutions. As a result of the acidic treatment, the specific surface area and micropore volume decreased, whereas the volume of the mesopores increased significantly. Pt, Pd, and Cu metal salts were used for metal impregnation on both ZSM-5 zeolites (initial and acid-treated analogues) through a wetness technique (Pt = 0.5 wt %, Pd = 0.5 wt %, and Cu = 5 wt %). Finally, ZSM-5 samples impregnated with metallic components were applied as heterogenous catalysts for CO and benzene combustion. After performing the catalytic studies, it was possible to verify the importance of the acid pre-treatment. In the case of benzene, as it is a molecule of larger size than CO, the existence of mesoporosity in the structure of the zeolite facilitated its access to the active centers and had a more prominent effect. In the oxidation of CO, most of the catalysts prepared with the acid pre-treated support showed increased activity [37]. The activity of the Pd nanoparticle supported SiO2 catalysts (0.5 wt%) (Pd-SiO2) were tested by Guan et al. in the oxidation of several simple and mixed aromatic hydrocarbons (benzene, toluene, and o-xylene) [38]. The high-resolution transmission electron microscopy (HRTEM) characterization technique showed that higher temperatures induced agglomeration of palladium particles. The catalytic oxidation of aromatic hydrocarbons showed that toluene is more easily oxidized over Pd-SiO2 composite than the binary mixture of benzene and toluene and the ternary mixture of benzene, toluene, and o-xylene [38]. In recent years, advanced oxidation processes (AOPs) have been considered to provide the most promising techniques to eliminate VOCs in both indoor and outdoor air environments. Among these, processes such as photocatalysis, ozone catalytic oxidation, and plasma processes stand out due to their high efficiency under environmentally friendly conditions [39]. In these processes, the oxidation involves the generation of reactive oxygen species (ROS) such as hydroxyl radicals

165

166

9  Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions

(•OH), superoxide radicals (•O2), hydroperoxyl radicals (•HO2), hydrogen peroxide (H2O2), and ozone (O3) to oxidize VOCs into CO2 and H2O. The vacuum ultraviolet (VUV) oxidation combines processes such as VUV photodegradation, photocatalytic oxidation (PCO), and ozone catalytic oxidation (OZCO). These processes may involve the generation of ozone by VUV irradiation, or the decomposition of O2 into oxygen atoms that may subsequently combine to give rise to O3. In turn, the decomposition of ozone and the photolysis of water by UV radiation may generate various reactive oxygen species (ROS). For example, Huang et al. attempted to oxidize gaseous benzene using VUV photooxidation [40]. The process efficiency was optimized by varying parameters such as air flow, relative humidity, concentration, and temperature of the pollutant stream, with the best conditions involving higher relative humidity and contact time, whereas high pollutant concentrations resulted in lower benzene degradation efficiency. The degradation mechanism of benzene is expected to involve hydroxyl radicals formed by action of VUV radiation in the presence of water [40]. PCO is a relevant process in VUV-based catalytic oxidation. Semiconductors are usually exploited as catalysts in PCO because the applied radiation energy is higher than their band gap and promotes the transfer of electrons from the valence band (VB) to the conductive band (CB), forming the pair (e−/h+) at the surface of the catalyst that can act as electron acceptor or donor in the reaction with VOCs. Among the semi-conductors, zinc oxide (ZnO) has been explored in several VOC degrading processes. In addition, several materials have been added to increase their photosensitivity, including materials with carbon structure such as graphene and graphene oxide. With that in mind, Jafari et al. deposited ZnO on silicaGO/fiberglass and succeed in the photocatalytic degradation of benzene from polluted air stream. The optimum moisture range for the catalytic removal of benzene was reported to be 35–45%. The degradation of benzene by this process had an inverse relationship with flow rate (i.e. with increasing flow rate, the removal efficiency decreased) [41]. Titanium oxide is another common oxide-based semiconductor material that was explored as a photocatalyst in several reactions of environmental and industrial interest due to its high efficiency, high stability, and low cost, in addition to being environmentally tolerable. In the presence of TiO2 and solar light, water vapor adsorbed to the catalyst surface can produce hydroxyl groups (OH−), which, in turn, by interaction with the electron-hole pair (e−/h+) created at the surface of the catalyst, gave rise to active species such as hydroxyl radicals (•OH), superoxide radicals (•O2−), oxygen atom radicals (•O−), or hydroperoxyl radicals (•HO2) that can be produced by subsequent reactions occurring on the TiO2 surface (Figure 9.4) [42].

Figure 9.4  Basic principles of photodegradation of volatile organic compounds (VOCs) using semiconductors.

9.3  Toluene

Recently, Mahmud et al. explored TiO2 P25 as a photocatalyst in the photodegradation of benzene, toluene, and p-xylene in the gas phase, either as a mixture or isolated, using several Langmuir-Hinshelwood kinetic models to better understand experimental outputs [43]. In that sense, it was possible to verify that the efficiency of benzene degradation was quite low (c. 10%) when it was mixed with other VOCs compared to the isolated mode (67%). The reason for the low degradation efficiency of benzene in the presence of the other VOCs is thought to be related to the weak interaction of benzene with the TiO2 surface when compared to toluene and p-xylene. It was also found that oxygen vacancy (Ov) on the surface of TiO2 remarkably improved the overall adsorption of VOCs and that the equilibrium constant of adsorption of VOCs was greater than the equilibrium constant of adsorption of water [43].

9.3  Toluene Toluene has been identified as one of the most serious pollutants both for the natural environment and human wellbeing, so therefore its elimination is of great interest at present. As a result, several technologies for its reduction have been explored, among the most promising being thermal or non-thermal catalytic oxidation that degrades toluene completely to CO2 and H2O [44, 45], adsorption [46, 47], photocatalytic oxidation [48–53], and thermal plasma-catalysis [54–57]. The adsorption process, despite being the simplest and least expensive, still has limitations such as the low adsorption capacity of some materials, the effect of relative humidity (RH), and the need for treatment of the desorbed material [58]. Photocatalytic oxidation, on the other hand, is similarly simple to operate, requires a lower oxidation temperature, and produces secondary pollutants [59–61]. Although catalytic oxidation of toluene is considered an alternative solution due to its high efficiency in toluene removal, this technique remains a challenge. The success of the catalytic oxidation of toluene is mostly dependent on the reaction conditions and catalyst performance. In turn, the activity of the catalyst is highly affected by its morphology, crystallinity, surface area and reducibility, active sites, and durability [62]. Among the most commonly applied catalysts for oxidative degradation are noble metal catalysts [63], transition metal catalysts, metal-organic based catalysts (metal-organic frameworks or MOFs) [64–66], perovskite catalysts [67–69], and spinel-based catalysts [70]. Despite the evidently better catalytic activity of noble metal-based catalysts [63] for VOC oxidation within the metallic family, largely related to their high efficiency and durability, they have the disadvantage of being scarce and therefore highly expensive, as well as being extremely sensitive to chloride poisoning, and this limits their industrial application. On the other hand, catalysts based on non-noble transition metal oxides [71, 72] and/or composite oxides [73–76] have demonstrated a high performance as catalysts for VOC oxidation, in addition to being inexpensive. Among these, manganese oxides (MnOx) have attracted growing attention for their exceptional redox properties and good poisoning resistance [77, 78]. To combine the advantages of both types of metals, substantial work on the synthesis and design of novel catalysts composed of noble metal nanoparticles (Pd, Pt, Au, Ru, and Ag) supported on different materials and metal oxide-based catalysts (Co, Mn, Ce, Al, La, Zn, Ni, and Cu) have been developed for various environmentally catalytic reactions (i.e. CO/VOCs degradation) and particularly for toluene oxidation [13, 62, 79–83]. Moreover, the addition of a second noble metal to produce a supported noble metals alloy has also been explored for toluene oxidation, achieving a better performance than the supported single metal catalyst [84]. The design of new catalysts remains a highly challenging task, and their selection is very dependent on the type of VOCs to be degraded and on the degradation method adopted. Several studies on supported noble metals have shown that the catalytic performance depends not only on

167

168

9  Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions

the supported noble metal but also on the nature of the support and its acid-base properties [85–87]. In some work, the role of metals and supports in the activation of molecular oxygen and toluene has been reported, as well as improvement of the surface dispersion of metal active sites. However, for these catalysts to become suitable it is necessary that, besides the synergistic effect, their design is aligned with the mechanism of toluene oxidation [88–90]. Hao et al. [91] and He et al. [92], for example, demonstrated the promoting effect of support acidity on the catalytic combustion of toluene when they used a series Figure 9.5  Promoting effect of support acidity on the catalytic combustion of of molecular sieves (of the ZSM-5 type) to support Pd toluene. catalysts (Figure 9.5). The benefit of having a greater number of acidic sites on the catalysts and consequently in the dispersion of active sites and desorption of CO2 from the catalysts may be the explanation for the improved catalytic performance. In 2019, Gan et al. prepared highly efficient Pt/Al2O3 type catalysts with ultralow Pt-loading for the catalytic combustion of toluene [93]. In this case, a synergistic promotional effect of Pt nanoparticles and Al2O3 support was observed. In fact, the in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) and electron paramagnetic resonance (EPR) results show that the metallic Pt nanoparticles (Pt0) are responsible for the activation of molecular oxygen while the support provides sites for the adsorption of toluene and desorption oxidation products, processes that are favored by the weak and medium strength acid-based sites. The doping of precious metals such as Pt, Pd, Au, and Ru has also been explored in the catalytic combustion of VOCs. In most cases, a synergistic effect was found due to modification of the electronic and/or geometric structure of the doped catalysts [79, 82, 94–96]. Hosseini et al. found a synergetic effect between noble metals on Pd and/or Au supported on a nanostructured mesoporous TiO2/ZrO2 catalyst. The mixed Au–Pd catalysts exhibited a higher catalytic activity in the combustion of toluene when compared with that of catalysts containing only one of the noble metals [97]. The synergistic effect between gold and palladium was also exploited for the combustion of toluene with Nb or Va doped bimetallic Pd-Au supported TiO2 catalysts [98], revealing that the doping of Nb and V enhanced the catalytic activity, particularly for the niobium doped catalyst (PdAu5NbTi) when compared with the doping-free sample (PdAu5VTi). A cyclic-like performance, additional production of mobile oxygen species, and the participation of these species in Pd particle reduction are suggested. Doping of catalysts (Pt/ZSM-5 type) with alkali metals (e.g. Na, K, and Cs) was also reported by Chen et al. [99] as resulting in an improved catalytic performance in toluene oxidation due to the occurrence of electronic promotion. The Pt/KZSM-5 catalyst is presented as the most active one among the doped catalysts with a remarkable activity even after 50 hours of contact. Analogous results were observed for aluminium-rich platinum Beta zeolite nanoparticles [87] with the presence of K+ in Pt/KBeta-SDS (i.e. seed-directed synthesis [SDS] of Beta zeolite) favoring the formation of more Pt0 species, and both the higher K+ content and the lower number of terminal silanol defects favoring toluene adsorption. In turn, the preparation of core-shell type catalysts based on different metals also seems to be an interesting technique to increase the catalytic activity in the oxidation of toluene, either by electronic modification or by lowering the segregation in bimetallic structures. Structured coreshell type catalysts containing Pd in the shell, such as Ag@Pd and Au@Pd, have shown high

9.3  Toluene

Li et al. reported highly active palladium bimetallic core structure catalysts supported on TiO2 substrate (Ag@Pd/TiO2) with a low palladium loading for toluene oxidation. UV-Vis absorption and selected area electron diffraction studies showed that the proximity between Ag and Pd and the different electronegativity in core-shell structure results in electron transfer from the Ag core to the Pd shell, which increases the local electron density around the Pd atom [101]. Recently, a series of Co3O4-based monolithic heterostructured catalysts were applied in the gaseous co-oxidation of CO and toluene [102]. The Co3O4@Co3O4 catalyst exhibited the most remarkable catalytic performances for the individual oxidation and co-oxidation of CO and toluene, being even comparable to those of Pt-based catalysts. The good performance was attributed to the large surface area, prominent surface oxygen vacancy and low-temperature reducibility. In situ DRIFT spectroscopy confirmed the bidentate carbonate species and benzoate species as reaction intermediary species in the oxidation of CO and toluene, respectively [102]. More recently, the core–shell structure Mn2O3@MnO2 (MnOx) has been prepared using KMnO4 modifying spherical Mn2O3 assembled by nanoparticles, and successfully applied for catalytic toluene oxidation [103]. The MnOx sample calcined at 350 °C (MnOx-1) exhibited a much higher catalytic activity compared with Mn2O3 and MnOx-2 (calcined at 450 °C) resulting from the higher surface Mn4+ content and many surface lattice oxygen species derived from the special core–shell structure, which improved the oxidation of toluene. In addition to eliminating VOCs, their selective oxidation is also a key issue (Figure 9.6) because it not only reduces CO2 formation but also allows the conversion of toluene into value-added organic compounds [104–106], namely benzaldehyde and benzoic acid, although the C-H bond cleavage in large-scale is difficult to control at high temperatures and pressure [107, 108]. Therefore, partial oxidation of toluene under mild and environmentally friendly conditions has become a real challenge. While new catalytic routes have been investigated [13, 109], several transition metal catalysts have been synthesized in the past few decades for this purpose, namely homogeneous and heterogeneous ones involving metal complexes of Cu [106, 110–113]. In 2019, Ta et al. reported a highly selective dinuclear oxovanadium(V) complex [V(PTANAP) (μ2-MeO)]2 prepared from an amide-imine conjugate, (E)-N’-((2-hydroxynaphthalen-1-yl) methylene)-4-methylbenzohydrazide, for the selective peroxidative oxidation (using a 70% t-BuOOH aq. sol. as an oxidant) of toluene and obtained, under conventional heating, at 60 °C for 18 h, a conversion of toluene of 35% with 100% selectivity for the benzaldehyde product [112]. The Cd(II) complex [Cd(1κONO’:2κO-HL2)(κOO’-Ac)]2·2DMF was tested as an heterogeneous catalyst for the microwave (MW)-assisted peroxidative oxidation of toluene [114]. The reaction was carried out oxidized toluene at moderate temperature (50 ºC) under MW, achieving a total (benzaldehyde+benzyl alcohol) product yield of 49% after one hour using t-BuOOH (70% aq. sol.) as oxidant, whereas 68% of the total product yield was obtained after 24 hour reaction using conventional heating (Figure 9.7). The simple Cd acetate salt Cd(OAc)2.2H2O, as well as the pro-ligand H2L2 used in the synthesis of the Cd(II) complex, were also tested for comparative purposes and, under typical reaction conditions (1 h, 50 ºC, 5 W), less than 1% total

Figure 9.6  Oxygenation of the methyl group in toluene.

169

170

9  Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions

Figure 9.7  MW-assisted peroxidative oxidation of toluene by Cd(II) compound [Cd(1κONO’:2κO-HL2) (κOO’-Ac)]2·2DMF.

yield was reached for the salt whereas no activity was detected for the ligand. This shows the important role of the coordination sphere, composed of an N and O donor at the Cd metal center. The mechanism of this reaction is suggested to involve the formation of the oxygenbased radicals t-BuOO⦁ and t-BuO⦁. The advantage of using MW irradiation was further studied by Sutradhar et al. with a new series of Cu(II) complexes derived from (5-bromo-2-hydroxybenzylidene)-2-hydroxybenzohydrazide (H2L) toward the peroxidative oxidation of toluene [110]. A maximum benzaldehyde yield of 44% at 50 °C in two and one-half hours with a polynuclear aroylhydrazone Cu(II) complex[Cu(L)]n was achieved. The reaction performed under the same reaction conditions but replacing MW by a conventional oil bath as a heating source, resulted only in 7% of benzaldehyde, showing the acceleration effect of the MW irradiation. The mechanism appears to follow several steps involving oxygen(t-BuO⦁or t-BuOO⦁) and carbon (R⦁) -based radicals. To obtain further insight into the active oxidant, the oxidation reaction was performed with the addition of di-tert-butylperoxide, t-BuOOt-Bu (no formation of t-BuOO⦁ by homolytic O-O cleavage), and an important yield drop occurred, suggesting the important role of t-BuOO⦁ in the radical mechanism [110]. Interestingly, for the catalytic oxidation of toluene in the presence of the tetranuclear Cu(II) cubane complex [Cu2(μ-1κONO’:2κOO’:3κO-HL)(μ-1κONO’:2κOO’-HL)]2�4DMF, under the explored reaction conditions, both methyl group and ring oxidations occur, with the formation of benzaldehyde and ortho-cresol as the main products (Figure 9.8) and a maximum product yield of 11% is achieved after three hours of reaction, with TBHP (70% aq.) as oxidant, at 80 ºC [115]. Advanced oxidation processes (AOP) were recently explored by Sutradhar et al. for the peroxidative oxidation of toluene with the dinuclear oxidovanadium(V) complexes [VO(HL1)(µ-O)]2 and [{VO(OEt)(EtOH)}2(L2)] derived from the aroylhydrazone Schiff bases 2-hydroxy-N′-(1-(pyrazin2-yl)ethylidene)benzohydrazide (H2L1) and N,N′-bis(2-hydroxybenzylidene)oxalohydrazide (H4L2), respectively. Besides the usual conventional heating, other energy inputs were explored (i.e. ultrasonic [bath or probe] and photo-assisted methods). The yields of oxygenated products from the peroxidative oxidation (H2O2, 30 % aq.) of toluene for one hour at room temperature followed the trend: US_PROBE (8.1%)>US_BATH (5.4%)>PHOTO (3.9%)>CONV (no oxygenated products were produced in this case) [116].

9.4  Ethylbenzene CH3

CH 3 Toluene

HO

OH

OH

O

O H N O

O

HN N

O

Cu

O

Cu

HO

OH Cu

Cu

O N HN

O

o-cresol

N

O

H

O

O N NH

O

HO

O

Benzyl alcohol

Benzaldehyde

Figure 9.8  Conversion of hazardous toluene in value-added products.

9.4  Ethylbenzene The oxidation of ethylbenzene leads to acetophenone, benzaldehyde, and benzoic acid as main products. Due to the importance of these products, the update of these catalytic systems has attracted considerable attention from researchers [117]. Additionally, the traditional synthetic routes for these compounds are not suitable for the preferred hazard-free industrial applications. For example, acetophenone is a valuable intermediate for manufacturing some pharmaceuticals, resins, alcohols, aldehydes, etc [118], and is traditionally prepared by Friedel-Crafts acylation of benzene and acetyl chloride using stoichiometric AlCl3 with the unavoidable production of chloride-containing wastes. Hence, this explains the growing interest in the development of ecofriendly heterogeneous catalysts for the oxidation of ethylbenzene. The catalyst activity of a cobalt oxide supported on mesoporous silicas SBA-15 (Santa-Barbara Amorphous) and KIT-6, prepared by wet impregnation, was investigated through ethylbenzene oxidation in acetonitrile in the presence of TPHB as the oxidizing agent. The aim of this study was to enhance the formation of benzoic acid and thus an excess of oxidant over substrate (15:1) was used. A maximum conversion of 37.1% and a selectivity of 88% for benzoic acid for 2% Co/SBA-15 were achieved (Figure 9.9). The use of the catalysts with KIT-6 as support produced more acetophenone than the catalysts supported on SBA-15, although the selectivity for the acid always remained high [119].

171

172

9  Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions

Figure 9.9  Ethylbenzene oxidation mechanism by Co/SBA-15 catalyst Reproduced with permission from Ref[119] / Elsevier.

ZSM-5 zeolite, incorporated in tetrahedral manganese (Mn-ZSM-5) was successfully prepared by an alternative method to the traditional one (namely, without an organic template and without calcination), which guarantees a higher commercial value. The Mn-ZSM-5-n favors the oxidation of the benzylic C-H bond in ethylbenzene using TBHP through a radical chain to produce acetophenone and 1-phenyl ethanol, and using the tetrahedral Mn(III) coordination as the active site of the catalytic redox process. Furthermore, it was possible to reuse MnZSM-5–50 for five cycles without a loss of activity, which may be due to the stability of the incorporated Mn [120]. The oxidation of ethylbenzene catalyzed by Mn-ZSM-5–50 is expected to start by the activation of the oxidant TBHP by coordination to the tetrahedral Mn(III) center, affording in an initial stage the oxygen-based radicals t-BuO⦁ and ⦁OH. Moreover, the t-BuO⦁ radical reacts with the α-H of ethylbenzene to form α-ethylbenzene radicals that can react with TBHP to produce 1-phenyl ethanol. The secondary alcohol reacts with t-BuO⦁ to produce ethylbenzene hydroperoxide. Finally, molecular dehydration of ethylbenzene hydroperoxide with t-BuO⦁ and ⦁OH occurred to yield acetophenone. Pd nanoparticles were immobilized on a composite based on modified carbonaceous material, g-C3N4-rGO, and their catalytic activity was investigated through the oxidation of ethylbenzene in acetonitrile in the presence of TBHP [121]. The oxidation occurred via a free-radical mechanism and resulted in the formation of acetophenone as the main product and benzaldehyde as a by-product. The best results were obtained with the Pd(4 wt%)/g-C3N4-rGO composite, showing 67% conversion and 97% selectivity for acetophenone, after optimizing the amount of catalyst (and Pd loading), reaction time and temperature. The hybrid material kept its activity and selectivity unchanged after five sequential runs. According to the authors, the catalytic reaction involves the formation of ethyl benzene hydroperoxide as an intermediate which could be activated by the catalyst via two different routes, C–C or a C–H bond break (Figure 9.10), the former being favored by lower temperatures.

9.5  Xylene p-Xylene is a very relevant raw material in the production of terephthalic acid (TA), which in turn is applied on a large scale in the production of polyethylene terephthalate (PET). Currently, the majority of terephthalic acid is produced via a liquid phase aerobic oxidation of p-xylene with air

9.5  Xylene

Figure 9.10  Conversion of ethyl benzene hydroperoxide to benzaldehyde (route A) or acetophenone (route B).

(a)

(b)

B

A

in acetic medium and a homogeneous Co/Mn catalytic system with hydrobromic acid (HBr) as a source of bromide ions, in the commonly known AMOCO (Co/Mn/Br) process. The corrosive and environmentally unfriendly nature of HBr is a major limitation in this process. For this reason, several studies have been developed aiming at improving the production conditions, either by developing new catalysts or by softening the reaction conditions [122–124]. p-Xylene oxidation occurs in steps with intermediate formation of alcohols and aldehydes and can be conventionally subdivided into a fast step from p-xylene to p-toluic acid followed by a slower step from p-toluic acid to terephthalic acid (Figure 9.11) [122]. Recently, Co3O4 catalysts derived from MOFs with different structures were shown to be efficient for o-xylene complete oxidation to CO2 and H2O [24]. The crucial roles of shape and the more active lattice oxygen were demonstrated because a rod-type Co3O4-R catalyst could adsorb and oxidize more o-xylene, forming intermediates such as alkoxide, carboxylate, and anhydride species and leaving oxygen vacancies due to the weak bonding of surface O2f (Figure 9.12). After the refilling of gaseous oxygen, the earlier disappearance of these intermediates on Co3O4-R indicated its good oxygen mobility. In fact, Co3O4-R exhibited enhanced catalytic performance with a xylene conversion of 90% at lower temperature, superior stability, and water resistance compared to spherical Co3O4-S [24]. Homogeneous catalytic oxidation of p-xylene has been the most common procedure explored for both commercial and laboratory synthesis of terephthalic acid and other intermediates. For example, Mendes et al. reported in 2017 the successful use of the iron tripodal complex C-scorpionate(II) [FeCl2{κ3-HC(pz)3}] under homogeneous conditions (in NCMe) for the oxidation of xylenes to their respective alcohols, aldehydes, and acids (no aromatic ring hydroxylation was observed) under mild reaction conditions, namely at low temperature (35 °C), with an environmentally friendly oxidant (H2O2, 30% aq. sol.) and a low catalyst loading, and in the presence of HNO3 as co-catalyst (n(HNO3)/n(catalyst) = 10). Under optimized conditions, high overall yield of 22% oxygenates (TOF = 1.3 x 102 h−1) (Figure 9.13) was obtained after only five minutes of reaction using p-xylene as substrate, whereas the maximum yield value obtained was 34% after six hours of reaction. The lower steric constraints are appointed to explain the better results achieved with the para isomer [125].

173

174

9  Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions

Figure 9.11  Stepwise oxidation of p-xylene to terephthalic acid.

Figure 9.12  Schematic representation of the catalytic mechanism for the catalytic oxidation of o-xylene over metal-organic framework (MOF)-derived Co3O4 with different shapes. Reproduced with permission from Ref [24] / American Chemical Society.

Figure 9.13  Oxidation of para-xylene to methylbenzyl alcohol, para-tolualdehyde and para-toluic acid catalyzed by a tripodal C-scorpionate iron(II) complex.

9.6  Final Remarks

Despite the advantages of heterogeneous catalysts concerning the easy catalyst separation from the reaction products and the possibility of its reuse, their application in p-xylene oxidation is still under development, although with encouraging results. As an example, vanadium C-scorpionate(IV) complexes supported on functionalized carbon nanotubes (CNTs) were screened for the heterogeneous oxidation of o-, m-, or p-xylene, with TBHP (sol. aq. 70%) for at least six cycles with preservation of their activity. Under MW irradiation for 12 hours at 80 °C, p-xylene was converted to the corresponding toluic acid (main product), tolualdehyde, and methylbenzyl alcohol with a total yield of 43% (TON = 1.34 × 103) and with 73% selectivity to p-toluic acid [126]. The 1D coordination polymer [Cu3(μ3-1κN3,2κN2O,3κN-L)(μ-NO3)(NO3)3(H2O)3]n⋅(NO3)n derived from aroylhydrazone N’-(di(pyridin-2-yl)methylene)pyrazine-2-carbohydrazide (HL) was also explored for the heterogeneous oxidation of the xylene isomers. Oxidation occurs at the methyl group, with the reactivity following the order p-xylene>m-xylene>o-xylene and methyl benzyl alcohol being the main product. A total yield up to 37% was achieved with 4-methylbenzyl alcohol, p-tolualdehyde, and p-toluic acid as the main products. In this case, no subsequent oxygenated products, such as p-carboxybenzaldehyde or terephthalic acid, were detected. Additionally, no oxidation of the aromatic ring was observed. Radical trapping experiments suggested that the reaction mechanism may involve radicals, conceivably produced in the catalyst-promoted decomposition of H2O2 with possible formation of hydroxo-Cu species [127]. The main reaction product from p-xylene selective oxidation obtained with manganese-iron mixed oxide materials (Mn/Fe/O) is terephthalic acid (TA), with small amounts of p-methylbenzyl alcohol, p-tolualdehyde, cresol, or 4-carboxylic benzaldehyde (4-CBA). The type of oxidizing agent (molecular O2, hydrogen peroxide, or TBHP) influences the behavior of the Mn/Fe/O catalysts, with the highest conversion (up to 85%) and a preferential selectivity to TA achieved for TBHP with an optimal n(p-xylene)/n(oxidant)=1:4 molar ratio beyond which the generation of side products increases and competes for active sites of the catalyst [128].

9.6  Final Remarks The use of VOCs in industrial and consumer products has increased dramatically and, consequently, maximum air contamination levels have been reached repeatedly that are causing adverse effects on human health and the environment. This chapter attempts to review the most recently explored strategies for the mitigation of the most common aromatic hydrocarbons (BTEX). VOCs can be removed by various techniques, including adsorption, catalytic oxidation, biological degradation, NTP oxidation, and photocatalytic decomposition, among others. The concern with parameters such as the regeneration and reuse of any adsorbent or catalyst and its effect on the evaluation of the costs and viability of the processes is notorious. One of the problems is the deactivation of catalysts during the process, and, in several cases, there has been a need for pre-­treatment of the catalyst to ensure its durability and stability. Improved efficiency has been achieved by combining certain metal oxides with hybrid adsorbents, or doping of precious metals such as Pt, Pd, Au, and Ru. The support of various metal catalysts on different materials allows a significant reduction in the amount of noble metals and metal oxides, as well as a greater dispersion of metal centers on their surface, thus increasing the efficiency and sustainability of the processes. This chapter pays special attention to the total or partial oxidation of BTEX pollutants, under homogeneous and heterogeneous conditions, based on various removal strategies. The understanding of catalytic mechanisms (e.g. interface boundary sites and synergistic effects), has been fundamental for the development of highly efficient and stable catalysts. Furthermore, the use of

175

176

9  Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions

more advanced characterization techniques will allow not only the knowledge of reaction routes, but also the in situ identification of possible intermediates. Finally, the aim of this work was to highlight developments in VOC mitigation, so that it may be useful in establishing more efficient technical approaches for the removal of the pollutant from the environment.

Acknowledgments The authors acknowledge the Fundação para Ciência e Tecnologia for the financial support through the multiannual funding to Centro de Química Estrutural (UIDB/00100/2020, UIDP/00100/2020 and LA/P/0056/2020) and for the EXPL/QUI-QOR/2069/2022 and 2022.02069.PTDC projects, and from the Instituto Politécnico de Lisboa for the IPL/2022/MMOF4CO2_ISEL project.

References 1 Khawaja, R., Veerapandian, S.K.P., Bitar, R. et al. (2022). Chem. Synth. 2: 13. 2 Heck, R.M., Farrauto, R.J., and Gulati, S.T. (2016). Catalytic Pollution Control, 3e. New York: Wiley Interscience. 3 Kim, K.-H. (ed.) (2017). Volatile Organic Compounds in Environment. Basel, Switzerland: MDPI. 4 Sirotkin, A.V. (2019). Reproductive effects of oil-related environmental pollutants. In: Encyclopedia of Environmental Health, 2e (ed. J. Nriagu), 493–498. Elsevier. 5 Balzer, R., Probst, L.F.D., Drago, V. et al. (2014). Braz. J. Chem. Eng. 31: 757. 6 Dehghani, M., Fazlzadeh, M., Sorooshian, A. et al. (2018). Ecotoxicol. Environ. Saf. 155: 133–143. 7 da Silva, A.G.M., Fajardo, H.V., Balzer, R. et al. (2015). J. Power Sources 285: 460. 8 Tang, W.X., Liu, G., Li, D.Y. et al. (2015). Sci. China: Chem. 58: 1359. 9 Trinh, Q.H., Lee, S.B., and Mok, Y.S. (2015). J. Hazard. Mater. 285: 525. 10 Słomińska, M., Król, S., and Namieśnik, J. (2013). Crit. Rev. Environ. Sci. Technol. 43 (14): 1417–1445. 11 Li, X., Wang, J., Guo, Y. et al. (2021). Chem. Eng. J. 411: 128558. 12 Wang, Y., Su, X., Xu, Z. et al. (2016). Appl. Surf. Sci. 363: 113–121. 13 Guo, Y., Wen, M., Li, G., and An, T. (2021). Appl. Catal. B: Environ. 281: 119447. 14 Urbutis, A. and Kitrys, S. (2014). Open Chem. J. 12: 492–501. 15 Wang, Y., Yang, D., Li, S. et al. (2018). Microporous Mesoporous Mater. 258: 17–25. 16 Minh, N.T., Thanh, L.D., Trung, B.C. et al. (2018). Clean Techn. Environ. Policy 20: 1861–1873. 17 Kim, K., Kang, C., You, Y. et al. (2006). Catal. Today 111: 223–228. 18 Joung, H., Kim, J., Oh, J. et al. (2014). Appl. Surf. Sci. 2014 (290): 267–273. 19 Shayegan, Z., Lee, C., and Haghighat, F. (2018). Chem. Eng. J. 334: 2408–2439. 20 Zou, W., Gao, B., Ok, Y.S., and Dong, L. (2019). Chemosphere 218: 845–859. 21 Huang, Y., Ho, S.S., Lu, Y. et al. (2016). Molecules 21: 56. 22 Chang, T., Shen, Z., Huang, Y. et al. (2018). Chem. Eng. J. 348: 15. 23 Ollegott, K., Wirth, P., Oberstebeulmann, C. et al. (2020). Chemie Ingenieur Technik 92: 1542–1545. 24 Ma, Y., Wang, L., Ma, J. et al. (2021). ACS Catalysis 11 (11): 6614–6625. 25 Fu, X.F., Zeltner, W.A., and Anderson, M.A. (1995). Appl. Catal. B Environ. 6: 209. 26 Yang, -H.-H., Du, J., Wu, M. et al. (2021). Chem. Eng. Technol. 427: 132075.

References

27 Lee, H.J., Yang, J.H., You, J.H., and Yoon, B.Y. (2020). J. Ind. Eng. Chem. 89: 156–165. 28 Li, L., Yang, Q., Wang, D. et al. (2021). Chem. Eng. Journal 421 (2): 127828. 29 Guo, H., Zhang, Z., Jiang, Z. et al. (2020). Res. J. Environ. Sci. 98: 196–204. 30 Zhang, K., Dai, L., Liu, Y. et al. (2020). Appl. Catal. B: Environ. 279: 119372. 31 Li, Z., Yang, D.-P., Chen, Y. et al. (2020). Mol. Catal 483: 110766. 32 Ilieva, L., Petrova, P., Venezia, A.M. et al. (2021). Mechanochemically prepared Co3O4-CeO2 catalysts for complete benzene oxidation. Catalysts 11: 1316. 33 Ma, X., Yu, X., and Ge, M. (2021). Catal. Today 376: 262–268. 34 Shamma, E., Said, S., Riad, M., and Mikhail, S. (2021). Environ. Technol. doi:10.1080/09593330.20 21.2007288. 35 Dong, A., Gao, S., Wan, X. et al. (2020). Appl. Catal. B: Environ. 271: 118932. 36 Liu, R., Zhou, B., Liu, L. et al. (2021). J. Colloid Interface Sci. 585: 302–311. 37 Todorova, T., Petrova, P., and Kalvachev, Y. (2021). Molecules 26: 5893. 38 Guan, Y., Deng, G., Cheng, Y. et al. (2020). Chem. Phys. Lett. 754: 137508. 39 Wu, M., Huang, H., and Leung, D.Y.C. (2022). J. Environ. Manage 307: 114559. 40 Huang, H., Liu, G., Zhan, Y. et al. (2017). Catal. Today 281 (3): 649–655. 41 Jafari, A.J., Kalantary, R.R., Esrafili, A., and Arfaeinia, H. (2018). Process Saf. Environ. Prot. 116: 377–387. 42 Dharma, H.N.C., Jaafar, J., Widiastuti, N. et al. (2022). Membranes 12: 345. 43 Mahmood, A., Wang, X., Xie, X., and Sun, J. (2021). J. Environ. Chem. Eng. 9 (2): 105069. 44 Krishnamurthy, A., Adebayo, B., Gelles, T. et al. (2020). Catal. Today 350: 100–119. 45 Hoseini, S., Rahemi, N., Allahyari, S., and Tasbihi, M. (2019). J Clean Prod. 232: 1134–1147. 46 Yin, K., Zhang, H., and Yan, Y. (2019). J. Solid State Chem. 279: 120976. 47 Song, M., Yu, L., Song, B. et al. (2019). Environ. Sci. Pollut. Res. 26: 22284–22294. 48 Xie, R., Lei, D., Zhan, Y. et al. (2020). Chem. Eng. J. 386: 121025. 49 Zhu, B., Zhang, L., Yan, Y. et al. (2019). Plasma Sci. Technol. 21: 115503. 50 Lee, Y.E., Chung, W.C., and Chang, M.B. (2019). Environ Sci. Pollut. Res. 26: 20908–20919. 51 Jung, S., Fang, J., Chadha, T.S., and Biswas, P. (2018). Phys. D: Appl. Phys. 51: 445206. 52 Li, -J.-J., Cai, S.-C., Yu, E.-Q. et al. (2018). Appl. Catal. B: Environ 233: 260–271. 53 Okunaka, S., Tokudome, H., and Hitomi, Y. (2020). J. Catal. 391: 480–484. 54 Liu, R., Song, H., Li, B. et al. (2021). Chemosphere 263: 127893. 55 Yang, S., Yang, H., Yang, J. et al. (2020). Chem. Eng. J. 402: 126154. 56 Yu, X., Dang, X., Li, S. et al. (2020). J. Clean. Prod. 276: 124251. 57 Yang, S., Bo, Z., Yang, H. et al. (2018). Ind. Eng. Chem. Res. 57 (45): 15291–15300. 58 Peta, S., Zhang, T., Dubovoy, V. et al. (2018). Mol. Catal. 444: 34–41. 59 Li, Y., Li, W., Liu, F. et al. (2020). J Nanopart. Res. 22: 122. 60 Li, Y.W. and Ma, W.L. (2021). Chemosphere 280: 130667. 61 Ezeh, C.I., Tomatis, M., Yang, X. et al. (2018). Ultrason. Sonochem. 40: 341–352. 62 Zhang, Y., Liu, Y., Xie, S. et al. (2019). Environ. Int. 128: 335–342. 63 Liotta, L.F. (2010). Appl. Catal. B: Environ 100: 403–412. 64 Zhao, L., Zhang, Z., Li, Y. et al. (2019). Appl. Catal. B Environ. 245: 502–512. 65 Chen, X., Cai, S., Yu, E. et al. (2019). Appl. Surf. Sci. 475: 312–324. 66 Peedikakkal, A.M.P., Jimoh, A.A., Shaikh, M.N., and Bassam, E.A. (2017). Arab. J. Sci. Eng. 42: 4383–4390. 67 Dong, S., Chen, T., Xu, F. et al. (2022). Catalysts 12: 763. 68 Zang, M., Zhao, C., Wang, Y., and Chen, S. (2019). J. Saudi. Chem. Soc. 23 (6): 645–654. 69 Pan, K.L., Pan, G.T., Chong, S., and Chang, M.B. (2018). J. Environ. Sci. 69: 205–216.

177

178

9  Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions

70 Wang, Y., Arandiyan, H., Liu, Y. et al. (2018). ChemCatChem 10: 3429–3434. 71 Brunet, J., Genty, E., Barroo, C. et al. (2018). Catalysts 8: 64. 72 Sanchis, R., Alonso-Domínguez, D., Dejoz, A. et al. (2018). Materials 11: 1387. 73 Chen, G., You, K., Gong, X. et al. (2022). React. Chem. Eng. 7: 898–907. 74 Chen, G., You, K., Zhao, F. et al. (2022). Res. Chem. Intermed. 48: 2593–2606. 75 Zeng, J., Xie, H., Zhang, G. et al. (2020). Ceram. Int. 46 (13): 21542–21550. 76 Xu, Y., Qu, Z., Ren, Y., and Dong, C. (2021). Appl. Surf. Sci. 560: 149983. 77 Chen, J., Chen, X., Yan, D. et al. (2019). Appl. Catal. B-Environ. 250: 396–407. 78 Yang, X., Yu, X., Jing, M. et al. (2019). ACS Appl. Mater. Inter. 11: 730–739. 79 Carabineiro, S.A.C., Chen, X., Martynyuk, O. et al. (2015). Catal. Today 244: 103–114. 80 Zhang, X., Zhao, J., Song, Z. et al. (2019). Chem. Sel. 4: 8902–8909. 81 Zhu, B., Yan, Y., Li, M. et al. (2018). Plasma Process. Polym. 15: 1700215. 82 Torrente-murciano, L., Solsona, B., Agouram, S., and Sanchis, R. (2017). Catal. Sci. Technol. 7: 2886–2896. 83 Fu, X., Liu, Y., Yao, W., and Wu, Z. (2016). Catal Commun. 83: 22–26. 84 Xie, S., Deng, J., Zang, S. et al. (2015). J. Catal. 322: 38–48. 85 Liu, W., Xiang, W., Guan, N. et al. (2021). Sep. Purif. Technol 278: 119590. 86 Zeng, X., Cheng, G., Liu, Q. et al. (2019). Ind. Eng. Chem. Res. 58 (31): 13926–13934. 87 Chen, C., Wu, Q., Chen, F. et al. (2015). Mater. Chem. A. 3: 5556. 88 Jiang, B., Xu, K., Li, J. et al. (2021). J. Hazard. Mater. 405: 124203. 89 Zhang, C., Wang, C., Huang, H. et al. (2019). Appl. Surf. Sci. 486: 108–120. 90 Chen, C., Chen, F., Zhang, L. et al. (2015). Chem. Commun. 51: 5936–5938. 91 He, C., Li, J., Cheng, J. et al. (2009). Ind. Eng. Chem. Res. 48: 6930–6936. 92 He, C., Shen, Q., and Liu, M. (2014). J. Porous Mater. 21: 551–563. 93 Gan, T., Chu, X., Qi, H. et al. (2019). Appl. Catal. B Environ. 257: 117943. 94 Priya, M., Kiruthika, S., and Muthukumaran, B. (2017). Ionics 23: 1209–1218. 95 Chen, G., Zhao, Y., Fu, G. et al. (2014). Science 344: 495–499. 96 García, T., Solsona, B., and Taylor, S.H. (2014). Catal. Lett. 97: 99–103. 97 Hosseini, M., Siffert, S., Cousin, R. et al. (2009). C. R. Chim. 12: 654–659. 98 Barakat, T., Rooke, J.C., Chlala, D. et al. (2018). Catalysis 8: 574. 99 Chen, C., Wang, X., Zhang, J. et al. (2015). Catal. Today 258: 190–195. 100 Abdel-Fattah, W.I., Eid, M.M., Abd El-Moez, S.I. et al. (2017). Life Sci. 183: 28–36. 101 Li, Y., Liu, F., Fan, Y. et al. (2018). Appl. Surf. Sci. 462: 207–212. 102 Mo, S., Zhang, Q., Sun, Y. et al. (2019). J. Mater. Chem. A. 7: 16197. 103 Zhang, C., Li, M., Wang, X. et al. (2022). J. Chem. Technol. Biotechnol. 97: 1138–1148. 104 Kesavan, L., Tiruvalam, R., Ab Rahim, M.H. et al. (2011). Science 331: 195–199. 105 Huang, H., Ye, W., Song, C. et al. (2021). J. Mater. Chem. A 9: 14710. 106 Lisicki, D., Maciej, A., and Orlińska, B. (2021). Ind. Eng. Chem. Res 60 (30): 11579–11589. 107 Gunay, A. and Theopold, K.H. (2010). Chem. Rev. 110: 1060. 108 Feng, J.-B. and Wu, X.-F. (2015). Appl. Organomet. Chem. 29: 63–86. 109 Wang, P., Wang, J., An, X. et al. (2021). Appl. Catal. B: Environ 282: 119560. 110 Sutradhar, M., Barman, T.R., Pombeiro, A.J.L., and Martins, L.M.D.R.S. (2018). Molecules 24: 47. 111 Sutradhar, M., Alegria, E.C.B.A., Roy Barman, T. et al. (2017). Mol. Catal. 439: 224. 112 Ta, S., Ghosh, M., Ghosh, K. et al. (2019). ACS Appl. Bio Mater. 2: 2802. 113 Lapari, S.S. and Parham, S. (2013). Int. J. Eng. Sci. Invent 2: 62–67. 114 Sutradhar, M., Roy Barman, T., Alegria, E.C.B.A. et al. (2020). New J. Chem. 44: 9163.

References

115 Sutradhar, M., Alegria, E.C.B.A., Barman, T.R. et al. (2021). Inorg. Chim. Acta 520: 120314. https://doi.org/10.1016/j.ica.2021.120314. 116 Sutradhar, M., Martins, M.G., Simões, D.H.B.G.O.R. et al. (2022). Appl. Catal. A: Gen. 638: 118623. 117 Rahman, M.M., Ara, M.G., Rahman, M.S. et al. (2020). J. Nanomater 2020:7532767. https://doi. org/10.1155/2020/7532767. 118 Kirar, J.S. and Khare, S. (2018). Appl. Organometal. Chem. 32: e4408. 119 Unnarkat, A.P., Singh, S., and Kalan, S. (2021). Materials Today: Proceedings 45 (4): 3991–3996. 120 Liu, X., Gao, S., Yang, F. et al. (2020). Res. Chem. Intermed. 46: 2817–2832. 121 Nilforoushan, S., Ghiaci, M., Hosseini, S.M. et al. (2019). New J. Chem. 43: 6921. 122 Karakhanov, E.A., Maksimov, A.L., Zolotukhina, A.V., and Vinokurov, V.A. (2018). Russ. J. Appl. Chem. 91: 707–727. 123 Fadzil, N.A.M., Rahim, M.H.A.R., and Maniam, G.P. (2014). Chinese J. Catal. 35 (10): 1641–1652. 124 Tomás, R.A., Bordado, J.C., and Gomes, J.F. (2013). Chem Rev. 113 (10): 7421–7469. 125 Mendes, M., Ribeiro, A.P.C., Elisabete, C.B.A. et al. (2017). Polyhedron 125: 151–155. 126 Wang, J., Martins, L.M.D.R.S., Ribeiro, A.P.C. et al. (2017). Chem. - An Asian J. 12: 1915. 127 Sutradhar, M., Roy Barman, T., Alegria, E.C.B.A. et al. (2019). Dalton. Trans. 48: 12839. 128 Nicolae, S., Neaţu, F., and Florea, M. (2018). C. R. Chim 21 (3–4): 354–361.

179

181

10 Catalytic Cyclohexane Oxyfunctionalization Manas Sutradhar1,*, Elisabete C.B.A. Alegria2, M. Fátima C. Guedes da Silva3, and Armando J.L. Pombeiro3,* 1 Centro de Química Estrutural and Departamento de Engenharia Química, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal and Faculdade de Engenharia, Universidade Lusófona - Centro Universitário de Lisboa, Campo Grande 376, Lisboa, Portugal 2 Centro de Química Estrutural and Departamento de Engenharia Química, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal and Departamento de Engenharia Química, ISEL, Instituto Politécnico de Lisboa, Portugal 3 Centro de Química Estrutural and Departamento de Engenharia Química, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal * Corresponding authors

10.1  Introduction In this century, the utilization of hydrocarbons has become a predominant topic of high importance for global carbon management in terms of economic and sustainable aspects. Natural oil and gas are the greatest source of many hydrocarobons, but the activation of CH bonds of alkanes for their functionalization is a challenge [1]. Important points concern, for instance: (i) the relative stability and inert nature of the CH bond in saturated hydrocarbons; and (ii) the further reactivity of the desired products, and (iii) the need to avoid overoxidation. Despite all of these difficulties, the selective oxidation of alkanes remains an important but challenging area in modern catalysis to obtained desired oxygenated products from a less expensive and still readily available, although very stable, hydrocarbon feedstock [1–5]. The selective oxidation of cyclohexane (CyH) for the production of cyclohexanol and cyclohexanone (also known as KA oil) is of particular industrial relevance. These species are used for adipic acid and caprolactam syntheses, which are important precursors for the manufacture of nylon-6,6 and nylon-6 polymers, respectively [6, 7]. The commercial cyclohexane oxidation process is carried out in liquid phase at 150–160 °C, and 10–20 bars of oxygen or air pressure, using cobalt or manganese salts as homogeneous catalysts [6–10]. As KA oil is more reactive than cyclohexane, the desired high selectivity to the products of cyclohexane oxidation is obtained in industry only at a low cyclohexane conversion (4–6%) to prevent further oxidation of the desired products into acids and esters. Moreover, it is costly to separate the homogeneous catalysts from the products. Therefore, the development of efficient routes for the catalytic and selective oxidation of cyclohexane to KA oil under mild conditions is still a challenging task in the field of oxidation catalysis.

Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 1, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

182

10  Catalytic Cyclohexane Oxyfunctionalization

In this chapter, we discuss various transition metal-based catalysts (focused often on the work of our group) with respect to the efficient and selective oxidation of cyclohexane primarily to KA oil under mild conditions. Several catalytic systems involving the use of non-conventional techniques [microwave (MW) irradiation or ultrasounds], the application of ionic liquids or scCO2 as a reaction medium, and catalytic recycling are discussed, as well as the eventual replacement of traditional homogeneous catalysts by more efficient, selective and environmentally benign ones.

10.2  Transition Metal Catalysts for Cyclohexane Oxidation Various transitional metal catalysts show efficient homogeneous catalytic activity toward the oxidation of cyclohexane (via the formation of cyclohexylhydroperoxide, CyOOH) to cyclohexanol and cyclohexanone (KA oil mixture) (Scheme 10.1) using H2O2 or tert-butyl hydroperoxide (TBHP) as an oxidant under mild conditions. In some cases, the addition of a small amount of an acid such as 2-pyrazinecarboxylic acid (PCA) [11–17], nitric acid [18], sulfuric acid [19], or oxalic acid [20], as a promoter can significantly improve the activity.

Scheme 10.1  Metal catalyzed oxidation of cyclohexane to cyclohexanol and cyclohexanone.

10.2.1  Vanadium Catalysts Oxidovanadium(V) complexes are good candidates as catalysts or catalyst precursors toward oxidation of cyclohexane to cyclohexanol and cyclohexanone under mild conditions. Several of these complexes with azine fragment ligands and different nuclearities show efficient catalytic properties for this reaction. For example, the dinuclear oxidovanadium(V) complexes [{VO(OEt)(EtOH)}2L] (1) (H4L = bis(2-hydroxybenzylidene)terephthalohydrazide) and [{VO(OEt)(EtOH)}2L1] (2) (H4L1 = bis(2-hydroxybenzylidene)oxalohydrazide), the hexanuclear mixed-valence oxidovanadium(IV/V) complex [V3O3(OEt)(ashz)2(μ–OEt)]2 (3) (H3ashz = N′-acetylsalicylhydrazide), and the three heterometallic dioxidovanadium(V)/ alkali metal coordination polymers derived from H4L1 [(VO2)2(μ4-L1){Na2(μ-H2O)2(H2O)2}]n (4), [{V(μ-O)2}2(μ4-L1){K2(μ-H2O)2(H2O)2}]n (5), [{V(μ-O)(μ3-O)}2(μ8-L1){Cs2(μ-H2O)2(H2O)2}]n (6) [12, 13, 20] and [VO{N(CH2CH2O)3}] (7) [16, 21] are highly active as catalysts or catalytic precusors for the cyclohexane oxidation under mild conditions. The dinuclear oxidovanadium(V) complex 2, the mixed-valence oxidovanadium(IV/V) complex 3 and the heterometallic VV/Cs coordination polymer 6 showed the highest catalytic activities with total yields of up to 32.5% of cyclohexanol and cyclohexanone in the presence of PCA as a promoter under mild conditions [12, 13, 20]. The 1/PCA/H2O2 system led to the highest turnover number (TON), 4.4 × 104 at 50 °C, with a low catalytic load. In the presence of PCA and under optimized conditions, the aqua-soluble coordination polymers 4−6 having similar [(VO2)2(μ4/8-L)]2− cores follow the catalytic activity order 4scCO2>CH3CN. 27 is more active than 28 in each medium. In [bmim][PF6], 27 led to the highest oxygenates yield of 21%. Both catalysts showed a significantly high selectivity toward cyclohexanone formation (above 95%) and the catalysts can be recycled in the scCO2-IL media, preserving their activity up to three consecutive cycles [35].

10.2  Transition Metal Catalysts for Cyclohexane Oxidation

Heterogenization of the molecular catalyst on carbon materials can result in highly active systems with possibility of catalyst recycling. For instance, 27 deposited on carbon nanotubes leads to yields up to 21% and TONs up to 5.6 × 103 [36, 37]. By using ozone (O3) instead of a peroxide as oxidant, [FeCl2{HC(pz)3}] (27) catalyzes the cyclohexane oxidation to a higher oxidation level, with formation of adipic acid, via cyclohexanol and cyclohexanone (Scheme 10.3). A maximum yield of 96% of adipic acid was obtained in 6 h at room temperature with 27. In contrast to the industrial process, this process occurs via single-pot, is added solvent free, and does not require high temperatures, neither the use of HNO3 which would lead to the formation of noxious NO2 [38].

Scheme 10.3  Single-pot oxidation of cyclohexane to adipic acid catalyzed by 27 [38] / Royal Society of Chemistry.

Heteronuclear iron complexes can also catalyze the peroxidative oxidation of cyclohexane, such as [Cu4Fe2(OH)(Piv)4(tBuDea)4Cl]·0.5CH3CN (28) where PivH=pivalic acid and tBuDea=N-tertbutyldiethanolamine, Under mild rection conditions (50 °C, in CH3CN, 2 h) up to 17% total yield of KA oil was obtained with 28 using PCA acid as a promoter [39]. Other mixed iron-copper and iron-cobalt catalysts are discussed in the following sections. Iron(III) complexes can also be catalysts toward other types of functionalization of cyclohexane (e.g. its conversion to chlorocyclohexane). The dinuclear iron(III) compound [Fe2(BPA)2(µOCH3)2(Cl)2] (29), where HBPA=N-(2-hydroxybenzyl)-N-(pyridin-2-ylmethyl)amine, and the mononuclear [Fe(HBPClNOL)(Cl)2] (30), where H2BPClNOL=N-(2-hydroxybenzyl)-N-(2pyridylmethyl)(3-chloro)(2-hydroxy)propylamine, efficiently catalyze (with an excellent selectivity) the functionalization of cyclohexane to chlorocyclohexane as the unique product using TCCA (trichloroisocyanuric acid) as chlorinating agent. A maximum yield of c. 35% is obtained with 30 at 50 °C [40].

10.2.3  Cobalt Catalysts The bis(μ-chlorido)-bridged cobalt(II) complex [Co2(μ-Cl)2(HL2)4][CoCl4] (31) derived from the silyl-containing Schiff base HL2 (in situ reaction of 2,6-diformyl-4-methylphenol with trimethylsilylmethyl p-aminobenzoate in a 1:2 molar ratio) behaves as an efficient homogeneous (pre)catalyst

189

190

10  Catalytic Cyclohexane Oxyfunctionalization

in the MW-assisted neat oxidation of cyclohexane with aqueous TBHP. A total yield of 37.7% with a selectivity >99% was obtained in 1h 30m in the presence of PCA as co-catalyst at 100 °C. The solvent-free protocol with the short reaction time under MW irradiation is of significance toward the development of a sustainable oxidation of cyclohexane to KA oil [41].

The binuclear Co(II) complex [CoCl(µ-Cl)(HpzPh)3]2 (32) was obtained by reaction of tris(pyrazol-1-yl)methane [HC(pzPh)3] with CoCl2, involving C(sp3)-N(pyrazolyl) bond rupture in HC(pzPh)3. The mononuclear [CoCl2(HpzPh)4] (33) was obtained by the reaction of CoCl2 with HpzPh (HpzPh=3-phenylpyrazole). Both 32 and 33 act as catalyst precursors for peroxidative oxidation of cyclohexane to cyclohexanol (main product) and cyclohexanone in acetonitrile-water medium using H2O2 as oxidant. An overall TON up to 223 is obtained with 32 in 6 h [42]. The heterometallic complex [Co4Fe2OSae8]·4DMF·H2O (34) derived from salicylidene-2-ethanolamine (H2Sae) shows an excellent activity toward oxidation of cycloalkanes with H2O2, under mild conditions (in CH3CN, 5 h reaction time and an ambient temperature of c. 20 °C) with a total yield of 26% and TON = 3.57 × 103 (in the case of cyclohexane substrate) [43]. The ESI-MS of a solution of 34 indicated the presence of two major species, [Co(HSae)2]+ and [Co2Fe(Sae)4]+ [43]. The oxidative behaviour of m-chloroperoxybenzoic acid (m-CPBA) was investigated toward the oxidation of the sp3 C–H bond with the heterometallic pre-catalyst [Co4Fe2O(Sae)8]·4DMF·H2O (35) (H2Sae = salicylidene-2-ethanolamine) and HNO3 as a promoter [44]. This system catalyzes the mild hydroxylation of tertiary C–H bonds with 99% retention of stereoconfiguration of a model

10.2  Transition Metal Catalysts for Cyclohexane Oxidation

alkane substrate (cis-1,2-dimethylcyclohexane) with high TOF (up to 2 s−1) and TON (up to 1.4 × 104) values at 50 °C. Combined kinetic studies (including isotope effects), isotopic labeling (18O2, H218O, D2O), ESI-MS spectroscopy, and density functional theory (DFT) theoretical studies suggest that the oxidation proceeds through a concerted mechanism involving a cobalt-peroxo C–H attacking species or via a cobalt-oxyl species (rebound process) instead of a free-radical pathway (see also section on Mechanisms). The oxidation state of Co(III) catalyst remain unchanged during the most energetically favored pathway suggesting the involvement of metal–ligand cooperativity [44].

10.2.4  Copper Catalysts Many copper complexes have already been reported as catalysts toward the oxidation of cyclohexane to cyclohexanol and cyclohexanone. For instance, the mononuclear aroylhydrazone Cu(II) complexes [Cu(H2L)(NO3)(H2O)] (36) and [Cu(H2L)Cl]·2MeOH (37) were obtained from the reaction of Schiff base (3,5-di-tert-butyl-2-

191

192

10  Catalytic Cyclohexane Oxyfunctionalization

hydroxybenzylidene)-2-hydroxybenzohydrazide (H3L) with an appropriate copper(II) salt [45]. Both 36 and 37 present the benzohydrazide ligand in the keto form and effectively catalyze the peroxidative oxidation of cyclohexane to cyclohexyl hydroperoxide, cyclohexanol, and cyclohexanone in NCMe/H2O media, under mild conditions and in the absence of a promoter or co-catalyst. Catalyst 36 gives a maximum overall yield of 30% and TON 294 after 24 h (at room temperature and without any additive) [45].

The water soluble Cu(II) complex [Cu(H2L1)(H2O)(im)]∙3H2O (38) (im = imidazole) [46], derived from the ortho-substituted arylhydrazone of barbituric acid 5-(2-(2-hydroxyphenyl)hydrazono) pyrimidine-2,4,6(1H,3H,5H)-trione (H4L1) (im = imidazole) [46], catalyzes the peroxidative (with H2O2) oxidation of cyclohexane at room temperature (yield up to 21% and TON up to 213). The ease of 38 to undergo redox processes, as observed by cyclic voltammetry, supports the possibility of the free-radical cyclohexane oxidation.

The mononuclear copper(II) tetrazolato complexes [Cu(pmtz)2(phen)] (39) and [Cu(pytz)2(phen)] (40) (pmtz = 5-(2-pyrimidyl)tetrazolate; pytz = 5-(2-pyridyl)tetrazolate) can be obtained by metal mediated [2+3] cycloaddition reaction between a copper bound azide polymer and appropriate organonitriles [47]. Both 39 and 40 were examined as homogeneous catalysts for the oxidation of cyclohexane to cyclohexanol and cyclohexanone with TBHP under solvent- and additive-free mild conditions using MW irradiation. The obtained yields (up to c. 21%) are comparable to those of other efficient catalysts, but the use of MW irradiation features the significance of the catalytc process in terms of sustainability. The mononuclear Cu(II) complexes with different numbers of phenantroline ligands [Cu(phen)3] Cl2·7H2O (41), [CuCl(phen)2]Cl·5H2O (42), and [CuCl2(phen)] (43) have been explored in the oxidation of cyclohexane using hydrogen peroxide as oxidant, in acetonitrile-water solution at a 25 °C to 70 °C temperature range [48]. The catalyst 42 gives a high conversion with a total yield of 67% (TON = 737) at 70 °C. At a higher temperature range (50 °C to 70 °C), a small percentage of adipic acid was also formed.

10.2  Transition Metal Catalysts for Cyclohexane Oxidation

Multinuclear copper complexes receive a significant interest in terms of mimicking the action of particulate methane monooxygenase (pMMO), a multi-copper enzyme that catalyzes the oxidation of alkanes. The use of the tetranuclear μ-oxido complex [O⊂Cu4(tea)4(BOH)4][BF4]2 (44, H3tea=triethanolamine), the trinuclear [Cu3(H2tea)2(4-OC6H4COO)2(H2O)]·4H2O (45), the

193

194

10  Catalytic Cyclohexane Oxyfunctionalization

dinuclear [Cu2(H2tea)2(C6H5COO)2]·2H2O (46), and the coordination polymer [Cu2(H2tea)2{μC6H4(COO)2-1,4}]n·2nH2O (47) [49] represents a pioneering work in this area. These multicopper compounds produce highly selective and active catalytic systems for the oxidation of alkanes (namely cyclohexane, methane, and ethane) in acetonitrile medium at room temperature [49]. The highest activity was observed for the tetra- and trinuclear complexes 44 and 45 with overall yields (based on cyclohexane) up to 32% for a typical acid (promoter)-to-catalyst molar ratio of 10:1. TON values of c. 360–380 can be obtained [49]. Thus, the multicopper catalysts display a higher catalytic activity than the mononuclear ones [49–51].

Copper(I) and copper(II) pyrazolate compounds have also been explored as catalysts. The dinuclear cyclic Cu(I) complex [Cu2(µ-N,N-3,5-(NO2)2pz)2(PPh3)2] (48) was formed with 3,5-NO2-pyrazole ligands, while a spontaneous rearrangement to the hexanuclear cyclic [trans-Cu6(µ-OH)6(µ-3,5(CF3)2pz)6] (49) was observed in the case of 3,5-bis(trifluoromethyl)-pyrazole [52].

10.2  Transition Metal Catalysts for Cyclohexane Oxidation

These compounds (48 and 49) were successfully applied as catalysts for the MW-assisted neat oxidation (with TBHP) of cyclohexane to the cyclohexanol and cyclohexanone mixture. Relatively high yields (up to 58%) in rather short reaction times (30 min) were achieved with 49. The yields obtained by this catalytic system are higher than those reported (although under considerably different conditions) for copper complexes with related pyrazole ligands [53, 54]. Moreover, this catalytic process (TBHP/MW) is very fast and solvent-free in addition to using a greener energy source [52]. In the three isomeric trinuclear copper complexes [Cu3(L)2(MeOH)4] (50 or 51) and [Cu3(L)2(MeOH)2]·2MeOH (52), where H3L = 3,5-di-tert-butyl-2-hydroxybenzylidene-2-hydroxybenzohydrazide, the L3- ligand exhibits trianionic pentadentate 1κO,O’,N:2κN’,O’’ chelation modes [55]. These play an efficient role as catalyst precursors for the peroxidative oxidation of cyclohexane to cyclohexyl hydroperoxide, cyclohexanol, and cyclohexanone using H2O2 as oxidant in aqueous-acetonitrile medium, at room temperature. Overall yields up to 31% (based on the cyclohexane) and TONs up to 1.55 × 103 are obtained after six hours in the presence of PCA as a promoter [55].

195

196

10  Catalytic Cyclohexane Oxyfunctionalization

The catalytic oxidation of cyclohexane using H2O2 as an oxidant in the ionic liquid (IL) 1-butyl3-methylimidazolium hexafluorophosphate [bmim][PF6] and the tetracopper(II) complex [(CuL)2(μ4-O,O’,O’’,O’’’-CDC)]2·2H2O (53) [HL = 2-(2-pyridylmethyleneamino)benzenesulfonic acid, CDC=cyclohexane-1,4-dicarboxylate] as a catalyst was also reported [56].

10.2  Transition Metal Catalysts for Cyclohexane Oxidation

By using the [bmim][PF6] IL instead of the commonly used organic solvent acetonitrile, this system shows significant improvements on the catalytic performance, reaction time, product yield (up to 36%), TON (up to 529), selectivity toward cyclohexanone, and ease of recycling (negligible loss in activity after three consecutive runs) [56]. The possible interactions between ILs and the catalyst, substrate, oxidant, and even reaction intermediates can allow ILs to act as multi-functional solvents for such catalytic oxidation reactions. The heterometallic complexes [Cu2Fe2(HL1)2(H2L1)2]·10DMSO (54) and [Cu2Fe2 (HL2)2(H2L2)2]· 2DMF (55) where HL1 and HL2 are polydentate Schiff bases formed in situ via condensation of salicylaldehyde (in the case of 54) or 5-bromo-salicylaldehyde (in the case of 55) with tris(hydroxymethyl) aminomethane. Both 54 and 55 exhibit exceptionally high catalytic activity in the oxidation of ­cyclohexane with H2O2 under mild conditions (in CH3CN, 5 h, r.t.) with the best observed yield of 36% (TON = 596) and 44% (TON = 1.1 × 103) for 54 and 55, respectively [57].

197

198

10  Catalytic Cyclohexane Oxyfunctionalization

A cage Cu,Na-silsesquioxane, [Cu9Na6Si20O42Ph20(EtOH)7.5(H2O)2] (56) (Figure 10.2), exhibits a good catalytic activity toward cyclohexane oxidation with H2O2 in acetonitrile with a maximum products yield of 21% [58]. Under an 18O2 atmosphere, the oxygenation of cyclohexane exhibits the highest incorporation of 18O into cyclohexyl hydroperoxide at the beginning of the reaction (44% in 10 min) and with time the value drops to 24% in 3 h. This is probably due to the decrease of percentage of 18O2 in the atmosphere due to formation of 16O2 as a reaction by-product. The amounts of 18O in the ketone are much lower (18% av.) before the addition of PPh3 and no 18O was detected after addition of PPh3. H216O present in a large excess in the reaction mixture possibly exchanges its oxygen with 18O labeled cyclohexanone formed during the reaction [58].

10.2.5  Molybdenum Catalysts The cis-dioxidomolybdenum(VI) complexes [MoO2(L1)] (57), [MoO2(L2)]·MeOH (58), and [MoO2(L3)] (59) obtained from the three different aroylhydrazone Schiff bases H2L1 = 2,3-dihydroxy benzylidene-2-hydroxybenzohydrazide, H3L2 = 2,3-dihydroxybenzylidene-benzohydrazide, and H2L3 = (3,5-di-tert-butyl-2-hydroxybenzylidene)-2-hydroxybenzohydrazide, respectively, exhibit good catalytic activities toward cyclohexane oxidation in CH3CN, in the ionic liquid [bmim][PF6], in supercritical carbon dioxide (scCO2), and in scCO2/[bmim][PF6] mixed solvent [59]. The catalysts show higher catalytic activities in the ionic liquid than in CH3CN, and exhibit very high selectivity toward cyclohexanol in the scCO2 medium. A maximum TON value of 26 is obtained with 57 in the scCO2 medium with selectivity of 96% toward cyclohexanol. The catalytic activity follows the order 57>58>59 in each medium, which corelates with the steric hindrance of groups present in the aldehyde moiety of the catalysts. An advantage of this catalytic system is the possibility of catalyst recycling (i.e. the catalyst can be recycled up to three cycles with full preservation of their activity in the scCO2 and scCO2/IL media) [59].

Figure 10.2  X-ray crystal structure of 56. Cyan balls represent Cu atoms and violet balls represent Na atoms. H-atoms and solvent molecules are omitted for clarity. Reproduced with permission from Ref [58] / American Chemical Society.

10.2  Transition Metal Catalysts for Cyclohexane Oxidation

10.2.6  Rhenium Catalysts Rhenium complexes with a diversity of ligands can exhibit good catalytic activities toward cyclohexane oxidation. Examples include the benzoylhydrazide complexes [ReCl2{η2-N,O-N2C(O)Ph}(PPh3)2] (60), [ReCl2{N2C(O)Ph}(Hpz)2(PPh3)] (61), [ReClF{N2C(O)Ph}(Hpz)2(PPh3)] (62) (where Hpz = pyrazole), and the parent [ReOCl3(PPh3)2] (63), as well as the tris(pyrazolyl)methane [HC(pz)3] complex [ReCl3{HC(pz)3}] (64) and [ReO3{SO3C(pz)3}] (65), where SO3C(pz)3 = tris(pyrazolyl)methane)sulfonate [60]. At room temperature and with a six hour reaction time, a maximum total TON value of 243 is obtained with 62 in acetonitrile medium using aqueous H2O2 as the oxidant in the absence of an acid promoter. However, in the presence of acid, the catalytic activity of 65 is higher than that of 63. A total TON value of 270 is observed in the presence of HNO3 as a promoter. Methyltrioxorhenium(VII) [MeReO3] (MTO, 66) is also a catalyst for oxidation of alkanes. A detailed DFT study on the [MeReO3]/H2O2/H2O-CH3CN system established the most plausible pathway. It proceeds through the monoperoxo complex [MeReO2(O2)(H2O)], followed by formation of an H2O2 adduct that undergoes water-assisted H+-transfer from H2O2 to the peroxo ligand that,

Scheme 10.4  Catalytic cycle for the formation of free HOO• and HO• radicals by 66. Reproduced with permission from Ref [61] / American Chemical Society.

199

200

10  Catalytic Cyclohexane Oxyfunctionalization

upon oxidation by the metal, generates HOO• (Scheme 10.4). The HO• radical formation takes place by the water-assisted H+-transfer and O–OH bond reductive cleavage using the reduced ReVI complex [MeReO2(OOH)(H2O)]. The HO• radical then reacts with the alkane [61].

10.2.7  Gold Catalysts The commercial Au(I) and Au(III) complexes 67–71 supported on various carbon materials [activated carbon, carbon xerogel, and carbon nanotubes (CNT)] with different surface treatments of original forms, oxidized (-ox), and oxidized with nitric acid followed by subsequent treatment with sodium hydroxide (-ox-Na) exhibit a good catalytic activity toward the oxidation of cyclohexane to cyclohexanol and cyclohexanone using TBHP as oxidant under mild conditions (CH3CN, 6 h, 20  °C). 71/CNT-ox-Na material showed the best results with an overall yield up to c. 26% (TON=653) after six hours of reaction time [62].

Three Au(III) tris(pyrazol-1-yl)methane complexes (72–74) supported on carbon materials (activated carbon, carbon xerogel, and CNT) were also studied for the oxidation of cyclohexane to cyclohexanol and cyclohexanone using aqueous H2O2 as an oxidant under mild conditions [63]. 73 is found to be the most effective catalyst, leading to a maximum yield of 16% (TON = 8 × 102) under heterogeneous conditions (on CNT-ox-Na). Under homogeneous conditions, comparatively lower yield (10%) and TON value (3 × 102) are obtained. Immobilization of the complexes on a CNT-ox-Na support formed the most active catalytic system [63]. Various gold nanoparticle/carbon catalysts were made by loading 1 wt.% gold on various carbon materials (activated carbon, polymer based carbon xerogels, multi-walled CNTs, nanodiamonds, microdiamonds, graphite, and silicon carbide) using a sol immobilisation or colloidal method (COL) and double impregnation methods. These Au/carbon materials play as catalysts for the oxidation of cyclohexane to cyclohexanol and cyclohexanone under mild conditions using aqueous

10.3  Mechanisms

H2O2 as an oxidant [64]. Au/CNT-COL showed the highest activity with a TON value of c. 171 and a yield of 3.6% in 6 h, which is comparable to the industrial process. Relatively low catalyst loads (Au catalyst to substrate molar ratio always lower than 1 × 10−3) under ambient temperature and catalyst recycling are advantages of this catalytic system. The differences in catalytic performance of these Au/carbon catalytic systems can be corelated in terms of gold nanoparticle size. The catalytic activity of Au nanoparticles supported on different supports (Al2O3, Fe2O3, ZnO, and TiO2) was tested for the peroxidative oxidation of cyclohexane to cyclohexanol and cyclohexanone with H2O2 or TBHP as oxidant under mild conditions (60 ºC). The best catalytic activity was found with Au on Fe2O3 (total yield of 13.5% in four hours). The catalyst maintained almost the same level of activity up to three cycles with no significant leaching [65].

10.3  Mechanisms Metal complex-catalyzed oxidation of cyclohexane with hydrogen peroxide or an alkylhydroperoxide (ROOH) is usually believed to proceed via formation of both carbon- and oxygen-centred free radicals [11, 15–17, 61, 66–71]. The formation of free radicals has been established by the combination of experimental data (including analysis of various selectivity parameters, tests with radical traps, and 18O-labeled essays) and theoretical DFT calculations. The proposed mechanisms for various Mn+1/n (e.g. V, Re, Fe, Cu) systems are given for the case of H2O2 as oxidant and cyclohexane as substrate. Metal-catalyzed decomposition of H2O2 leads to the hydroxyl radical (HO•), which acts as H-abstractor from the alkane. The formation of the hydroxyl radical (the first two overall reactions shown) is presented in detail (Scheme 10.4) for the case of a Re(VII/VI) catalytic system. Mn+1 + H2O2 → HOO• + H+ + Mn+ Mn+ + H2O2 → HO• + Mn+1 + HO− HO• + CyH → H2O + Cy• Cy• + O2 → CyOO• CyOO• + H2O2 → CyOOH + HOO• CyOOH + Mn+ → CyO• + Mn+1 + HO− CyOOH + Mn+1 → CyOO• + H+ + Mn+ CyO• + CyH → CyOH + Cy• 2CyOO• → CyOH + Cy−H = O + O2 In such a mechanism, pyrazine carboxylic acid (Scheme 10.2) [11, 15, 22, 23] or water [16, 61] can play a promoting role by assisting proton-transfer steps, namely in the stage of generation of the hydroxyl radical from H2O2. Examples of the possible water-assisted H+-transfer 6-membered transition states (TSs) are indicated in Scheme 10.5.

201

202

10  Catalytic Cyclohexane Oxyfunctionalization

Scheme 10.5  Water-assisted H+-transfer TSs from coordinated H2O2 to an oxido ligand (i) and from a coordinated hydroperoxido to an hydroxido ligand (ii).

The involvement of free radicals is typically associated to low bond-, regio-, and stereo-selectivities. However, the selectivities can be improved if the mechanism occurs via metal-based oxidants. A recent and detailed example concerns the cyclohexane oxidation to cyclohexanol and cyclohexanone with m-chloroperoxybenzoic acid and the heterometalic complex 35 (see section 10.2.3 Cobalt Catalysts) in which the alkane C-H bond attacking species is a peroxybenzoate complex of cobalt (75) as the model species [44].

10.4  Final Comments A variety of transition metal complexes (or composites) that catalyze the peroxidative oxidation of cyclohexane to cyclohexanol and cyclohexanone under mild conditions is presented herein, focusing on examples usually investigated in our laboratories. Particularly active are: (i) oxidovanadium(IV and V) complexes containing azine fragment (C-N=N-C) ligands, carbohydrazone ligands or pyrazole based ligands; (ii) iron(III) complexes with carboxylates, Schiff base aminoalcohol, carbohydrazone or scorpionate ligands; (iii) cobalt(II) complexes with silyl-containing Schiff bases or pyrazolyl ligands; (iv) copper(II) complexes with hydrazone Schiff base or pyrazole based ligands; (v) molybdenum(VI) complexes with hydrazone Schiff base ligands; (vi) rhenium complexes with pyrazole ligands; and (vii) gold nanocomposites. The promoting effect of co-catalysts (e.g PCA, HNO3, or even water) has also been observed. Growing attention is being paid toward the development of sustainable catalytic systems (e.g. in the use of low energy microwave irradiation or ultrasounds as a alternative source of energy, the application of an ionic liquid or scCO2 instead of an organic solvent, and recycling of the metal catalysts). General mechanisms of cyclohaxane oxidation are also outlined with their implications for selectivity depending markedly on the type of oxidant and on the type of mechanism involved. Although promising development of metal-catalyzed cyclohexane oxyfunctionalization processes has already evolved, various challenges still need to be overcome to develop more

References

sustainable green catalytic processes. These include the preparation of more effective and selective catalysts, the development of greener methods, reuse and recyclability of the catalysts, and further exploration of kinetic and theoretical studies for deeper mechanistic insights and better catalyst design. A promising approach concerns the attempts to the establishment of catalytic systems based on a non-transitional metal [e.g. Al(lII)] catalyst with a low environmental impact. This has been developed in our laboratories and supported by DFT studies [72–76], but is not addressed in this chapter.

Acknowledgments The authors are grateful: (i) to the Fundação para a Ciência e a Tecnologia (FCT) for the Centro de Químical Estrutural UIDB/00100/2020 and UIDP/00100/2020 projects, and for the Institute of Molecular Sciences LA/P/0056/2020 project, and (ii) to the Instituto Politécnico de Lisboa the IPL/ IDI&CA2023/SMARTCAT_ISEL and 2022.02069.PTDC projects.

References 1 Pombeiro, A.J.L. and Guedes da Silva, M.F.C. (eds) (2019). Alkane Functionalization. Hoboken: John Wiley and Sons Ltd. 2 Labinger, J.A. and Bercaw, J.E. (2002). Nature 417: 507–514. 3 Newhouse, T. and Baran, P.S. (2011). Angew. Chem. Int. Ed. 50: 3362–3374. 4 Neuenschwander, U., Turra, N., Aellig, C. et al. (2010). Chimia 64: 225–230. 5 Shilov, A.E. and Shul’pin, G.B. (1997). Chem. Rev. 97: 2879–2932. 6 McKetta, J.J. and Cunningham, W.A. (1977). Encyclopedia of Chemical Processing and Design. New York: Marcel Dekker. 7 Raja, R. (2007). Chapter 37. Strategically designed single-site heterogeneous catalysts for clean technology, green chemistry and sustainable development. In: Turning Points in Solid-State, Materials and Surface Science, chapter 37. (eds. K.D.M. Harris and P.P. Edwards), 623–638. London: Royal Society of Chemistry. 8 Jevtic, R., Ramachandran, P.A., and Dudukovic, M.P. (2009). Ind. Eng. Chem. Res. 48: 7986–7993. 9 Li, H., She, Y., and Wang, T. (2012). Front. Chem. Sci. Eng. 6: 356–368. 10 Khirsariya, P. and Mewada, R. (2014). Ijedr 2: 3911–3914. 11 Shul’pin, G.B. (2009). Mini-Rev. Org. Chem. 6: 95–104. 12 Crabtree, R.H. (2001). J. Chem. Soc., Dalton Trans 2437–2450. 13 Sutradhar, M., Shvydkiy, N.V., Guedes da Silva, M.F.C. et al. (2013). Dalton Trans. 42: 11791–11803. 14 Sutradhar, M., Kirillova, M.V., Guedes da Silva, M.F.C. et al. (2012). Inorg. Chem. 51: 11229–11231. 15 Shul’pin, G.B., Kozlov, Y.N., Nizova, G.V. et al. (2001). J. Chem. Soc. Perkin Trans. 2: 1351–1371. 16 Kirillova, M.V., Kuznetsov, M.L., Romakh, V.B. et al. (2009). J. Catal. 267: 140–157. 17 Shul’pin, G.B. (2013). Dalton Trans 42: 12794–12818. 18 Kuznetsov, M.L. and Pombeiro, A.J.L. (2009). Inorg. Chem. 48: 307–318. 19 Reis, P.M., Silva, J.A.L., Frausto da Silva, J.J.R., and Pombeiro, A.J.L. (2000). Chem. Commun. 1845–1846. 20 Gupta, S., Kirillova, M.V., Guedes da Silva, M.F.C. et al. (2013). Inorg. Chem. 52: 8601–8611.

203

204

10  Catalytic Cyclohexane Oxyfunctionalization

21 Reis, P.M., Silva, J.A.L., Fraústo da Silva, J.J.R., and Pombeiro, A.J.L. (2000). Chem. Commun. 1845–1846. 22 Shul’pin, G.B., Beller, M., and Bolm, C. (eds.) (2004). Transition Metals for Organic Synthesis, vol. 2, 2e, 215. New York: Wiley-VCH. 23 Kirillova, M.V., Kuznetsov, M.L., Kozlov, Y.N. et al. (2011). ACS Catal. 1: 1511–1520. 24 Dragancea, D., Talmaci, N., Shova, S. et al. (2016). Inorg. Chem. 55: 9187–9203. 25 Süss-Fink, G., Cuervo, L.G., Therrien, B. et al. (2004). Inorg. Chim. Acta 357: 475–484. 26 Silva, T.F.S., Luzyanin, K.V., Kirillova, M.V. et al. (2010). Adv. Synth. Catal. 352: 171–187. 27 Silva, T.F.S., Mac Leod, T.C.O., Martins, L.M.D.R.S. et al. (2013). J. Mol. Catal. A: Chemical 367: 52–60. 28 Silva, T.F.S., Alegria, E.C.B.A., Martins, L.M.D.R.S., and Pombeiro, A.J.L. (2008). Adv., Synth. Cat. 350: 706–716. 29 Milunovic, M.N.M., Martins, L.M.D.R.S., Alegria, E.C.B.A. et al. (2013). Dalton Trans. 42: 14388–14401. 30 Nesterov, D.S., Nesterova, O.V., Guedes da Silva, M.F.C., and Pombeiro, A.J.L. (2015). Catal. Sci. Technol. 5: 1801–1812. 31 Bilyachenko, A.N., Levitsky, M.M., Yalymov, A.I. et al. (2016). RSC Adv. 6: 48165–48180. 32 Vinogradov, M.M., Kozlov, Y.N., Bilyachenko, A.N. et al. (2015). New J. Chem. 39: 187–199. 33 Vinogradov, M.M., Kozlov, Y.N., Nesterov, D.S. et al. (2014). Catal. Sci. Technol. 4: 3214–3226. 34 Roy Barman, T., Sutradhar, M., Alegria, E.C.B.A. et al. (2020). Catalysts 10: 1175. 35 Ribeiro, A.P.C., Martins, L.M.D.R.S., Alegria, E.C.B.A. et al. (2017). Catalysts 7: 230. 36 Martins, L.M.D.R.S., Almeida, M.P., Carabineiro, S.A.C. et al. (2013). ChemCatChem 5: 3847–3856. 37 Ribeiro, A.P.C., Martins, L.M.D.R.S., Carabineiro, S.A.C. et al. (2018). ChemCatChem 10: 1821–1828. 38 Ribeiro, A.P.C., Martins, L.M.D.R.S., and Pombeiro, A.J.L. (2017). Green Chem 19: 1499–1501. 39 Nesterova, O.V., Nesterov, D.S., Vranovičová, B. et al. (2018). Dalton Trans 47: 10941–10952. 40 Gomes, C.A., Lube, L.M., Fernandes, C. et al. (2017). New J. Chem. 41: 11498–11502. 41 Zaltariov, M.-F., Vieru, V., Zalibera, M. et al. (2017). Eur. J. Inorg. Chem. 2017 (37): 4324–4332. 42 Silva, T.F.S., Martins, L.M.D.R.S., Guedes da Silva, M.F.C. et al. (2014). Chem. Asian J. 9: 1132–1143. 43 Nesterov, D.S., Chygorin, E.N., Kokozay, V.N. et al. (2012). Inorg. Chem. 51: 9110–9122. 44 Nesterova, O.V., Kuznetsov, M.L., Pombeiro, A.J.L. et al. (2022). Catal. Sci. Technol. 12: 282–299. 45 Sutradhar, M., Alegria, E.C.B.A., Guedes da Silva, M.F.C. et al. (2016). Molecules 21: 425. 46 Palmucci, J., Mahmudov, K.T., Guedes da Silva, M.F.C. et al. (2015). RSC Adv. 5: 84142–84152. 47 Saha, M., Vyas, K.M., Martins, L.M.D.R.S. et al. (2017). Polyhedron 132: 53–63. 48 Detoni, C., Carvalho, N.M.F., Aranda, D.A.G. et al. (2009). Appl. Catal. A: General 365: 281–286. 49 Kirillov, A.M., Kopylovich, M.N., Kirillova, M.V. et al. (2005). Angew. Chem. Int. Ed. 44: 4345–4349. 50 Kirillov, A.M., Kirillova, M.V., and Pombeiro, A.J.L. (2012). Coord. Chem. Rev. 256: 2741–2759. 51 Adams, R.D., Rassolov, V., and Wong, Y.O. (2016). Angew. Chem. Int. Ed. 55: 1324–1327. 52 Galassi, R., Simon, O.C., Burini, A. et al. (2017). Polyhedron 134: 143–152. 53 Silva, T.F.S., Silva, M.F.C.G., Mishra, G.S. et al. (2011). J. Organomet. Chem. 696: 1310–1318. 54 Contaldi, S., Di Nicola, C., Garau, F. et al. (2009). Dalton Trans. 4928–4941. 55 Sutradhar, M., Martins, L.M.D.R.S., Guedes da Silva, M.F.C. et al. (2015). Eur. J. Inorg. Chem. 2015 (23): 3959–3969. 56 Ribeiro, A.P.C., Martins, L.M.D.R.S., Hazra, S., and Pombeiro, A.J.L. (2015). C. R. Chim. 18: 758. 57 Nesterova, O.V., Chygorin, E.N., Kokozay, V.N. et al. (2013). Dalton Trans 42: 16909–16919.

References

58 Astakhov, G.S., Bilyachenko, A.N., Korlyukov, A.A. et al. (2018). Inorg. Chem. 57: 11524–11529. 59 Sutradhar, M., Ribeiro, A.P.C., Guedes da Silva, M.F.C. et al. (2020). Mol. Catal. 482: 100356. 60 Alegria, E.C.B.A., Kirillova, M.V., Martins, L.M.D.R.S., and Pombeiro, A.J.L. (2007). Appl. Catal. A: Gen. 317: 43–52. 61 Kuznetsov, M.L. and Pombeiro, A.J.L. (2009). Inorg. Chem. 48: 307–318. 62 Carabineiro, S.A.C., Martins, L.M.D.R.S., Pombeiro, A.J.L., and Figueiredo, J.L. (2018). ChemCatChem 10: 1804–1813. 63 Almeida, M.P., Martins, L.M.D.R.S., Carabineiro, S.A.C. et al. (2013). Catal. Sci. Technol. 3: 3056–3069. 64 Carabineiro, S.A.C., Martins, L.M.D.R.S., Avalos-Borja, M. et al. (2013). Appl. Catal. A: Gen. 467: 279–290. 65 Martins, L.M.D.R.S., Carabineiro, S.A.C., Wang, J. et al. (2017). ChemCatChem 9: 1211–1221. 66 Carabineiro, S.A.C., Bastos, S.S.T., Orfao, J.J.M. et al. (2010). Appl. Catal. A: Gen. 381: 150–160. 67 Carabineiro, S.A.C., Machado, B.F., Bacsa, R.R. et al. (2010). J. Catal. 273: 191–198. 68 Mishra, G.S., Silva, T.F.S., Martins, L.M.D.R.S., and Pombeiro, A.J.L. (2009). Pure Appl. Chem. 81: 1217–1227. 69 Di Nicola, C., Garau, F., Karabach, Y.Y. et al. (2009). Eur. J. Inorg. Chem. 2009 (5): 666–676. 70 Kopylovich, M.N., Nunes, A.C.C., Mahmudov, K.T. et al. (2011). Dalton Trans 40: 2822–2836. 71 Shul’pin, G.B. (2019). Alkane-oxidizing systems based on metal complexes. Radical versus nonradical mechanisms. In: Alkane Functionalization, ch 3 (eds. A.J.L. Pombeiro and M.F.C. Guedes da Silva), 47–70. Hoboken: John Wiley and Sons, Ltd. 72 Rocha, B.G.M., Kuznetsov, M.L., Kozlov, Y.N. et al. (2015). Catal. Sci. Technol. 5: 2174–2187. 73 Novikov, A.S., Kuznetsov, M.L., Rocha, B.G.M. et al. (2016). Catal. Sci. Technol. 6: 1343–1356. 74 Kuznetsov, M.L., Rocha, B.G.M., Pombeiro, A.J.L., and Shul’pin, G.B. (2015). ACS Catal. 5: 3823–3835. 75 Novikov, A.S., Kuznetsov, M.L., Pombeiro, A.J.L. et al. (2013). ACS Catal. 3: 1195–1208. 76 Kuznetsov, M.L., Kozlov, Y.N., Mandelli, D. et al. (2011). Inorg. Chem. 50: 3996–4005.

205

207

Part III Carbon-based Catalysis

209

11 Carbon-based Catalysts for Sustainable Chemical Processes Katarzyna Morawa Eblagon1,2, Raquel P. Rocha1,2, M. Fernando R. Pereira1,2, and José Luís Figueiredo1,2,3 1 Laboratory of Separation and Reaction Engineering – Laboratory of Catalysis and Materials (LSRE–LCM), Faculty of Engineering, University of Porto, Portugal 2 ALiCE - Associate Laboratory in Chemical Engineering, Faculty of Engineering, University of Porto, Portugal 3 Academia das Ciências de Lisboa

11.1  Introduction Being a green technology, catalysis is at the very heart of sustainability, minimizing the production of wastes and maximizing atom efficiency in chemical processes. In addition, catalysts are used in various technologies for pollution abatement and control, providing for cleaner air and water. In particular, heterogeneous catalysis is often preferred, due to the easy separation of the solid catalyst from the reaction medium, allowing for continuous operation. Recently, the potential of carbon materials in catalysis has been realized, as they may offer a more sustainable alternative to the conventional metal and metal oxide catalysts used in industry by replacing critical materials (noble metals and rare earth elements) in relevant catalytic processes [1]. In this chapter, we will review the methods used for the synthesis of nanostructured carbons, and their use as catalysts and catalyst supports will be discussed in the context of environmental protection and the sustainable production of chemicals and fuels from biomass.

11.1.1  Nanostructured Carbon Materials Carbon materials used in catalysis include activated carbons, carbon gels, templated carbons, bulk graphite, carbon blacks, carbon nanotubes and nanofibers, and graphene-based materials. All of them present a graphenic structure, and it is possible to tune their properties (texture and surface chemistry) by chemical and thermal post-treatments, or by adequate synthesis strategies [2]. Activated carbons (ACs) exhibit a hierarchical pore structure, consisting of micro (< 2 nm), meso (2–50 nm), and macropores (> 50 nm), which is well suited for adsorption; the meso- and ­macropores provide access to the solutes to the large micropore volume (up to 0.5 cm3/g) where adsorption mainly occurs. However, a large mesopore surface area is preferred for catalysis in order to maximize the concentration of active sites while minimizing diffusion limitations and catalyst deactivation. Another disadvantage of microporous ACs is that they may contain large Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 1, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

210

11  Carbon-based Catalysts for Sustainable Chemical Processes

amounts of mineral impurities (up to 20 wt%, depending on the precursor) that may exhibit catalytic activity on their own, promoting undesirable reactions [3]. Carbon gels and templated carbons are two of the most interesting types of mesoporous materials for catalytic applications since they are free from transition metal contaminations. Carbon gels (aerogels, xerogels and cryogels, according to the drying method) are produced by carbonization of organic gels obtained by polycondensation of hydroxybenzenes with aldehydes [4]. Their textures depend mainly on the synthesis pH and dilution ratio (solvents to reactants ratio), leading to materials with a wide range of pore sizes (from 2 nm to 1000 nm) [5–7]. Ordered mesoporous carbons (OMCs) can be prepared with the aid of templates by two different methods. In nanocasting (or hard-templating, or exotemplating), a mesoporous silica is impregnated with a carbon precursor, subsequently carbonized inside the pores; then, the silica template is dissolved [8, 9]. Pore sizes ranging from 3 to 30 nm can be obtained, depending on the template used. Another route for preparing OMCs (known as soft-templating, or endotemplating) consists in the self-assembly of thermally decomposable surfactants (e.g. block copolymers) and carbon precursors (e.g. thermosetting polymers), followed by carbonization. In this case, the carbon material forms around the template, subsequently removed by heat treatment. This method involves fewer steps and avoids using strong acids (such as HF) to remove the template [10]. Carbon nanofibers (CNFs) and carbon nanotubes (CNTs) are filamentous carbons; the individual filaments (with diameters lower than a few hundred nanometers) are mostly non-porous, but they tend to agglomerate, forming bundles or aggregates, with relatively large surface areas (up to 300 m2/g). CNTs present a tubular morphology formed by concentric graphene layers, whereas the graphene layers are at an angle with the fiber axis in the case of CNFs [11]. Graphene-based materials [graphene oxide GO, and reduced graphene oxide, (rGO)] have been successfully used in composites with TiO2 for the photocatalytic degradation of organic pollutants [12–14], whereas graphene-derived catalytic membranes were used for the removal of aqueous organic contaminants by photocatalysis, electrocatalysis, and persulfate/peroxymonosulfate activation catalysis, demonstrating their potential for water treatment [15]. GO can be obtained from natural graphite by oxidation and exfoliation, most frequently by the modified Hummers method using sulfuric acid and potassium permanganate [16, 17]. Next, rGO can be obtained by different reduction methods. Hydrothermal carbonization (HTC) of biomass-derived precursors, such as carbohydrates, can produce porous carbons, but these materials have poor mechanical resistance. More robust carbon hybrid structures are obtained in the presence of GO or oxidized CNTs, when the acidic surface groups promoting dehydration/condensation reactions of the carbohydrates, while providing a scaffold for the growth of carbon xerogels [18, 19]. The nature, concentration, and accessibility of the active sites are the major factors that determine catalyst performance; therefore, both the physical (textural) and chemical (surface) properties of carbon materials are important in the context of this review.

11.1.2  Carbon Surface Chemistry Active sites are found essentially at the edges of the graphene layers and at basal plane defects, where the unsaturated carbon atoms may react with different compounds forming a variety of surface functional groups. Most relevant in the present context are those containing oxygen, nitrogen, and sulfur, shown schematically in Figure 11.1, which also includes data for their identification and quantification by X-ray photoelectron spectroscopy (XPS) and temperature-programmed desorption (TPD); the CHNSO composition can be determined by elemental analysis (EA) [20].

11.1  Introduction

Figure 11.1  Oxygen, nitrogen and sulfur groups on the surface of carbon materials and data for their identification/quantification by temperature-programmed desorption (TPD) and X-ray photoelectron spectroscopy (XPS). Reprinted with permission from Ref [20].

Further details about these techniques can be found in the literature [21–24]. In addition, the point of zero charge (pHPZC) and/or the isoelectric point (pHIEP) are quite useful parameters, providing a global description of the surface chemistry of the carbon material [25]. Most of the oxygenated groups are acidic (carboxylic acids and anhydrides, lactones, lactols, and phenols), whereas carbonyl and ether groups are neutral or may form surface basic structures such as quinone, chromene, and pyrone groups [25]. Nitrogen groups include pyridine (N-6), pyrrole or pyridone (N-5), oxidized nitrogen (NX) at the edges, and quaternary nitrogen (N-Q) incorporated into the graphene structure [26]. The nitrogen atoms provide additional electrons, enhancing the catalytic activity in oxidation reactions. These groups, together with the π electron system of the basal planes, can play the role of active sites in various reactions or may be involved in the formation of active species, such as free radicals. Small amounts of oxygenated groups are always present on the surface of carbon materials, being formed spontaneously by exposure to the atmosphere. Further functionalization can be achieved by different oxidative treatments, in the gas phase (with oxygen, ozone, or nitrogen oxides) or in the liquid phase (with nitric acid, hydrogen peroxide, or ammonium persulphate). These treatments are rather unselective, forming various types of functional groups, but subsequent thermal treatments under an inert atmosphere can then be used to remove selectively the most labile groups. By adjusting the temperature and duration of the treatment, samples with different amounts of the desired groups can be prepared from the original oxidized carbon, with minimal changes in the textural properties [3, 27].

211

212

11  Carbon-based Catalysts for Sustainable Chemical Processes

The introduction of nitrogen groups enhances both surface basicity and the catalytic activity of carbon materials in oxidation reactions [28, 29]. Surface nitrogen groups can be introduced by treating the carbon materials with nitrogen-containing compounds (e.g. ammonia at high temperatures) [28, 30, 31], or by reaction between the surface oxygen groups with amine compounds [32]. Another approach consists of the carbonization of nitrogen containing polymers [33]. Nitrogendoped carbon gels can be synthesized by the addition of a nitrogen precursor during the sol-gel polycondensation [34, 35] in particular, carbon xerogels with large mesoporous surface areas were obtained using urea or melamine [36]. On the other hand, nitrogen-doped OMCs can be obtained by nanocasting using polyacrylonitrile (PAN) as a C/N precursor [37]. Sulfur groups are also relevant in the context of the present review, in particular sulfonic acid groups. The most widely used method for the incorporation of such groups consists in treating the carbon material with concentrated sulfuric acid [38, 39], but alternative functionalization methodologies have been reported [40, 41]. A mechanothermal method for doping carbon materials with heteroatoms has been described, consisting in ball-milling a mixture of the carbon material with the heteroatom precursor, followed by heating under nitrogen up to the decomposition temperature of the precursor [42]. The method was used to incorporate N, S, P, and B onto CNTs, with appropriate precursors (melamine, sodium thiosulfate, sodium dihydrogen phosphate, and boric acid, respectively) [43].

11.2  Metal-free Carbon Catalysts for Environmental Applications Carbon materials are versatile catalysts, performing well in many reactions that are traditionally catalyzed by metals or transition metal oxides. In addition, upon functionalization with sulfonic acid groups, they can be used in acid catalysis, in such reactions as esterification, alkylation, acetalization, and acylation. On the other hand, carbon materials with basic surface properties can be used as catalysts for the elimination of organic pollutants in water and wastewaters by advanced oxidation processes (AOPs) [44, 45]. Advanced oxidation technologies are based on the action of extremely reactive species (namely free radicals, such as the hydroxyl radical, or other oxygenated species, such as singlet oxygen or the superoxide ion) capable of oxidizing organic compounds. These species can be formed as a result of chemical, photochemical, electrochemical or sonochemical processes. Oxidants such as oxygen, hydrogen peroxide, or ozone can be used to generate radicals, and the corresponding chemical processes are named wet oxidation, wet peroxide oxidation and ozonation, respectively. Ideally, these processes should lead to the complete mineralization of the organic pollutants, yielding carbon dioxide, water and inorganic ions, or at least converting them into less toxic compounds. Catalysts are used to increase conversions and to decrease the severity of the operating conditions, and carbon materials are particularly well suited for such applications.

11.2.1  Wet Air Oxidation and Ozonation with Carbon Catalysts CNTs are often used in catalytic AOP studies, due to their chemical inertness in alkaline and acidic media, easy functionalization, and absence of diffusion limitations; on the other hand, oxalic acid and phenol are most frequently used as model substrates. The main conclusions of a review of the role of surface functional groups on the performance of CNTs in catalytic wet air oxidation (CWAO) and catalytic ozonation (COZ) were the following: pristine CNTs exhibit adequate catalytic activity for the oxidation of oxalic acid by CWAO and COZ, but the oxidation of phenol is more difficult; the catalytic performance is improved by nitrogen doping, whereas the presence of acidic oxygen

11.2  Metal-free Carbon Catalysts for Environmental Applications

groups (carboxylic acids, phenols, anhydrides) decreases the catalytic activity; experiments in the presence of free-radical scavenger species show that the oxidation reactions occur mainly on the catalyst surface for CWAO and both in the liquid phase and on the catalyst surface for COZ, as illustrated in Figure 11.2 [20]. On the other hand, N-doped CNTs prepared by the ball-milling method were found to exhibit outstanding performances as catalysts for these AOPs: oxalic acid was completely mineralized in 5 min by CWAO (at 140 °C and 40 bar total pressure) and in 4 h by COZ (at room temperature and pressure), and the catalyst could be reused in cyclic experiments without loss of activity for CWAO, as a result of the stability of the nitrogen functional groups incorporated (N6, N5, and NQ) [46]. The promoting effect of nitrogen groups with unpaired electrons, such as the N6 groups, has been explained by the enhanced chemisorption of oxygen, which favors the surface oxidation pathway. Thus, the oxidation reaction can proceed on the catalyst surface, even in the absence of hydroxyl radicals in the liquid phase. Good performance in the CWAO of oxalic acid was also obtained with a honeycomb cordierite monolith coated with N-doped CNTs in continuous operation [47]. On the other hand, the influence of the surface chemistry of carbon xerogels on their ability to remove phenol from wastewaters by catalytic wet peroxide oxidation (CWPO) was studied [48].

Figure 11.2  Schematic representation of the reactions that might occur in the liquid phase and on the surface of carbon nanotubes (CNTs) during catalytic wet air oxidation (CWAO) and catalytic ozonation (COZ) experiments. from Ref [20] / MDPI / CC BY 4.0.

213

214

11  Carbon-based Catalysts for Sustainable Chemical Processes

The adsorption of phenol was improved in the presence of quaternary nitrogen in the carbon structure and hindered in the presence of carboxylic groups. Moreover, the activity of these catalysts in CWPO was governed by the total acidity/basicity of their surfaces and the type of functionalities present.

11.3  Carbon-based Catalysts for Sustainable Production of Chemicals and Fuels from Biomass The world is faced with serious climate changes driven by rising greenhouse gas emissions resul­ ting from burning fossil fuels. Technologies using alternative renewable carbon sources are thus urgently sought. Furthermore, easily available petroleum resources are becoming extremely scarce, which brings about supply shortages, increased oil prices, and energy insecurity worldwide. Limiting the economy’s dependence on oil by replacing it with renewable carbon sources can help fight climate change and mitigate the energy crisis. Moreover, the chemistry of renewables is not only able to replace petroleum-based products, but also opens an opportunity to develop a portfolio of completely new products that possess no equivalent in the current market [49].

11.3.1  Carbon Materials as Catalysts and Supports Catalysts active in petroleum-based technologies cannot be directly applied to biomass feedstocks. Instead, new materials and processes must be developed because the properties of biomass feedstocks are very different from those of petroleum substrates. For instance, biomass compounds are highly functionalized, have low volatility, and are highly reactive, unlike the oil-derived feedstocks containing low functionalities [50]. As a result, biomass valorization reactions are suitable for liquid-phase technologies, but they also often require hydrothermal conditions and the presence of strong acids or bases. Carbon materials are highly suitable for application with biomass-derived substrates because of their excellent chemical and hydrothermal stability, easily tailorable porous structure and surface chemistry. Carbons have much higher hydrothermal resistance than other commonly used materials in heterogeneous catalysis. For instance, zeolites or metal oxides suffer from irreversible deactivation in the presence of water [51]. Another property boosting carbon activity in hydrothermal biomass valorization reactions is their strong hydrophobic nature. Unlike zeolites or metal oxides, carbon materials are widely available and can be easily prepared, even from biomass resources [52]. For instance, very active carbon catalysts have been prepared from biomass-derived sugars or biomass waste [53–55]. Cellulose is the most abundant biopolymer on Earth, making it the most promising feedstock for chemicals, fuels, and materials [56–58]. Cellulose is a robust biopolymer containing strong intraand intermolecular hydrogen bonds and consists in many glucose units connected through β-1,4glycosidic bonds [59]. The chemical structure makes it resistant to chemical and biological treatments and highly insoluble in water or other common solvents [59, 60]. In this context, cost- and energyeffective glucose production via hydrolysis of cellulose derived from lignocellulosic biomass has been a bottleneck to the industrial success of biorefineries based on the sugar platform [61]. Recently, solid acids consisting of a framework (usually carbon) with surface acidic sites (Lewis or Brønsted acid sites) have gained much attention for the hydrolysis of cellulose. Many solid acid catalysts have been developed to the date, including sulfonated ACs or graphene oxide [59, 62, 63]. Conventional strong solid Brønsted acid catalysts, including niobic acid, H-mordenite, Nafion, or

11.3  Carbon-based Catalysts for Sustainable Production of Chemicals and Fuels from Biomass

Amberlyst-15 cannot hydrolyze crystalline cellulose. However, high yields of soluble sugars from crystalline cellulose were obtained using amorphous carbons consisting of flexible polycyclic carbon sheets with –SO3H, –COOH, and phenolic hydroxyl (–OH) groups in a three-dimensional network [64]. The biomass-derived catalyst was obtained by carbonization of cellulose, followed by treatment of the resulting carbon char with sulfuric acid. Outstanding catalytic performance was attributed to the strong ability of this carbon char to adsorb cellulose by forming hydrogen bonds with single bond OH groups of cellulose. Simultaneously, neighbouring Brønsted acid functionalities (–SO3H groups) facilitated the hydrolysis of cellulose [64]. On the other hand, Foo and Sievers found that the weakly acidic –OH functionalities, not only take part in adsorption of cellulose but also catalyze its hydrolysis [65]. The authors reported that the activity of phenolic functionalities depended on the synergistic effects between the functional groups and the defect sites on the carbon, which forced the close contact between cellulose and catalytic sites. Shen et al. studied novel carbon solid acid catalysts derived from the co-carbonisation of starch and polyvinylchloride in the hydrolysis of cellulose to reductive sugars. Their results showed that the interaction between the carbon catalyst and cellulose can be strengthened by the presence of chlorine groups on the carbon material. Chlorine atoms are electron donors and interact with protons from OH groups. These OH groups were forming numerous hydrogen bonds in cellulose. As a result of this interaction, numerous intra- and inter-molecular hydrogen bonds of cellulose were broken, exposing the glycosidic bond and increasing hydrolysis yield [66]. Our group studied the influence of surface chemistry of carbon xerogels (CX) on the hydrolysis of glycosidic linkages in cellobiose, which is a water-soluble model molecule of cellulose [67]. Various solid acid or basic catalysts were prepared by functionalising CXs with solutions of H3PO4, H2SO4, ammonium persulfate, or in the flow of oxygen or ammonia at elevated temperatures. The obtained materials contained many different functional groups, including heteroatoms such as O, P, S, and basic N. CXs containing 3.4 wt% of P were found to be the best performing catalysts, reaching 90% conversion of cellobiose with 72% selectivity to glucose, under oxidative atmosphere and in a short reaction time of only 4 h. The presence of phosphonates (–P–C–) increased the selectivity to glucose up to an unprecedented 87%. In contrast, CXs containing other active sites, such as carbonyls and quinones, nitrogen or sulphonic groups, were found to promote glucose degradation, leading to lower final yields [67]. Sulfonated carbons as solid acids showed very high activity in many reactions such as hydrolysis [64], dehydration [68], or esterification [55]. However, despite the high stability of the carbon framework, the sulfonic acid groups are prone to leaching under hydrothermal reaction environments mainly due to the tendency of the C-S bond to undergo acid-catalyzed hydrolysis (desulfonation) through proton leaching from the surface [63, 69, 70]. In this respect, a diazonium coupling strategy leads to the functionalization of carbon materials with more stable sulfonic groups than traditional treatments with concentrated sulfuric acid. This is attributed to introducing a strong covalent bond between carbon and sulfur [71]. We have recently reported that functionalization with diazonium salts introduces a higher concentration of sulfonic groups on carbon materials compared to boiling in concentrated sulfuric acid [53]. Carlier and Hermans studied the performance of different carbon materials, including CNTs, ACs, CNFs, and graphene nanoplatelets in the hydrolysis of cellobiose. Sulfur functionalities were introduced to these materials by applying the diazonium coupling strategy. The authors reported a maximum of 84% conversion of cellobiose with 95% selectivity toward glucose obtained using functionalized graphene nanoplatelets solid acid. The robustness of the sulfonic functions has been shown by the stable performance of this catalyst in 6 consecutive runs [71].

215

216

11  Carbon-based Catalysts for Sustainable Chemical Processes

11.3.2  Cascade Valorization of Biomass with Multifunctional Catalysts The conversion of biomass into valuable bio-products using catalytic routes can adopt one of the following technological approaches: (i) multistep synthesis based on the production of a platform molecule and its further conversion in another process; and (ii) conversion of biomass derivatives through one-pot (cascade) processes involving several catalytic steps to obtain a value-added product [49]. The latter approach is generally preferred because the cascade reactions can improve the atom economy by limiting the number of energy-intensive steps [51]. These reactions require multifunctional catalysts with tuned active sites to facilitate the subsequent steps of the process [72]. The strategy for cascade conversion of cellulose is shown in Figure 11.3. In the one-pot reaction, the substrate is hydrolyzed into glucose in the first step. Subsequently, glucose is converted to platform chemicals or biofuels via various catalytic reactions, including hydrogenation, oxidation, and others. The initial platform molecules obtained from glucose can be further converted to other valuable products [58]. Carbon materials are frequently used as supports in biomass valorization reactions [73]. In fact, our group showed many different methods to functionalize carbon materials with acidic and basic sites [2, 27, 74, 75]. When used as supports for multifunctional catalysts, the functional groups on the surface of carbon can improve the anchoring of the active phase [76]. It should be pointed out that carbon supports can be inert and act as a stabilizer to prevent the agglomeration and inhibit the sintering of metal nanoparticles during heat treatments of materials [73]. However, the functionalized carbons can also assist the active metal sites to promote simultaneous or intermediate reactions or aid in the adsorption of substrates [72, 76, 77]. Synergistic effects between acidic functionalities and metallic sites on carbon support can improve the catalytic performance of the materials [78]. Straightforward recycling of the active phase (i.e. noble metal nanoparticles) via simple burning out the carbon is another advantage of applying carbon supports. Kobayashi et al. discovered that the hydrolysis of cellulose increases in the presence of noble metal catalysts, which led to the discovery of cascade processes [79]. Since then, many routes have been developed to valorize cellulose or cellobiose in one-pot cascade processes. Well-performing multifunctional catalysts have been reported for various cascade processes by supporting active phases (metal nanoparticles) onto functionalized carbon materials.

Figure 11.3  One-pot cascade conversion of cellulose to platform chemicals and biofuels. Adapted from Ref [50].

11.3  Carbon-based Catalysts for Sustainable Production of Chemicals and Fuels from Biomass

For instance, gluconic acid can be obtained from cellulose in one-pot, as shown in Figure 11.3, using bifunctional catalysts containing acidic and metallic sites. Gluconic acid is currently produced by biochemical oxidation of glucose (enzymatic oxidation) and it is widely used in food, cosmetics, and pharmaceutical industries. Thus, there is a large interest to substitute the expensive and unstable enzymes with cost-effective heterogeneous catalysts. Promising results in this cascade reaction were obtained using functionalized CNTs [80], ACs [81], or CXs [72, 77, 82] as supports for noble metal nanoparticles. Our group obtained a remarkably high selectivity to gluconic acid of nearly 80% using gold nanoparticles supported on CX-bearing phenolic groups. This work compared the catalytic performance of 5 wt% gold supported on oxidized CNTs, OMCs and CXs with two different pore sizes. Functionalized and pristine supports were used for catalyst preparation. It was found that the selectivity to gluconic acid in cascade hydrolytic oxidation of cellobiose was strongly influenced by the combined effects of the texture and surface chemistry of the support. As shown in Figure 11.4, CXs with larger mesopore size led to a conformational change of cellobiose, which exposed the glycosidic bond to hydrolysis, whereas the phenolic groups on the surface played a double role as adsorption sites for cellobiose and active sites for hydrolysis. We have also shown that the degradation of glucose and gluconic acid was less pronounced on the CXs containing phenolic groups on the surface, which was attributed to the preferred hydrophilic interactions of the catalyst surface with water, limiting the re-adsorption of the reactants [72]. In another study focused on cascade conversion of cellobiose to gluconic acid, we found that the activity of bifunctional catalysts containing gold supported on CXs was dependent not only on the surface chemistry of the carbon supports, but also on the method used for deposition of the metallic sites. Several gold preparation methods, including wet-impregnation, reduction with NaBH4, deposition-precipitation, sol immobilisation, or reduction with citric acid were applied for the preparation of bifunctional catalysts. The experimental results combined with theoretical calculations of the rate constants of intermediate reactions showed that the catalyst prepared by reduction with citric acid was the most efficient in this cascade reaction. This catalyst was also found to be the least active in the degradation of glucose and gluconic acid, which was attributed to the hydrophilic properties of the CX support introduced during the reduction of the gold precursor with citric acid. The presence of oxygen functionalities on the catalyst obtained via reduction with citric

Figure 11.4  The combined effect of the texture and surface chemistry of carbon xerogels (CX) support containing large pores and phenolic groups on the surface on the first step of cascade conversion of cellobiose to gluconic acid. Adapted from Ref [65].

217

218

11  Carbon-based Catalysts for Sustainable Chemical Processes

acid made the adsorption of water molecules on the carbon surface more favorable than that of the reagents (glucose and gluconic acid), leading to a decrease in re-adsorption and degradation of the latter [77]. Our group has also studied the performance of gold nanoparticles supported on a composite material containing TiO2 and reduced graphene oxide (rGO) [83]. It was found that the presence of a small quantity of rGO in TiO2 changes the strong metal support electronic interaction between gold and TiO2, leading to a decrease in yields of gluconic acid obtained. In this work, we showed that the addition of rGO to reducible oxides like TiO2 could be used for tuning the metal-support interaction. This finding can be applied in other environmentally friendly reactions in which the strong metal-support interaction hinders the catalytic performance [83]. Changing the composition of metallic sites supported on carbon in bifunctional catalysts allows the transformation of cellulose to be tuned within the broad array of value-added products shown in Figure 11.3. For example, a bifunctional catalyst containing nanoparticles active for hydrogenation such as Pt, Rh, or Ru, loaded on CNTs bearing oxygen functionalities, catalyzes the direct conversion of biomass into polyols via hydrolysis–hydrogenation route [57]. Polyols such as sorbitol, mannitol, xylitol, ethylene glycol (EG), and propylene glycol (PG) are often used in pharmaceuticals, food, cosmetics, and textiles industries [60, 84]. Kobayashi et al. reported that a significant reduction of cellulose crystallinity could be obtained by ball-milling. The authors obtained a significant increase in cellulose conversion from 66% to 82% with ball-milling, and the total yield of the sugar alcohols reached 58% [79]. Following this approach, high yields of sorbitol from cellulose were obtained in our laboratories using noble metals supported on pristine AC (i.e. without any acidic functionalities). In our strategy, the hydrolysis step was facilitated by ballmilling. The applied treatment weakened the hydrogen bonding in microcrystalline cellulose, and made the glycosidic bonds more prone to breakage [85]. In another work focused on the one-pot conversion of ball-milled cellulose to sorbitol, we have compared various carbon materials as supports for Ru in the bifunctional catalyst for this process. Most of the catalysts prepared showed high conversions of around 75% of ball-milled cellulose. Interestingly, Ru supported on CNTs obtained the highest selectivity to sorbitol, despite being one of the least active among the studied catalysts. In addition, we have found that the mix-milling of bifunctional catalyst and microcrystalline cellulose significantly shortened the time required to obtain 60% selectivity to sorbitol from 5 h (needed when ball-milled cellulose was used) to only 1h after mix-milling [86]. Moreover, following the mix-milling strategy, selectivities toward sorbitol as high as 80% were obtained in our laboratories using Ru supported on pristine AC [85]. We have also compared AC with CNT as supports in cellulose conversion to sorbitol. Higher dispersion of Ru nanoparticles was obtained on CNT than on AC support, which resulted in increased yields of sorbitol obtained on the former support. The conversion of cellulose and the yield of sorbitol could be further improved by the addition of Ni to Ru forming a bimetallic catalyst supported on CNT [87]. Deng et al. studied the influence of the pre-treatment of CNT supports with HNO3 on the performance of bifunctional Ru/CNT catalysts in cascade conversion of crystalline cellulose to sorbitol. The authors concluded that the carboxylic groups (i.e. acidic sites) on CNT and the hydrogen spillover phenomenon contributed to the high yields of sorbitol obtained [88]. On the other hand, our group studied the influence of the surface chemistry of CNT on the catalytic performance of supported ruthenium in conversion of ball-milled cellulose to sorbitol [89]. In this work, CNTs were first pre-treated with diluted HNO3. Subsequently, the materials were thermally treated to selectively remove some oxygen functionalities from the CNT surface. The catalytic results showed that Ru supported on oxidized CNT achieved higher conversions of cellulose but lower selectivity to sorbitol as compared to its counterpart supported on pristine CNT. We

11.3  Carbon-based Catalysts for Sustainable Production of Chemicals and Fuels from Biomass

Figure 11.5  Schematic pathway of one-pot cascade conversion of cellulose to ethylene glycol (EG) using a physical mixture of Ru/CNT (oxidized) and W/CNT (pristine) developed in our laboratories. Adapted from Ref [91].

attributed the drop in selectivity to sorbitol to the increased amount of side reactions of glucose catalyzed by acidic sites [89]. Interestingly, introducing W to Ru catalysts can lead to the formation of EG or PG instead of hexitols [90]. We have explored the conversion of ball-milled cellulose to EG by applying a physical mixture of Ru supported on CNT containing acidic sites (pre-treated with nitric acid) and W supported on pristine CNT. Based on our experimental results using the physical mixture of these catalysts, we developed the reaction pathway of EG production shown in Figure 11.5. The oxygen functionalities catalyzed the depolymerisation of cellulose to glucose. Glucose was then further converted to glycolaldehyde via retro-aldol condensation with the aid of W sites. In the last step, glycolaldehyde was hydrogenated to EG on Ru nanoparticles. Modified CNTs showed a higher surface area than their pristine counterparts, which facilitated hydrolysis of cellulose. Moreover, the presence of oxygen functionalities on CNT limited the side reaction of glucose, increasing the final yield of the targeted product. The activity of the optimised mixture of the catalysts turned out to be higher than that of the bimetallic Ru-W supported on CNT. The developed catalytic system obtained promising yields of 35% of EG from biomass wastes, such as paper tissue or a sample of eucalyptus wood [91].

11.3.3  Carbon Catalysts Produced from Biomass As mentioned before, biomass waste and biomass-derived sugars can be used to prepare carbon catalysts to replace conventionally used materials such as CNT, ACs, or CXs. These carbonaceous materials, called carbon hydrochars or carbon chars, decrease the carbon footprint of the catalytic processes and limit their overall cost, which is especially important in large scale applications [90]. Recently, we have shown that carbon materials derived from glycerol or carbohydrates (i.e. starch,

219

220

11  Carbon-based Catalysts for Sustainable Chemical Processes

glucose, or sucrose) can be easily prepared by partial carbonization in the presence of H2SO4. These carbons were successfully applied to valorise glycerol (a byproduct of biodiesel production) via transesterification reaction with ethyl acetate to acetins [53]. Acetins are the most interesting among the glycerol derivatives because of their extensive industrial applications, such as fuel additives, cosmetics, solvents, plasticizers, etc [92]. High yields of acetins were obtained, which confirmed the large potential of these materials as solid acid catalysts [53]. Our team has also successfully prepared composite supports applying hydrothermal carbonization of biomass-derived glucose in the presence of CNT [93, 94]. Ru supported on these novel composite materials showed an improved catalytic performance in cellulose conversion to sorbitol when compared to Ru supported on CNT. The higher selectivity to sorbitol obtained on carbon hydrochar-CNT supports was attributed to the basic surface groups on CNT, which prevented sorbitol hydrogenolysis. The yields of sorbitol were further improved on these composite supports by increasing their surface areas, which was accomplished by activating the materials with CO2 [93]. The aforementioned carbon char-CNT supports were also used to deposit Ru-W bimetallic nanoparticles and applied in cellulose conversion to EG in one-pot [94]. Significant yields of EG of around 30% in reaction time of 5 h were obtained using multifunctional catalysts containing carbonized supports. The functionalization of these composite supports with acidic groups increased the yields of EG to 48%, which was attributed to limited side reactions of glucose in the presence of oxygen species. These works demonstrated that low-cost carbon supports obtained from biomass can outperform the more expensive conventional materials. It was also shown that biomass could be a feedstock for the production of both materials and chemicals, similarly to petroleum resources [93, 94].

11.4  Summary and Outlook This chapter was focused on carbon-based catalysts for sustainable chemical processes. Three takeaway messages can be summarized: Carbon is a fascinating material. It can be obtained in several allotropic forms, with different textural properties and surface chemistries. Various strategies are available to design the carbon material to be applied as a catalyst or catalyst support for specific reactions, which usually implies obtaining a pore size distribution that maximizes the surface area and the access of the reactants to the active/anchoring sites. These sites are typically assigned to surface defects or heteroatoms, and the most relevant methodologies for their introduction on carbons were briefly revised. ● Catalytic wet air oxidation and catalytic ozonation for the mineralization of organic pollutants from water sources are two excellent examples of using metal-free carbon materials as catalysts for sustainable processes. Mesoporous materials such as multiwalled CNTs (MWCNTs), to avoid mass transfer limitations, and surface basic groups, easily obtained by N-doping using a solventfree milling methodology, were shown among the most promising catalysts for these reactions. ● Biomass valorization, in a biorefinery concept to replace fossil fuels, is in the spotlight as a sustainable approach to obtain added value platform chemicals and bio-fuels. Bifunctional catalysts, combining a selected acid/base carbon surface and an appropriate metal phase, were shown to be very active for different tandem reactions using a green solvent (water) to obtain added-value products such as gluconic acid, sorbitol, xylitol, EG, among others. It was ­additionally shown that the carbon based catalysts could be obtained by hydrothermal carbonization of biomassderived glucose. ●

References

Acknowledgments This work was financially supported by LA/P/0045/2020 (ALiCE), UIDB/50020/2020 and UIDP/50020/2020 (LSRE-LCM), funded by national funds through FCT/MCTES (PIDDAC). KME acknowledges Junior Researcher Grant (grant no.2021.00535.CEECIND) received from FCT.

References 1 Serp, P. and Figueiredo, J.L. (eds) (2009). Carbon Materials for Catalysis. Hoboken: John Wiley & Sons, Inc. 2 Enterría, M. and Figueiredo, J.L. (2016). Carbon 108: 79–102. 3 Figueiredo, J.L. (2013). J. Mater. Chemistry 1 (33): 9351–9364. 4 Pekala, R.W. (1989). J. Mater. Sci. 24 (9): 3221–3227. 5 Lin, C. and Ritter, J.A. (1997). Carbon 35 (9): 1271–1278. 6 Job, N., Pirard, R., Marien, J., and Pirard, J.-P. (2004). Carbon 42 (3): 619–628. 7 Rey-Raap, N., Arenillas, A., and Menéndez, J.A. (2016). Microporous and Mesoporous Mater. 223: 89–93. 8 Ryoo, R., Joo, S.H., and Jun, S. (1999). J. Phys. Chem. B 103 (37): 7743–7746. 9 Jun, S., Joo, S.H., Ryoo, R. et al. (2000). J. Am. Chem. Soc. 122 (43): 10712–10713. 10 Tanaka, S., Nishiyama, N., Egashira, Y., and Ueyama, K. (2005). Chem. Commun. (16): 2125–2127. 11 Serp, P. (2009). Carbon nanotubes and nanofibers in catalysis. In: Carbon Materials for Catalysis (eds. P. Serp and J.L. Figueiredo), 309–372. Hoboken: John Wiley & Sons, Inc. 12 Pastrana-Martínez, L.M., Morales-Torres, S., Likodimos, V. et al. (2012). Appl. Catal., B 123-124: 241–256. 13 Morales-Torres, S., Pastrana-Martínez, L.M., Figueiredo, J.L. et al. (2012). Environ. Sci. Pollut. Res 19 (9): 3676–3687. 14 Pedrosa, M., Pastrana-Martínez, L.M., Pereira, M.F.R. et al. (2018). J. Environ. Chem. Eng. 348: 888–897. 15 Pedrosa, M., Figueiredo, J.L., and Silva, A.M.T. (2021). Chem. Eng. J. 9 (1): 104930. 16 Hummers, W.S. and Offeman, R.E. (1958). J. Am. Chem. Soc 80 (6): 1339-1339. 17 Stankovich, S., Dikin, D.A., Piner, R.D. et al. (2007). Carbon 45 (7): 1558–1565. 18 Enterría, M., Martín-Jimeno, F.J., Suárez-García, F. et al. (2016). Carbon 105: 474–483. 19 Rey-Raap, N., Enterría, M., Martins, J.I. et al. (2019). ACS Appl. Mater. Interfaces 11 (6): 6066–6077. 20 Rocha, R.P., Soares, O.S.G.P., Figueiredo, J.L., and Pereira, M.F.R. (2016). C-J. Carbon Res. 2 (3): 17. 21 Figueiredo, J.L., Pereira, M.F.R., Freitas, M.M.A., and Órfão, J.J.M. (1999). Carbon 37 (9): 1379–1389. 22 Figueiredo, J.L., Pereira, M.F.R., Freitas, M.M.A., and Órfão, J.J.M. (2007). Ind. Eng. Chem. Res. 46 (12): 4110–4115. 23 Bandosz, T. (2009). Surface chemistry of carbon materials. In: Carbon Materials for Catalysis (eds. P. Serp and J.L. Figueiredo), 45–92. Hoboken: John Wiley & Sons, Inc. 24 Ania, C.O., Armstrong, P.A., Bandosz, T.J. et al. (2020). Carbon 164: 69–84. 25 Boehm, H.P. (2002). Carbon 40 (2): 145–149. 26 Kapteijn, F., Moulijn, J.A., Matzner, S., and Boehm, H.P. (1999). Carbon 37 (7): 1143–1150. 27 Figueiredo, J.L. and Pereira, M.F.R. (2010). Catal. Today. 150 (1): 2–7. 28 Stöhr, B., Boehm, H.P., and Schlögl, R. (1991). Carbon 29 (6): 707–720. 29 Bohem, H.P. (2009). Catalytic properties of nitrogen-containing carbons. In: Carbon Materials for Catalysis (eds. P. Serp and J.L. Figueiredo), 219–265. Hoboken: John Wiley & Sons, Inc.

221

222

11  Carbon-based Catalysts for Sustainable Chemical Processes

30 Biniak, S., Szymański, G., Siedlewski, J., and Świątkowski, A. (1997). Carbon 35 (12): 1799–1810. 31 Rechnia-Gorący, P., Malaika, A., and Kozłowski, M. (2020). Catal. Today. 357: 102–112. 32 El-Sayed, Y. and Bandosz, T.J. (2005). Langmuir. 21 (4): 1282–1289. 33 Lahaye, J. (1998). Fuel 77 (6): 543–547. 34 Zhang, R., Lu, Y., Zhan, L. et al. (2002). Carbon 411: 1645–1687. 35 Pérez-Cadenas, M., Moreno-Castilla, C., Carrasco-Marín, F., and Pérez-Cadenas, A.F. (2009). Langmuir. 25 (1): 466–470. 36 Gorgulho, H.F., Gonçalves, F., Pereira, M.F.R., and Figueiredo, J.L. (2009). Carbon 47 (8): 2032–2039. 37 Liu, G., Li, X., Ganesan, P., and Popov, B.N. (2009). Appl. Catal., B. 93 (1): 156–165. 38 Terzyk, A.P. (2001). Colloids Surf., A. 177 (1): 23–45. 39 Terzyk, A.P. (2003). J. Colloid Interface Sci. 268 (2): 301–329. 40 Wang, X., Liu, R., Waje, M.M. et al. (2007). Chem. Mater. 19 (10): 2395–2397. 41 Xiao, H., Guo, Y., Liang, X., and Qi, C. (2010). J. Solid State Chem. 183 (7): 1721–1725. 42 Soares, O.S.G.P., Rocha, R.P., Gonçalves, A.G. et al. (2015). Carbon 91: 114–121. 43 Soares, O.S.G.P., Rocha, R.P., Órfão, J.J.M. et al. (2019). C. 5 (2): 30. 44 Figueiredo, J.L. and Pereira, M.F.R. (2009). In: Carbon Materials for Catalysis (eds. P. Serp and J.L. Figueiredo), 177–217. Hoboken, NJ: John Wiley & Sons, Inc. 45 Figueiredo, J. (2017). Application of nanocarbon materials to catalysis. In: Nanotechnology in Catalysis: Applications in the Chemical Industry, Energy Development, and Environment Protection (eds. B. Sels and M. van de Voorde), 37–56. Weinheim: Wiley-VCH Verlag GmbH & Co. KGaA. 46 Soares, O.S.G.P., Rocha, R.P., Gonçalves, A.G. et al. (2016). Appl. Catal., B 192: 296–303. 47 Rocha, R.P., Santos, D.F.M., Soares, O.S.M.P. et al. (2018). Top. Catal. 61 (18): 1957–1966. 48 Malaika, A., Morawa Eblagon, K., Soares, O.S.G.P. et al. (2020). Appl. Surf. Sci. 511: 145467. 49 Gallezot, P. (2012). Chem. Soc. Rev. 41 (44): 1538–1558. 50 Chheda, J.N., Huber, G.W., and Dumesic, J.A. (2007). Angew. Chem. Int. Ed. 46 (38): 7164–7183. 51 Sudarsanam, P., Zhong, R., Van den Bosch, S. et al. (2018). Chem. Soc. Rev. 47 (22): 8349–8402. 52 Cao, X., Sun, S., and Sun, R. (2017). RSC Adv. 7 (77): 48793–48805. 53 Malaika, A., Ptaszyńska, K., and Kozłowski, M. (2021). Fuel 288: 119609. 54 Malaika, A., Ptaszyńska, K., Morawa Eblagon, K. et al. (2021). Fuel 304: 121381. 55 Malaika, A. and Kozłowski, M. (2019). Fuel Process. Technol. 184: 19–26. 56 Lin, L., Han, X., Han, B., and Yang, S. (2021). Chem. Soc. Rev. 50 (20): 11270–11292. 57 Ribeiro, L.S., Órfão, J.J.M., and Pereira, M.F.R. (2021). Mater. Today Sustainability. 11-12: 100058. 58 Mika, L.T., Cséfalvay, E., and Németh, Á. (2018). Chem. Rev. 118 (2): 505–613. 59 Huang, Y.-B. and Fu, Y. (2013). Green Chem. 15 (5): 1095–1111. 60 Deng, W., Tan, X., Fang, W. et al. (2009). Catal. Lett. 133 (1): 167. 61 Zeng, M. and Pan, X. (2020). Catal. Rev. Sci. Eng. 1–46. 62 Hu, L., Lin, L., Wu, Z. et al. (2015). Appl. Catal., B 174-175: 225–243. 63 Shrotri, A., Kobayashi, H., and Fukuoka, A. (2018). Acc. Chem. Res. 51 (3): 761–768. 64 Suganuma, S., Nakajima, K., Kitano, M. et al. (2008). J. Am. Chem. Soc. 130 (38): 12787–12793. 65 Foo, G.S. and Sievers, C. (2015). ChemSusChem. 8 (3): 534–543. 66 Shen, S., Cai, B., Wang, C. et al. (2014). Appl. Catal., A 473: 70–74. 67 Morawa Eblagon, K., Malaika, A., Pereira, M.F.R., and Figueiredo, J.L. (2018). ChemCatChem. 10 (21): 4934–4946. 68 Zhao, J., Zhou, C., He, C. et al. (2016). Catal. Today. 264: 123–130. 69 Scholz, D., Kröcher, O., and Vogel, F. (2018). ChemSusChem. 11 (13): 2189–2201. 70 Onda, A., Ochi, T., and Yanagisawa, K. (2009). Top. Catal. 52 (6): 801–807.

References

71 Carlier, S. and Hermans, S. (2020). Front. Chem. 8. 72 Morawa Eblagon, K., Pereira, M.F.R., and Figueiredo, J.L. (2016). Appl. Catal., B 184: 381–396. 73 Xia, D., Yu, H., Li, H. et al. (2022). Environ. Chem. Lett. 20 (3): 1719-1744. 74 Morawa Eblagon, K., Rey-Raap, N., Figueiredo, J.L., and Pereira, M.F.R. (2021). Appl. Surf. Sci. 548: 149242. 75 Sousa, J.P.S., Pereira, M.F.R., and Figueiredo, J.L. (2011). Catal. Today. 176 (1): 383–387. 76 Lam, E. and Luong, J.H.T. (2014). ACS Catal. 4 (10): 3393–3410. 77 Morawa Eblagon, K., Pereira, M.F.R., and Figueiredo, J.L. (2018). Catal. Today. 301: 55–64. 78 Yang, Z., Luo, B., Shu, R. et al. (2022). Fuel 319: 123617. 79 Kobayashi, H., Ito, Y., Komanoya, T. et al. (2011). Green Chem. 13 (2): 326–333. 80 Tan, X., Deng, W., Liu, M. et al. (2009). Chem. Commun. (46): 7179–7181. 81 Onda, A., Ochi, T., and Yanagisawa, K. (2011). Catal. Commun. 12 (6): 421–425. 82 Mager, N., Meyer, N., Léonard, A.F. et al. (2014). Appl. Catal., B 148-149: 424–435. 83 Morawa Eblagon, K., Pastrana-Martínez, L.M., Pereira, M.F.R., and Figueiredo, J.L. (2018). Energy Technol. 6 (9): 1675–1686. 84 Ruppert, A.M., Weinberg, K., and Palkovits, R. (2012). Angew. Chem. Int. Ed. 51 (11): 2564–2601. 85 Ribeiro, L.S., de Melo Órfão, J.J., and Pereira, M.F.R. (2015). Green Process. Synth. 4 (2): 71–78. 86 Ribeiro, L.S., Delgado, J.J., de Melo Órfão, J.J., and Pereira, M.F.R. (2017). Catal. Today 279: 244–251. 87 Ribeiro, L.S., Delgado, J.J., de Melo Órfão, J.J., and Pereira, M.F.R. (2017). Appl. Catal., B 217: 265–274. 88 Deng, W., Liu, M., Tan, X. et al. (2010). J. Catal. 271 (1): 22–32. 89 Ribeiro, L.S., Delgado, J.J., de Melo Órfão, J.J., and Pereira, M.F.R. (2017). ChemCatChem. 9 (5): 888–896. 90 Ji, N., Zhang, T., Zheng, M. et al. (2008). Angew. Chem. Int. Ed. 47 (44): 8510–8513. 91 Ribeiro, L.S., de Melo Órfão, J.J., and Pereira, M.F.R. (2018). Bioresour. Technol. 263: 402–409. 92 Zada, B., Kwon, M., and Kim, S.-W. (2022). Molecules 27 (7): 2255. 93 Rey-Raap, N., Ribeiro, L.S., de Melo Órfão, J.J. et al. (2019). Appl. Catal., B 256: 117826. 94 Ribeiro, L.S., Rey-Raap, N., Figueiredo, J.L. et al. (2019). Cellulose 26 (12): 7337–7353.

223

225

12 Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes Giulia Tuci1, Andrea Rossin1, Matteo Pugliesi1, Housseinou Ba2, Cuong Duong-Viet2, Yuefeng Liu3, Cuong Pham-Huu2, and Giuliano Giambastiani1 1 Institute of Chemistry of OrganoMetallic Compounds, ICCOM-CNR and Consorzio INSTM, Via Madonna del Piano, 10, Sesto F.no, Florence, Italy 2 Institute of Chemistry and Processes for Energy, Environment and Health (ICPEES), ECPM, UMR 7515 of the CNRS and University of Strasbourg, 25 rue Becquerel, 67087 Strasbourg Cedex 02, France 3 Dalian National Laboratory for Clean Energy (DNL), Dalian Institute of Chemical Physics, Chinese Academy of Science, 457 Zhongshan Road, Dalian, China

12.1  Introduction Catalysis is the driving force behind the most important processes of modern industrial chemistry. At present, 85–90% of chemical processes include at least one catalytic step. The ever-increasing involvement of catalysis in industrial processes meets with the urgency for greener solutions to the production of chemicals and commodities while reducing process costs and their environmental impact. Oxidation reactions represent a fundamental and wide branch of catalysis applied to the production of many fundamental compounds [1, 2]. Among heterogeneous catalysts industrially employed since mid-1950s, alkaline-, alkaline-earth, or lanthanide-promoted transition metal-oxide systems have become crucial in several refining and petrochemical processes. However, the urgency to make catalytic processes greener and more sustainable to meet with health, safety, and environmental regulations of modern society has prompted scientists to find cheaper and alternative solutions. Metal-based catalysts toxicity and the environmental impact linked to their storage/disposal at the end of their catalytic life represent additional concerns that need to be tackled. Carbon-based catalysts in the form of pure C-networks or light heteroatomdoped systems have witnessed a wonderful technological renaissance in catalysis. Indeed, many earth-abundant C-based 1D-3D porous architectures have shown unique performance as single-phase and metal-free catalysts in a number of industrially relevant transformations [3, 4]. This contribution focuses on two fundamental oxidation processes in the area of natural gas treatment and petrochemical specialty synthesis: the selective H2S oxidation to elemental sulfur for the natural tail gas purification and the alkane to olefins dehydrogenation. With this dedicated chapter, the authors would like to highlight the remarkable achievements of environmentally friendly, durable, and cheap C-based catalytic materials in this specific area of heterogeneous catalysis.

Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 1, First Edition. Edited by Armando J.L. Pombeiro, Manas Sutradhar, and Elisabete C.B.A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

226

12  Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes

12.2  The H2S Selective Oxidation to Elemental Sulfur The partial H2S oxidation to elemental sulfur represents a process of primary importance both from an environmental and an economic viewpoint. Hydrogen sulfide (H2S) is currently one of the most harmful gaseous pollutants for the human health and the environment, primarily coming from petrochemical industry throughout the management of natural fossil fuels such as coal, petroleum, and natural gas. It is also a highly corrosive, poisonous, and malodorous gas whose removal from industrial gaseous wastes is mandatory before it comes in contact with the atmosphere [5–8]. On the other hand, elemental sulfur, with its production from oil and gas processing that approaches 75 Mt per year, is a highly desirable compound for industrial manufacturing covering various fields, from chemistry and construction to the synthesis of fertilizers. At present, the so-called Claus process is the most widely used and consolidated technology for recovering elemental sulfur from the treatment of acidic gaseous streams with H2S concentrations even higher than 50%. The process was patented in 1883 by the German chemist Carl Friedrich Claus and it was significantly implemented later on [9, 10]. It consists of a sequence of thermal and catalytic steps whose thermodynamic limitations allow a maximum sulfur recovery of 98% [8]. Current methods of recovering sulfur from H2S gas streams typically combine the Claus process with a tail gas treatment unit (TGTU) as to improve the sulfur recovery efficiency (SUPERCLAUS® process) up to 99.7–99.9% (Eq. 12.1) [8]. The tail gas oxidation step is performed on acidic streams typically containing from 0.4 to 1.5 vol.% of H2S. The presence of molecular oxygen in the reagents stream requires a crucial control on the oxidation selectivity toward the generation of elemental sulfur, hence minimizing the formation of highly undesired over-oxidation by-products (mainly sulfur dioxide SO2; Eqs. 12.2 - 12.3) [11].

H2S + ½ O2 → S + H2O

(12.1)



S + O2 → SO2

(12.2)



H2S + 3 / 2  O2 → SO2 + H2O

(12.3)

Classical metal/metal oxide catalysts (i.e. Ti, Cr, V, and Fe-based materials) have been exploited for promoting the selective H2S oxidation to elemental sulfur for a long time [12–15]. However, they suffer from several drawbacks: a fast deactivation on run, their inherent toxicity as well as the costs and environmental concerns associated to their disposal/recovery at the end of the catalyst lifetime. For this reason, the scientific community is now trying to find valuable alternatives. Carbon-based materials represent an excellent replacement of classical metal-based catalysts. These metal-free and sustainable catalytic materials have shown excellent performance in the oxidation process since the pioneering work of Cariaso and Walker, who first reported in 1975 on the H2S partial oxidation promoted by activated microporous carbon samples [16]. Since that time, several aspects dealing with the activity of carbon materials in the process and the use of coreagents in the stream have systematically been investigated. In particular, researchers in this field have studied the beneficial role played by water as co-feed in the reagents stream for the reaction rate enhancement and the control of amount of sulfur gathered at the catalyst surface [17–19]. On the other hand, the carbon surface oxidation was found to have a negative impact on the catalyst activity, as it favors the occurrence of undesired over-oxidation paths with the generation of sulfur oxide species [20]. In early 2000s, a common strategy to increase the catalytic activity of activated carbons (ACs) in the process was the impregnation of carbonaceous networks with alkaline species

12.2  S Selective Oxidation to Elemental Sulfur

(i.e. NaOH or Na2CO3). Such an approach fosters H2S capture and limits catalyst deactivation [21, 22]. In spite of relatively high activity and fast reaction kinetics (already under low-operational temperatures), these materials suffer from very low sulfur saturation capacity due to their small pores and low pore volumes. Indeed, for low temperature processes (90% sulfur selectivity already at 180 °C [39]. However, low H2S concentrations and relatively high catalyst amounts were mandatory, because CN has a low specific surface area (SSA) and a low number of exposed active sites. Accordingly, the same authors reported later on the preparation of holey CN nanosheets by thermal exfoliation of bulk CN. The as prepared ordered and mesoporous CN-based architectures improved the diffusion of reactants and products, hence ensuring a better catalytic performance [40, 41]. Following this idea, the same group developed a series of N-rich mesoporous carbons [42], hierarchical porous polymers [43, 44] and 2D micro-mesoporous carbon-based materials [45], concluding that desulfurization activity is boosted by high SSA values combined with abundance of exposed pyridinic groups. Similar conclusions regarding the structure-reactivity relationship of N-doped carbon materials in H2S selective oxidation were drawn in the same years by an Iranian team who combined experimental outcomes with DFT simulations [46, 47]. In particular, they attributed the catalytic activity of their systems to appropriate morphology with a narrow pore size distribution (< 3 nm) and to the presence of pyridinic nitrogen in the form of pyridinic-oxide groups. The latter are found to boost H2S dissociation, a fundamental elementary step in the process. In a related work, Liu et al. prepared a series of highly porous N-doped porous carbon-based (NPC) samples with high uniformity and size tunability. These authors have demonstrated that desulfurization activity is affected by dimensions of NPC particles: the smaller the particle size, the higher their catalytic activity due to greater exposure of N-basic active sites and more rapid diffusion of reagents and products [48]. In the last few years, many research efforts have been devoted to unveiling the nature of active sites. Several authors affirmed the central role played by the overall N-content [49, 50] and, in particular, by pyridinic sites [51, 52] for H2S absorption and subsequent oxidation. Other groups drew different conclusions by combining DFT calculations with advanced materials characterization techniques. In particular, Jiang et al. identified both pyridinic and pyrrolic nitrogen as active sites for H2S adsorption and charge transfer [53], whereas Liu et al. uncovered a linear correlation between sulfur formation rates and catalysts surface defects density (Figure 12.2). The latter team demonstrated that edge and holey carbon structural defects induce a high exposure of pyridinic active sites that synergistically contribute to the observed catalytic activity (Figure 12.2) [54]. DFT evidence suggested that carbon atoms neighboring N-pyridinic are the preferential sites for HS− species adsorption and activation.

229

230

12  Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes

Figure 12.2  Relationship between sulfur formation rate and (a) defect density or (b) pyridinic N content normalized by specific surface area (% m−2) in various N-doped carbon nanoflakes. (c) Schematic illustration of the defect enriched N-doped carbon nanoflake model and its activity as carbocatalyst for selective oxidation of H2S to elemental sulfur. Reproduced with permission from Ref [54].

Very recently, a series of covalent triazine frameworks CTFs have been investigated as highly porous, N-rich organic polymers for the process. Their large surface area (up to 3900 m2/g) together with their micro-mesoporous structure ensure a fruitful accessibility of reagents to N-basic sites and promote mass transfer, therefore allowing complete H2S conversion to sulfur to be reached at 180 °C [55]. Pham-Huu et al. extended the carbon nanomaterials hetero-doping from nitrogen to oxygen treating CNTs and graphite felts with HNO3 vapors, hence introducing oxygenated functional groups at the outer materials surface [56, 57]. This group demonstrated that the oxidative treatment causes the formation of defective surface carbon sites and O-containing groups, both of which act as active sites in desulfurization. The as-prepared materials exhibit remarkable activity along with high resistance to deactivation even under harsh conditions (i.e. high WHSV and low O2/H2S molar ratio). However, in accordance with preliminary studies on AC [20], the presence of strong adsorption sites boosts the complete oxidation of sulfur to SO2, hence making O-doped carbon materials less selective than the iron-based benchmark catalyst. Despite this drawback, the same group has very recently prepared hierarchical materials based on CNTsdecorated carbon felt (CF) that underwent an oxidative post-treatment with HNO3 vapors. The as-obtained O-doped carbon composites were tested for H2S selective oxidation in a fixed-bed reactor using induction heating (IH) instead of the “classical” Joule-heated (JH) reactor for the control of the temperature at the catalytic bed. They exploited an electrically conductive material as both susceptor for the IH technology and catalyst for the oxidation process. Therefore, better control of the temperature at the catalyst bed can be obtained, while reducing energy wastes and dissipation phenomena [58]. The authors compared the performance of their CNTs/CF composites under traditional and inductive heating mode and demonstrated the superior desulfurization activity and the remarkable catalyst stability when catalysis is operated under IH [59]. However, sulfur selectivity remains moderate for both reactor configurations and irrespective of the different conversion values, as previously reported for O-decorated materials [56, 57]. Zhang, Qian et al.

12.3  Alkane Dehydrogenation

examined the catalytic activity of different O-functionalities at the carbon nanomaterials surface in more depth. Taking advantage of kinetic studies and DFT calculations they found that both pyrone [60] and benzoquinone [61] groups play a pivotal role in promoting H2S desulfurization due to the lower energy required for the electronic transfer. Selectivity toward the production of elemental sulfur did not exceed 75 % also in these cases.

12.3  Alkane Dehydrogenation The last few decades have witnessed an important renaissance in the catalyst technology applied to one of the most industrially relevant and energy-demanding catalytic transformations: the alkane-to-alkene dehydrogenation [62–64]. Olefins are fundamental chemical building blocks for the manufacturing of a variety of other chemical components and plastic materials. Their global market value is estimated in more than 330 USD billions by the next five to six years with an annual growth rate of 5% [65]. Direct dehydrogenation (DDH) is an endothermic process largely employed for alkane-to-alkene dehydrogenation and typically carried out at temperatures between 550–750 °C in industrial processes. From an atom economy viewpoint, it is highly favorable as the only by-product is molecular hydrogen than can be exploited as such as a carbon-free fuel or it can be reused as reagent for other industrial transformations. DDH is generally associated to the use of large amounts of superheated steam whose role is not only to provide the thermodynamic driving force for the dehydrogenation process but also to reduce the partial pressure of substrates, thus shifting the chemical equilibrium to higher olefins conversions. Oxygen co-fed streams (oxidative dehydrogenation [ODH] conditions) are generally more attractive for alkanes dehydrogenation, at least from a thermodynamic viewpoint. Indeed, ODH is an exothermic process that occurs under milder temperature conditions. Both protocols suffer (albeit to a different extent) from “catalysts coking” phenomena, responsible for their progressive fouling with the subsequent reduction of their performance over long-term catalytic runs. As a result, costly and energy-demanding catalysts regeneration treatments based on oxygen-rich streams are periodically required to burn out all undesired coke deposits and restore the pristine catalyst performance [66]. Although large steam amounts in DDH represent a costly and highly energy-demanding step in the process, the use of oxygen as co-reagent reduces the process selectivity, often favoring undesired over-oxidation paths and introducing a potentially dangerous (explosive) issue (particularly for industrial applications). Various transition metals (i.e. Co, Ni, Cu, Cr, Ce, V, Zr, Ga, Sb, Zn, Mn, and Mo) have been largely used as promoters for the more conventional iron-potassium oxide system [67, 68]. Regardless, the often rapid fouling of metallic active sites and high costs, combined with the toxicity and environmental concerns linked to the storage/disposal of the exhausted catalytic materials (with a high metallic content) have again boosted research to attain more environmentally friendly and sustainable catalytic solutions [69]. The following sections describe the efforts and achievements in the field of carbonbased materials as single-phase and metal-free catalysts in ODH and DDH.

12.3.1  Alkane Dehydrogenation under Oxidative Environment: The ODH Process Although several transition-metal promoted metal-oxide materials have been described since the early seventies as effective alkane dehydrogenation catalysts under an oxidative environment [70–75], it has become clear that carbonaceous overlayers (coke deposits) formed at the catalyst surface take an active part in the process [76, 77]. These findings opened a new branch of catalysis

231

232

12  Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes

based on the use of cheaper and more sustainable carbon-based metal-free systems for ODH. In 1990, Grunewald and Drago demonstrated the potentiality of purely carbon-based materials as active and selective catalysts for the EB ODH to styrene (ST) [78]; since then, many studies have been carried out with these compounds [79, 80]. Initially, the activity of the carbonaceous catalysts was analyzed in light of their morphological properties (i.e. SSA and pore size distribution). Later, other important factors such as the nature of the chemical groups present on their surface started to be taken into account as well to explain their reactivity. The Figueiredo group first demonstrated that the nature of functional groups at the surface of ACs deeply influences their dehydrogenation activity. In particular, oxygen-containing moieties were easily introduced upon materials surface oxidation and were found to improve the ultimate catalysts performance. The authors demonstrated the key role in ODH played by the carbonyl/quinone groups through a redox mechanism well described by means of appropriate kinetic models [81–83]. Ketonic groups are supposed to be reduced in the process, hence fostering the alkane to alkene oxidation while the oxygen co-feed restores the carbonyl moieties for the successive catalytic cycle. Despite these encouraging results, some limitations related to the low oxidation resistance of the catalytic materials and the slow ST desorption rate caused by the catalysts microporosity (leading to over-oxidation and consequent low process selectivity) limit the practical applications of ACs in alkanes ODH [84]. In addition, materials deactivation due to coke deposits formation was demonstrated, with the newly formed carbonaceous layer acting as new active phase with reduced performance [85]. To overcome these drawbacks, in the early 2000s, Schlögl’s group developed carbon nanofilaments (i.e. multiwalled carbon nanotubles [MWCNTs]) [86] and onion-like carbons (OLCs) [87] as Csp2-hybrized model nanomaterials, displaying high stability toward oxidation and ideal porosity as to ensure a good reagents and products diffusion. They reached ST selectivity values up to 68%. Starting from these seminal papers, several groups optimized different synthetic strategies to prepare sp2 hybridized carbon nanomaterials to be tested in alkane ODH in order to better define their structure-reactivity relationship and substantially confirming the key role of carbonyl-quinone/ hydroxyl groups as redox couples directly engaged in the process and showing remarkable stability on stream [88–91]. In 2007, Su et al. first reported an in-depth ODH mechanistic study testing different carbon allotropic forms: ACs and nanodiamonds (NDs) as a model for sp3-hybridized carbons and sp2-CNTs. They found that the ordered microstructure of the materials is essential to guarantee their stability on stream. Their study led to the conclusion that only nanocarbons (i.e. CNTs and NDs) were robust and durable catalysts in ODH (Figure 12.3). In addition, they demonstrated that the overall reaction rate is kinetically limited by EB C–H bond breaking and that both NDs and CNTs show similar kinetics but different selectivity [92]. In particular, the sp3-carbon configuration produces a relatively large amount of benzene in the first hours on stream before switching its productivity toward ST due to the gradual formation on NDs surface of a layer of the more selective sp2-hybridized carbon [92, 93]. Starting from these seminal findings, Su et al. in collaboration with the German group of Schlögl. pursued studies in alkane ODH for the following decade, setting many milestones in the field. In 2008, they tested CNTs in the ODH of light alkanes (i.e. butane) for the first time. This process represents a challenging transformation due to the high tendency of the low-weight unsaturated products to undergo over-oxidation, which affects the process selectivity. In particular, authors demonstrated that CNTs surface functionalization with ketonic groups followed by phosphorous passivation leads to high selectivity toward the corresponding alkenes with remarkable stability on a stream up to 100 h [94]. If ketonic groups are well-recognized as the active sites for the process, the introduction of phosphorous is found to significantly enhance selectivity by suppressing the

12.3  Alkane Dehydrogenation

Figure 12.3  Reaction rate (left) and styrene (ST) selectivity (right) for ethylbenzene (EB) oxidative dehydrogenation (ODH) promoted by activated carbon (AC), nanotubes (NTs), and nanodiamonds (NDs). Reproduced with permission from Ref [92].

combustion of undesired hydrocarbons. In a related work, they demonstrated the efficiency of borate- and phosphate-modified oxygen-decorated CNTs in promoting propane [95] and ethane [96] ODH with low activity but remarkable propene and ethylene selectivity. Additionally, kinetic studies demonstrated that the insertion of B and P compounds in these cases does not affect the formation rate of alkenes, but rather increases the process selectivity by suppressing the combustion side reaction. The exact promotion mechanism responsible for higher selectivity values in the process was elucidated only later, in 2015. These authors used phosphate-modified NDs as model materials and showed that phosphate groups interact preferentially with phenol moieties at the nanocarbons surface before (selectively) blocking the defective sites responsible for hydrocarbons over-oxidation. As a consequence, selectivity is enhanced while ketonic active groups for the alkane ODH are entirely preserved [97]. Notably, the ODH activity of carbon nanomaterials is dependent from the phosphate loading and with improved selectivity registered only for low P contents [98]. For too high phosphate loadings, also ketonic active groups start to be “passivated”, hence decreasing ODH conversion values until complete deactivation of materials occurs (for P contents >15 wt.%). Therefore, phosphate was used as a modulator for tuning ODH activity with a volcano-like distribution of selectivity values (Figure 12.4) [98, 99]. A similar study was also performed by the same authors, who examined the effect of borate groups as additives. In this respect, they observed a different ODH selectivity dependence from the additive content. They found that alkene selectivity increases until 5 wt.% of B2O3 loading, whereas higher borate charges do not change either alkane conversion or selectivity appreciably. This phenomenon is significantly different with respect to the volcano-type distribution observed for phosphate-modified carbon materials and it is ascribed to a more selective blockage of surface defects exerted by borate that avoids the indiscriminate coverage of active ketonic groups with subsequent catalyst deactivation [100]. Beyond the introduction of P and B- based additives, the same group investigated the role of carbon hybridization and structure on alkane ODH activity. In particular, they prepared a series of model sp2@sp3 core-shell materials finding that active oxygen species are generated in situ on the materials edges by molecular O2 present in the reactant feed. These studies further confirm the beneficial role of graphitic sp2-hybridized species in the generation of surface quinoidic groups suitable for promoting ODH while suppressing the formation of electrophilic oxygen fragments

233

234

12  Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes

Figure 12.4  Product distributions for isopentane oxidative dehydrogenation (ODH) promoted by carbon nanotubes (CNTs) with different phosphate loadings. Green, blue, and pale green bars (monoolefins, C5H10 isomers); white bar diolefin (isoprene); black line CO; red line CO2. Reproduced with permission from Ref [98].

detrimental for selectivity [101, 102]. At the same time, amorphous carbon is confirmed to have a negative influence on the ODH performance, for both EB and light alkanes [103]. Catalytic tests using discrete molecular catalysts (i.e. phenanthrequinone cyclotrimer) [104] and in situ analysis/18O isotope tracer experiments [105] were employed by Su et al. to confirm the role of ketonic functional groups in ODH through their initial reduction (formally by an hydrogen abstraction from the alkane reagent) followed by re-oxidation by molecular oxygen to restart the catalytic cycle [104–106]. Finally, they proposed a direct chemical titration method for the quantification of the active sites effectively available at the surface in order to unveil the intrinsic activity of ODH catalysts and to compare their performance [107, 108]. The carbonyl groups on oxidized CNTs were poisoned by phenyl hydrazine (selected because of its size and structure similar to EB) by means of selective and stoichiometric hydrazone formation. As shown in Figure 12.5, EB conversion dropped down once titration started, until complete suppression of catalyst activity occurred when the system was saturated with phenyl hydrazine. In addition to demonstrating the pivotal role played by carbonyl groups in ODH, this methodology effectively quantifies the accessible active sites and normalizes catalyst reaction rates when comparing the activity of different materials [107]. Continuing their long tradition in the field, Su et al. tested N-doped nanocarbons for the propane ODH for the first time in 2013 [109]. In that period, N-doped CNTs were found to activate molecular oxygen. The nitrogen incorporation within the C network reduces the highest occupied molecular orbital (HOMO)-lowest unoccupied molecular orbital (LUMO) gap and fosters the electronic transfer from CNTs to adsorbed O2 [110]. Accordingly, Su’s group demonstrated that N-CNTs exhibit enhanced catalytic performance with respect to undoped CNTs and the higher the nanomaterial N-loading, the higher the process activity and selectivity. As demonstrated by first principle studies [111], N-dopants speed up the dissociative adsorption of molecular O2 and reduce the overall O2 dissociation energy. In particular, authors identified graphitic nitrogen species in N-CNTs as the main players in propane dehydrogenation by correlating N1s X-ray photoelectron spectroscopy (XPS) analysis with the observed catalytic performance [109]. This conclusion has

12.3  Alkane Dehydrogenation

Figure 12.5  (left) Schematic representation of the in situ titration process for the oxidative dehydrogenation (ODH) active sites on nanocarbon catalysts. (right) Ethylbenzene (EB) conversion and titrant uptake as a function of time on stream during the steady-state activity measurement and in situ titration process. Reproduced with permission from Ref [107].

recently been re-discussed by Slabon and Kusstrowski, who performed a detailed computational study based on 15N solid state nuclear magnetic resonance (NMR) spectroscopy of N-doped mesoporous carbon in ODH [112]. These authors revealed that a high concentration of specific N-containing groups (i.e. N-pyridinic and N-pyrrolic sites) along with C=O/C–OH moieties are responsible of the observed EB-to-ST ODH catalytic activity. A more detailed and systematic study on the effects of B, P, and N-doping on carbon nanomaterials performance employed for alkane ODH was reported in 2014 by Garcia-Bordeje et al. [113]. These authors prepared a series of B, P, and N homo-decorated carbon nanofibers (CNFs) and confirmed the big influence of heteroatoms doping on the alkane-to-alkene ODH activity and selectivity. Boron and phosphorus doping is found to increase selectivity toward alkenes, with a P-content fixed to 1.5 wt.% and an optimum B-loading of 1.5 wt.% (higher B contents decrease propane conversion as a result of the blockage of active sites). On the other hand, N groups are found to be the most active (but also the least selective) for propane conversion [113]. Recently, Bordoloi et al. reported for the first time B-N co-doped carbon nanomaterials as metal-free catalysts for propane ODH. In particular, these authors proposed a single-step synthetic strategy to prepare B and N co-doped hierarchical porous materials, demonstrating through solid-state NMR and XPS analysis the formation of N-B-C species. The developed samples exhibit high prominence to activate alkanes C–H bond with up to 85% selectivity toward propylene together with an excellent stability under the operative reaction conditions [114]. More recently, Lu et al. reported on boron nitride (BN) embedded graphitic carbon to be tested in EB ODH. The authors proved the synergistic effect between BN and carbon, reaching high selectivity (94%) and a remarkable ST formation rate through the combination of mesoporosity (which prevents ST over-oxidation) and the hybrid BN/C structure (which provides abundant anchoring sites for active oxygen functional groups) [115]. Beyond heterodoping effects, another important factor affecting ODH activity is the catalyst morphology. Several studies were devoted to find a correlation between textural properties and catalytic performance. In 2012, Kozlowski et al. prepared two carbon catalysts differing only in their meso- or microporous structure, with the aim of analyzing the influence of this parameter on their ODH performance [116]. The authors found that the initial activity increases with the overall SSA values but no significant effects are associated with varying the materials micro- or mesoporous texture. Catalyst deactivation is unavoidable, even if it is slower for samples characterized by

235

236

12  Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes

higher total pore volume. Similar conclusions were drawn by Figueiredo’s group for isobutane ODH. The authors demonstrated a less pronounced deactivation of ACs due to the reduced generation of coke deposits for samples with higher surface area [117]. As previously discussed [81, 85], coke deposits cause catalyst fouling, preventing the regular contact of the reactants with the oxidized functional carbons active species in ODH [118]. With the aim of improving catalyst ODH activity and stability on stream, Su et al. developed a series of highly porous graphenic materials. In particular, porous reduced graphene oxide (GO) with ultra-high surface area (~2600 m2 g−1) [119] and AC decorated with few layer graphene [120] were prepared and successfully exploited in ODH of alkanes. In both cases, the abundant material edges were found to be ideal sites for the formation of active ketonic carbonyl groups. At the same time, the high SSA favors the diffusion of reagents and products, avoiding over-oxidation and ensuring an overall high conversion and selectivity. The graphitic structure enhances the oxidation resistance of materials under the operative catalytic conditions, making the catalyst considerably stable on stream. The findings of these studies are in accordance with conclusions drawn by Schwartz’s group, which demonstrated (both experimentally and theoretically) that few-layer graphene sheets are ideal ODH catalysts [121– 123]. In particular, their DFT analyses showed that dicarbonyl moieties at the zig-zag edges and quinone groups on armchair boards are beneficial for ODH because both functionalities easily undergo dehydrogenation/regeneration cycles [123]. Afterward, many groups investigated the role of carbonyl moieties in carbon-based nanomaterials for ODH both experimentally [124–130] and theoretically [131–133]. Among them, Su et al. performed a kinetic analysis in 2018 by means of active site titration, isotope, and temperature programmed surface reaction experiments. At the end of the study, the research team invoked a radical-type mechanism, in which EB is first adsorbed on the ketonic groups with abstraction of two hydrogen atoms to form ST and then O2 is assumed to be activated and dissociated restoring the active sites (Figure 12.6) [130].

Figure 12.6  Postulated reaction mechanism for ethylbenzene (EB) oxidative dehydrogenation (ODH) on ketonic carbonyl groups of nanocarbon catalysts. Reproduced with permission from Ref [130].

12.3  Alkane Dehydrogenation

Recently, Li et al. clarified this mechanism through DFT calculations combined with microkinetic modeling [133]. The simulation showed that carbonyl groups are the main active sites for the first hydrogen abstraction from the alkane. On the other hand, activated oxygenated species (i.e. O2* and/or HO2� free radical) are responsible for the abstraction of the second hydrogen from the hydrocarbon intermediate, hence actively contributing to the overall reaction rate. Among the possible mechanisms, these authors found that a radical path can be invoked when the process is run with a stoichiometric amount of molecular oxygen, whereas it must be excluded under standard conditions. Finally, they showed that defective sites do not affect the intrinsic catalyst performance directly, but are rather converted into carbonyl groups under the applied reaction conditions, increasing the number of active sites available [133]. Experimentally, Stowers et al. supplied direct evidence of the detrimental effect of carboxylic/anhydride moieties on ODH selectivity by submitting mildly oxidized CNTs to a preliminary H2 treatment to reduce carboxylic groups to phenolic moieties. After the material pre-treatment, the authors observed an increase of about 80% and 30% in selectivity and yield, respectively, unambiguously demonstrating the negative influence of electrophilic oxygen species on the process [134]. The main problem related to the use of carbon nanomaterials for ODH remains the low alkene selectivity (~50% for EB to ST) due to alkane cleavage and over-oxidation paths at high temperature and under excess oxygen (necessary to reach high alkanes conversion). In 2016, Su’s group performed the EB-to-ST ODH under oxygen-lean conditions (EB/O2 = 5) using NDs as catalytic materials [135]. The authors demonstrated that the process reaches EB conversion above 40% and unprecedented selectivity to ST up to 90%. The role of oxygen is to generate oxygenated active sites on NDs, whereas its optimized concentration in the feed raises the catalyst lifetime up to 500 h.

12.3.2  Alkane Dehydrogenation under Steam- and Oxygen-free Conditions: The DDH Reaction Despite the important technical limits and safety concerns associated to the use of oxygen as cofeed in the reagents stream, ODH remains the most widely exploited process, whereas much less work has been focused on unveiling the real potentiality of carbon-based catalysts in the alkaneto-alkene dehydrogenation operated under non-oxidative conditions. The first report on the alkane DDH catalysis promoted by metal-free carbon-based materials was presented in 2010 by Su et al., taking advantage of their background experience in ODH (see Section 12.3.1) [136]. As a first explorative study, they tested a series of classical carbon-based nanomaterials under steam-free conditions for the EB-to-ST DDH. As Figure 12.7 shows, their findings demonstrated the superior catalytic performance of NDs among the samples of this series. The authors link the superior activity of NDs to their unique sp2/sp3 hybrid structure, in particular to the presence of carbonyl groups present at the surface of sp2 carbons as Lewis basic sites for hydrocarbon activation (analogously to what already claimed for ODH). In a more recent study, they deepened the investigation by varying systematically the NDs surface microstructure through successive thermal annealing treatments at increasing temperatures. They found that a mixed sp2/sp3 structure combining a ND core with a defective sp2-surface (reactive for ketone-type functional group formation) shows the highest DDH activity if compared with ND@[amorphous carbon] and ND@[ordered onion-like structure] composites. Indeed, the former bears detrimental surface electrophilic oxygen species (i.e. carboxylic groups and anhydrides) whereas the latter shows a low concentration of defective sites. In both cases, this translates into a modest catalytic activity [137]. They confirmed that carbonyls anchored on defective sites are again the active groups for DDH of alkanes. In addition, they tentatively assumed that, in the absence of

237

238

12  Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes

Figure 12.7  Steady-state activity of various carbon materials. Reaction conditions: 0.05 g, 550 °C, 2.8% EB (diluted in helium), 10 mL/min. Reproduced with permission from Ref [136].

an external O2 supply and when all the oxygen-containing functional groups are consumed by the process, defective sites/vacancies themselves serve as active sites for the C–H bond activation. In partial contrast to these preliminary conclusions, Yuan’s group exploited a series of mesoporous carbon materials activated with HNO3 [138] or HCl [139] as catalysts for propane DDH. In their study, they demonstrated the beneficial effect raising from large and well-ordered porosity in terms of mass transport; again, they confirmed the role of carbonyl groups as active species. However, they also postulated that, in the absence of extra-oxygen in the reactants stream, ketonic groups are regenerated throughout a spontaneous thermal dehydrogenation path [138–140]. Several studies appeared later in the literature devoted to the identification of DDH active sites agreed on the key role played by C=O surface moieties [141–144], although their regeneration mechanism under oxygen-free conditions remains a controversial matter. Pham-Huu’s group recovered a hybrid nanocarbon material (i.e. NDs distributed on GO/few-layer graphene composites) after several hours on DDH conditions and demonstrated that their initial activity is completely restored only upon calcination on air [141]. An oxidative treatment is therefore necessary to recover the initial catalytic activity. The catalyst activity “revival” is ascribed to the regeneration of carbonyl groups and to the coke deposits removal. The hypothesis of a spontaneous ketonic groups regeneration was therefore excluded, while the residual low DDH activity of the catalyst after several hours on stream was attributed to the carbonaceous coke deposits formed on the catalyst surface. Successively, Hao’s group reported that nanocarbons without any functional surface group are effective in alkane DDH and demonstrated that their activity is directly correlated with the overall materials SSA [145]. They attributed the commonly observed decrease of activity to the formation of coke deposits that simply cause a decrease of the SSA. The fact that nanocarbons reach a steady-state DDH conversion and maintain this value for hundreds of hours is indicative that coke itself has a residual (even if low) dehydrogenation activity. All these data taken together suggested that the presence of carbonyl surface groups does not seem to be essential for DDH, that it is more realistically attributed to a combination of structural (porosity, pore volume, SSA) and morphological (surface defective sites) factors. Oxygenated groups (if present) contribute to the initial nanocarbons activity, as demonstrated by the direct correlation between alkene formation rate and the amount of carbonyl groups found by Yuan’s et al. [142, 146]. However, the activity observed at the steady-state conditions is more reliably due to defective carbon sites without any oxygenated moiety. The high activity of NDs in alkane DDH makes them particularly suitable for replacing traditional metal-based catalysts. However, their powdery form hampers their safe handling and

12.3  Alkane Dehydrogenation

utilization, potentially causing reactors blocks and detrimental pressure drops phenomena. To overcome this problem, Su [147, 148] and Pham-Huu’s [149] independently developed macroscopic monolithic supports to immobilize and disperse NDs, avoiding their aggregation and concomitant loss of catalytic activity. Both groups used a non-oxide ceramic (silicon carbide, SiC) as a macroscopic support for NDs because of its ideal properties [35]: high thermal conductivity, mechanical stability and open meso- and macro-porosity. Indeed, the NDs/SiC composite reaches higher activity and greater stability on stream with respect to pure NDs. As a further step forward, Zhao et al. prepared NDs/CNx hybrid composites through pyrolysis of NDs and melamine [150]. Starting from their preliminary results with bare CNx in EB DDH, in which they proved a promotional effect of N-atoms on the process [151], they added NDs in different ratios and demonstrated a clear synergistic effect between the two materials. DDH activity of the NDs/CNx composites is nearly doubled with respect to the single components tested under the same conditions. Authors attributed this effect to an increase of electron density and basicity due to the nitrogen incorporation into the carbon matrix [150, 152]. The evidence of the beneficial effects of carbon nanomaterials N-doping on alkane DDH activity and durability boosted an intense production of different N-doped nanocarbons with a variety of morphology and chemical compositions that demonstrated superior performance with respect to their undoped counterparts [153–162]. All of these studies highlighted the key role of nitrogen in increasing the basic and nucleophilic properties of the materials, enhancing the ability of active sites (recognized again as carbonyl groups and defect/edge) to interact with the C–H bond of alkanes and lower the process activation energy [161]. In the same line, Lv and Li’s group highlighted in 2016 that N-doping in oxidized CNTs increases the DDH performance because of the heteroatoms ability at enhancing the adsorption strength of the starting alkane while reducing that of the alkene product. The active sites were still identified as the ketonic species [163]. In contrast to all previous reports on the topic, Yuan et al. found that N-insertion in mesoporous carbon networks has a detrimental effect on the alkane DDH activity with worse productivity if compared with that of their undoped counterparts. They speculated that N-doping causes a decrease of the ketonic active groups content that is not compensated by the electron donating effect of nitrogen [164]. Beyond this singular case, the insertion of light heteroatoms in a carbon framework is globally recognized to foster the alkane DDH activity and, even if nitrogen remains the most commonly used, phosphorus [164–166], boron [164], and sulfur [167] were also found to improve nanocarbon performance through an increase in active carbonyl/quinone groups content and/or a beneficial electronic donation. Giambastiani and Pham-Huu’s groups first demonstrated in 2016 that N-doped carbon material morphology also plays a key role in alkane DDH activity and selectivity. They found that the overall SSA and microporous area have only marginal effects on the process, whereas a high density of small mesopores (3–5 nm) favors DDH selectivity toward ST [168]. Moreover, the higher surface basicity induced by N-sites limits the occurrence of alkane cracking side reactions (typically originating from the presence of oxygenated acid edge-site defects) [150, 160, 169], ensuring better catalyst stability and durability on stream with respect to its undoped counterpart [168, 170]. In 2017, the same groups reported a unique alkane DDH performance by exploiting highly porous N-rich organic polymers as catalysts [171]. In particular, they tested for the first time in the process a series of CTFs, which are easily prepared through the cyano-aryl cyclotrimerization reaction [171, 172]. A selected sample from this series (CTF-ph, Figure 12.8) shows an unprecedented stability on stream without any initial deactivation even under harsh reaction conditions, also outperforming the industrial Fe-based benchmark catalyst. An in-depth material characterization unveiled the key role played by the “chemically accessible” surface basicity for the inhibition of coke deposits formation, typically identified as the main reason for the deactivation of these catalytic systems on stream.

239

240

12  Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes

Figure 12.8  a and a′) Direct dehydrogenation (DDH) of ethylbenzene (EB) with (CTF)-ph and K-Fe catalysts at increasing reactor temperature, from 550 to 600 °C. b) DDH of EB with CTF-ph at increasing EB concentrations: 2.8, 5 and 10 vol.%. Reproduced with permission from Ref [171].

To complete the picture, a joint study from the same groups has recently outlined the importance of a complementary role for the coke deposits generated during alkane DDH [173]. With the unique exception of the aforementioned case, carbonaceous deposits (coke) formed during alkane DDH are commonly claimed to cause a progressive catalyst fouling and in turn its deactivation. Although “coke” dehydrogenation activity was already known [174, 175], these authors exploited its performance in the EB-to-ST DDH under conditions close to those commonly used in industrial plants (600 °C, 10 vol.% EB/He, GHSV = 3000 h−1), simply using a “coked” γ-Al2O3-composite as catalyst. Their study demonstrated that highly ordered-graphitized coke grown during the process and under rigorous non-oxidative conditions enhances the DDH activity remarkably. Besides offering a complementary perspective on the role of “coke” in DDH catalysis, their finding indirectly confirmed the importance of high-energy C-sites in coke deposits, in the form of oxygen-free surface defects/vacancies.

12.4  Conclusions Carbon-based catalysts in the form of pure C-networks or light heteroatom-doped systems have witnessed a wonderful technological renaissance in catalysis. Many earth-abundant C-based 1D-3D porous architectures have shown unique performance as single-phase and metal-free catalysts in a number of industrially relevant transformations. In particular, they have also been successfully exploited in two challenging oxidation processes at the heart of petrochemical industry: the selective H2S oxidation to elemental sulfur for the natural tail gas purification and the alkane dehydrogenation to olefins. The ever-growing interest in the development of straightforward synthetic procedures to obtain efficient catalytic materials starting from cheap and earth abundant elements has opened new horizons in sustainable catalysis. The excellent outcomes reached so far stem from the joint talents of synthetic chemists and experts in materials science and catalysis. In addition, advanced characterization techniques and theoretical calculations with high levels of accuracy have marked a real step-forward toward a better understanding of the nature and composition of the active sites at work. This has set the basis of the future rational design of more efficient and selective metal-free catalytic materials that will deeply affect the industrial transition in the years to come.

References

Acknowledgments G. G. and C. P.-H. thank the TRAINER project (Catalysts for Transition to Renewable Energy Future) of the “Make our Planet Great Again” program (Ref. ANR-17-MPGA-0017) for support. The Italian team would also like to thank the Italian MIUR through the PRIN 2017 Project Multi-e (20179337R7) “Multielectron transfer for the conversion of small molecules: an enabling technology for the chemical use of renewable energy” for financial support.

References 1 Guo, Z., Liu, B., Zhang, Q. et al. (2014). Chem. Soc. Rev. 43: 3480–3524. 2 Bhaduri, S. and Mukesh, D. (2000). Homogeneous Catalysis: Mechanisms and Industrial Applications, 247. John Wiley & Sons, Inc. 3 Liu, X. and Dai, L. (2016). Nat. Rev. Mater. 1: 16064. 4 Sadjadi, S. (2020). Emerging Carbon Materials for Catalysis, 431. Elsevier Inc. 5 Eow, J.S. (2002). Environ. Prog. 21: 143–162. 6 Faramawy, S., Zaki, T., and Sakr, A.A.-E. (2016). J. Nat. Gas Sci. Eng. 34: 34–54. 7 Khairulin, S., Kerzhentsev, M., Salnikov, A., and Ismagilov, Z.R. (2021). Catalysts 11: 1109. 8 Zhang, X., Tang, Y., Qu, S. et al. (2015). ACS Catal. 5: 1053–1067. 9 El-Bishtawi, R. and Haimour, N. (2004). Fuel Process. Technol. 86: 245–260. 10 Pieplù, A., Saur, O., Lavalley, J.C. et al. (1998). Catal. Rev.: Sci. Eng. 40: 409–450. 11 Yasyerli, S., Dogu, G., and Dogu, T. (2006). Catal. Today 117: 271–278. 12 Chun, S.W., Jang, J.Y., Park, D.W. et al. (1998). Appl. Catal. B, Environ. 16: 235–243. 13 Li, K.-T., Huang, M.-Y., and Cheng, W.-D. (1996). Ind. Eng. Chem. Res. 35: 621–626. 14 Li, K.-T., Yen, C.-S., and Shyu, N.-S. (1997). Appl. Catal. A, Gen. 156: 117–130. 15 Uhm, J.H., Shin, M.Y., Zhidong, J., and Chung, J.S. (1999). Appl. Catal. B, Environ. 22: 293–303. 16 Cariaso, O.C. and Walker, P.L. (1975). Carbon 13: 233–239. 17 Klein, J. and Henning, K.-D. (1984). Fuel 63: 1064–1067. 18 Primavera, A., Trovarelli, A., Andreussi, P., and Dolcetti, G. (1998). Appl. Catal. A, Gen. 173: 185–192. 19 Bagreev, A. and Bandosz, T.J. (2001). Carbon 39: 2303–2311. 20 Mikhalovsky, S.V. and Zaitsev, Y.P. (1997). Carbon 35: 1367–1374. 21 Xiao, Y., Wang, S., Wu, D., and Yuan, Q. (2008). Sep. Purif. Technol. 59: 326–332. 22 Bagreev, A. and Bandosz, T.J. (2002). Ind. Eng. Chem. Res. 41: 672–679. 23 Chen, Q., Wang, J., Liu, X. et al. (2011). Carbon 49: 3773–3780. 24 Long, D., Chen, Q., Qiao, W. et al. (2009). Chem. Commun. 3898–3900. 25 Chen, Q., Wang, J., Liu, X. et al. (2011). Microp. Mesop. Mater. 142: 641–648. 26 Sun, F., Liu, J., Chen, H. et al. (2013). ACS Catal. 3: 862–870. 27 Gong, K., Du, F., Xia, Z. et al. (2009). Science 323: 760–764. 28 Wang, J., Ke, C., Jia, X. et al. (2021). Appl. Catal. B, Env. 283: 119650. 29 Chizari, K., Deneuve, A., Ersen, O. et al. (2012). ChemSusChem. 5: 102–108. 30 Duong-Viet, C., Truong-Phuoc, L., Tran-Thanh, T. et al. (2014). Appl. Catal. A, Gen. 482: 397–406. 31 Ba, H., Duong-Viet, C., Liu, Y. et al. (2016). C. R. Chimie 19: 1303–1309. 32 Liu, Y., Duong-Viet, C., Luo, J. et al. (2015). ChemCatChem. 7: 2957–2964.

241

242

12  Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes

33 Pham-Huu, C., Giambastiani, G., Liu, Y. et al. (2016). Method for preparing highly nitrogen-doped mesoporous carbon composites  PCT/EP2016/051196. WO2016116542 (A1). Application number: 15152038.4. Publication date: 28-07-2016 Bulletin 2016/30. 34 Ba, H., Liu, Y., Truong-Phuoc, L. et al. (2015). Chem. Commun. 51: 14393–14396. 35 Tuci, G., Liu, Y., Rossin, A. et al. (2021). Chem. Rev. 121: 10559–10665. 36 Duong-Viet, C., Nhut, J.-M., Truong-Huu, T. et al. (2020). Catal. Sci. Technol. 10: 5487–5500. 37 Duong-Viet, C., Nhut, J.-M., Truong-Huu, T. et al. (2021). Catalysts 11: 226. 38 Xu, C., Gu, Q., Li, S. et al. (2021). ACS Catal. 11: 8591–8604. 39 Shen, L., Lei, G., Fang, Y. et al. (2018). Chem. Commun. 54: 2475–2478. 40 Lei, G., Cao, Y., Zhao, W. et al. (2019). ACS Sustainable Chem. Eng. 7: 4941–4950. 41 Lei, G., Dai, Z., Fan, Z. et al. (2019). Carbon 155: 204–214. 42 Liang, S., Mi, J., Liu, F. et al. (2020). Chem. Eng. Sci. 221: 115714. 43 Mi, J., Liu, F., Chen, W. et al. (2019). ACS Appl. Mater. Interfaces 11: 29950–29959. 44 Liu, X., Zhangsun, G., Zheng, Y. et al. (2021). Ind. Eng. Chem. Res. 60: 2101–2111. 45 Kan, X., Zhang, G., Luo, Y. et al. (2022). Green Energy Environ. 7 (5): 983–995. doi:10.1016/j. gee.2020.1012.1016. 46 Ghasemy, E., Motejadded, H.B., rashidi, A. et al. (2018). J. Taiwan Inst. Chem. Eng. 85: 121–131. 47 Kamali, F., Eskandari, M.M., Rashidi, A. et al. (2019). J. Hazard. Mater. 364: 218–226. 48 Zhang, X., Xu, C., Li, S. et al. (2021). Appl. Mater. Today 25: 101228. 49 Sun, M., Wang, X., Pan, X. et al. (2019). Fuel Process. Technol. 191: 121–128. 50 Li, S., Liu, Y., Gong, H. et al. (2019). ACS Appl. Nano Mater. 2: 3780–3792. 51 Yang, C., Ye, H., Byun, J. et al. (2020). Environ. Sci. Technol. 54: 12621–12630. 52 Xu, C., Chen, J., Li, S. et al. (2021). J. Hazard. Mater. 403: 123806. 53 Chen, L., Yuan, J., Li, T. et al. (2021). Sci. Total Environ. 768: 144452. 54 Li, S., Gu, Q., Cao, N. et al. (2020). J. Mater. Chem. A 8: 8892–8902. 55 Peng, W.-L., Kan, X., Chen, W. et al. (2021). ACS Appl. Mater. Interfaces 13: 34124–34133. 56 Duong-Viet, C., Liu, Y., Ba, H. et al. (2016). Appl. Catal. B, Env. 191: 29–41. 57 Xu, Z., Duong-Viet, C., Ba, H. et al. (2018). Catalysts 8: 145. 58 Wang, W., Tuci, G., Duong-Viet, C. et al. (2019). ACS Catal. 9: 7921–7935. 59 Truong-Huu, T., Duong-Viet, C., Duong-The, H. et al. (2021). Appl. Catal. A, Gen. 620: 118171. 60 Bian, C., Gao, Q., Zhang, J. et al. (2019). Sci. Total Environ. 695: 133875. 61 Wang, X., Zhang, W., Gao, Q. et al. (2019). Appl. Surf. Sci. 470: 1010–1017. 62 Meyers, R.A. (2005). Handbook of Petrochemicals Production Processes, 11.13–11.34. New York: McGraw-Hill Handbooks. 63 Carter, J.H., Bere, T., Pitchers, J.R. et al. (2021). Green Chem. 23: 9747–9799. 64 Cavani, F. and Trifiro, F. (1995). Appl. Catal. A, Gen. 133: 219–239. 65 Data Bridge Market Research. Global olefins market – industry trends and forecast to 2029. https:// www.databridgemarketresearch.com/reports/global-olefins-market (accessed 15 January 2022). 66 Argyle, M.D. and Bartholomew, C.H. (2015). Catalysts 5: 145–263. 67 Nesterenko, N.S., Ponomoreva, O.A., Yuschenko, V.V. et al. (2003). Appl. Catal. A, Gen. 254: 261–272. 68 Kotarba, A., Kruk, I., and Sojka, Z. (2004). J. Catal. 221: 650–652. 69 Sheng, J., Yan, B., Lu, W.-D. et al. (2021). Chem. Soc. Rev. 50: 1438–1468. 70 Murakami, Y., Iwayama, K., Uchida, H. et al. (1981). J. Catal. 71: 257–269. 71 Tagawa, T., Hattori, T., and Murakami, Y. (1982). J. Catal. 75: 56–65. 72 Emig, G. and Hofman, H. (1983). J. Catal. 84: 15–26. 73 Vrieland, G.E. (1988). J. Catal. 111: 1–13.

References

74 Cadus, L.E., Arrua, L.A., Gorriz, O.F., and Rivarola, J.B. (1988). Ind. Eng. Chem. Res. 27: 2241–2246. 75 Iwasawa, Y., Nobe, H., and Ogasawara, S. (1973). J. Catal. 31: 444–449. 76 Cadus, L.E., Gorriz, O.F., and Rivarola, J.B. (1990). Ind. Eng. Chem. Res. 29: 1143–1146. 77 Fiedorow, R., Przystajko, W., Sopa, M., and Dalla Lana, I.G. (1981). J. Catal. 68: 33–41. 78 Grunewald, G.C. and Drago, R.S. (1990). J. Mol. Catal. 58: 227–233. 79 Drago, R.S. and Jurczyk, K. (1994). Appl. Catal. A, Gen. 112: 117–124. 80 Guerrero-Ruiz, A. and Rodriguez-Ramos, I. (1999). Carbon 32: 23–29. 81 Pereira, M.F.R., Orfao, J.J.M., and Figueiredo, J.L. (1999). Appl. Catal. A, Gen. 184: 153–160. 82 Pereira, M.F.R., Orfao, J.J.M., and Figueiredo, J.L. (2000). Appl. Catal. A, Gen. 196: 43–54. 83 Velasquez, J.J.D., Suarez, L.M.C., and Figueiredo, J.L. (2006). Appl. Catal. A, Gen. 311: 51–57. 84 Pereira, M.F.R., Orfao, J.J.M., and Figueiredo, J.L. (2002). Carbon 40: 2393–2401. 85 Pereira, M.F.R., Orfao, J.J.M., and Figueiredo, J.L. (2001). Appl. Catal. A, Gen. 218: 307–318. 86 Mestl, G., Maksimova, N.I., Keller, N. et al. (2001). Angew. Chem. Int. Ed. 40: 2066–2068. 87 Keller, N., Maksimova, N.I., Roddatis, V.V. et al. (2002). Angew. Chem. Int. Ed. 41: 1885–1888. 88 Su, D.S., Maksimova, N.I., Delgado, J.J. et al. (2005). Catal. Today 102-103: 110–114. 89 Macia-Agullo, J.A., Cazorla-Amoros, D., Linares-Solano, A. et al. (2005). Catal. Today 102-103: 248–253. 90 Sui, Z.-J., Zhou, J.-H., Dai, Y.-C., and Yuan, W.-K. (2005). Catal. Today 106: 90–94. 91 Zhao, T.-J., Sun, W.-Z., Gu, X.-Y. et al. (2007). Appl. Catal. A, Gen. 323: 135–146. 92 Zhang, J., Su, D., Zhang, A. et al. (2007). Angew. Chem. Int. Ed. 46: 7319–7323. 93 Su, D., Maksimova, N.I., Mestl, G. et al. (2007). Carbon 45: 2145–2151. 94 Zhang, J., Liu, X., Blume, R. et al. (2008). Science 322: 73–77. 95 Frank, B., Zhang, J., Blume, R. et al. (2009). Angew. Chem. Int. Ed. 48: 6913–6917. 96 Frank, B., Morassutto, M., Schomacker, R. et al. (2010). ChemCatChem. 2: 644–648. 97 Sun, X., Ding, Y., Zhang, B. et al. (2015). ACS Catal. 5: 2436–2444. 98 Huang, R., Liu, H.Y., Zhang, B.S. et al. (2014). ChemSusChem. 7: 3476–3482. 99 Huang, R., Wang, J., Zhang, B. et al. (2018). Catal. Sci. Technol. 8: 1522–1527. 100 Huang, R., Liang, C.H., Su, D.S. et al. (2015). Catal. Today 249: 161–166. 101 Liu, X., Frank, B., Zhang, W. et al. (2011). Angew. Chem. Int. Ed. 50: 3318–3322. 102 Sun, X., Wang, R., Zhang, B. et al. (2014). ChemCatChem. 6: 2270–2225. 103 Rinaldi, A., Zhang, J., Frank, B. et al. (2010). ChemSusChem. 3: 254–260. 104 Zhang, J., Wang, X., Su, Q. et al. (2009). J. Am. Chem. Soc. 131: 11296–11297. 105 Guo, X., Qi, W., Liu, W. et al. (2017). ACS Catal. 7: 1424–1427. 106 Huang, R., Xu, J., Wang, J. et al. (2016). Carbon 96: 631–640. 107 Qi, W., Liu, W., Guo, X. et al. (2015). Angew. Chem. Int. Ed. 54: 13682–13685. 108 Wen, G., Diao, J., Wu, S. et al. (2015). ACS Catal. 5: 3600–3608. 109 Chen, C., Zhang, J., Zhang, B. et al. (2013). Chem. Commun. 49: 8151–8153. 110 Hu, X., Wu, Y., Li, H., and Zhang, Z. (2010). J. Phys. Chem. C 114: 9603–9607. 111 Mao, S., Li, B., and Su, D. (2014). J. Mater. Chem. A 2: 5287–5294. 112 Szewczyk, I., Rokicinska, A., Michalik, M. et al. (2020). Chem. Mater. 32: 7263–7273. 113 Marco, Y., Roldan, L., Munoz, E., and Garcia-Bordeje, E. (2014). ChemSusChem. 7: 2496–2504. 114 Goyal, R., Sarkar, B., Bag, A. et al. (2016). J. Mater. Chem. A 4: 18559–18569. 115 Sheng, J., Yan, B., He, B. et al. (2020). Catal. Sci. Technol. 10: 1809–1815. 116 Malaika, A., Rechnia, P., Krzyzynska, B., and Kozlowski, M. (2012). Microp. Mesop. Mater. 163: 300–306. 117 Martin-Sanchez, N., Soares, O.S.G.P., Pereira, M.F.R. et al. (2015). Appl. Catal. A, Gen. 502: 71–77.

243

244

12  Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes

118 Pelech, I., Soares, O.S.G.P., Pereira, M.F.R., and Figueiredo, J.L. (2015). Catal. Today 249: 176–183. 119 Diao, J., Liu, H., Wang, J. et al. (2015). Chem. Commun. 51: 3423–3425. 120 Zhang, Y., Diao, J., Rong, J. et al. (2018). ChemSusChem. 11: 536–541. 121 Liang, C., Xie, H., Schwartz, V. et al. (2009). J. Am. Chem. Soc. 131: 7735–7741. 122 Schwartz, V., Fu, W., Tsai, Y.-T. et al. (2013). ChemSusChem. 6: 840–846. 123 Dathar, G.K.P., Tsai, Y.-T., Gierszal, K. et al. (2014). ChemSusChem. 7: 483–491. 124 Janus, P., Janus, R., Kuśtrowski, P. et al. (2014). Catal. Today 235: 201–209. 125 Wegrzyniak, A., Jarczewski, S., Kuśtrowski, P., and Michorczyk, P. (2018). J. Porous Mater. 25: 687–696. 126 Kwon, H.C., Yook, S., Choi, S., and Choi, M. (2017). ACS Catal. 7: 5257–5267. 127 Wang, T., Chong, S., Wang, T. et al. (2018). Appl. Surf. Sci. 427: 1011–1018. 128 Roldan, L., Benito, A.M., and Garcia-Bordeje, E. (2015). J. Mater. Chem. A 3: 24379–24388. 129 Du, P., Zhang, -X.-X., Zhang, S. et al. (2021). ChemCatChem. 13: 610–616. 130 Liu, W., Wang, C., Su, D., and Qi, W. (2018). J. Catal. 368: 1–7. 131 Tang, S. and Cao, Z. (2012). Phys. Chem. Chem. Phys. 14: 16558–16565. 132 Cao, L., Dai, P., Zhu, L. et al. (2020). Appl. Catal. B, Environ. 262: 118277. 133 Lian, Z., Si, C., Jan, F. et al. (2020). ACS Catal. 10: 14006–14014. 134 Zhou, Z., Orcutt, E.K., Anderson, H.C., and Stowers, K.J. (2019). Carbon 152: 924–931. 135 Diao, J., Feng, Z., Huang, R. et al. (2016). ChemSusChem. 9: 662–666. 136 Zhang, J., Su, D.S., Blume, R. et al. (2010). Angew. Chem. Int. Ed. 49: 8640–8644. 137 Wang, R., Sun, X., Zhang, B. et al. (2014). Chem. Eur. J. 20: 6324–6331. 138 Liu, L., Deng, Q.-F., Liu, Y.-P. et al. (2011). Catal. Commun. 16: 81–85. 139 Liu, L., Deng, Q.-F., Agula, B. et al. (2012). Catal. Today 186: 35–41. 140 Liu, L., Deng, Q.-F., Agula, B. et al. (2011). Chem. Commun. 47: 8334–8336. 141 Tran Thanh, T., Ba, H., Truong-Phuoc, L. et al. (2014). J. Mater. Chem. A 2: 11349–11357. 142 Hu, Z.-P., Chen, C., Ren, J.-T., and Yuan, Z.-Y. (2018). Appl. Catal. A, Gen. 559: 85–93. 143 Hu, Z.-P., Zhao, H., Chen, C., and Yuan, Z.-Y. (2018). Catal. Today 316: 214–222. 144 Hu, Z.-P., Zhang, L.-F., Wang, Z., and Yuan, Z.-Y. (2018). J. Chem. Technol. Biotechnol. 93: 3410–3417. 145 Li, Y., Zhang, Z., Wang, J. et al. (2015). Chin. J. Catal. 36: 1214–1222. 146 Hu, Z.-P., Ren, J.-T., Yang, D. et al. (2019). Chin. J. Catal. 40: 1385–1394. 147 Liu, H., Diao, J., Wang, Q. et al. (2014). Chem. Commun. 50: 7810–7812. 148 Diao, J., Liu, H., Feng, Z. et al. (2015). Catal. Sci. Technol. 5: 4950–4953. 149 Ba, H., Liu, Y., Mu, X. et al. (2015). Appl. Catal. A, Gen. 499: 217–226. 150 Zhao, Z. and Dai, Y. (2014). J. Mater. Chem. A 2: 13442–13451. 151 Zhao, Z., Dai, Y., Lin, J., and Wang, G. (2014). Chem. Mater. 26: 3151–3161. 152 Ge, G. and Zhao, Z. (2019). Appl. Catal. A, Gen. 571: 82–88. 153 Wang, J., Liu, H., Diao, J. et al. (2015). J. Mater. Chem. A 3: 2305–2313. 154 Zhao, Z., Dai, Y., Ge, G. et al. (2015). RSC Adv. 5: 53095–53099. 155 Zhao, Z., Dai, Y., and Ge, G. (2015). Catal. Sci. Technol. 5: 1548–1557. 156 Zhao, Z., Dai, Y., Ge, G. et al. (2015). ChemCatChem. 7: 1070–1077. 157 Zhao, Z., Dai, Y., Ge, G., and Wang, G. (2015). ChemCatChem. 7: 1135–1144. 158 Zhao, Z., Dai, Y., Ge, G., and Wang, G. (2015). AIChE J. 61: 2543–2561. 159 Shi, L., Qi, W., Liu, W. et al. (2018). Catal. Today 301: 48–54. 160 Zhao, Z., Dai, Y., Ge, G. et al. (2015). Green Chem. 17: 3723–3727. 161 Tian, S., Yan, P., Li, F. et al. (2019). ChemCatChem. 11: 2073–2078. 162 Zhou, Q., Guo, X., Song, C., and Zhao, Z. (2019). ACS Appl. Nano Mater. 2: 2152–2159.

References

163 Mu, J., France, L.J., Liu, B. et al. (2016). Catal. Sci. Technol. 6: 8562–8570. 164 Song, Y., Liu, G., and Yuan, Z.-Y. (2016). RSC Adv. 6: 94636–94642. 165 Li, L., Zhu, W., Liu, Y. et al. (2015). RSC Adv. 5: 56304–56310. 166 Pan, S.-F., Yin, J.-L., Zhu, X.-L. et al. (2019). Carbon 152: 855–864. 167 Zhou, Q. and Zhao, Z. (2020). ChemCatChem. 12: 342–349. 168 Ba, H., Liu, Y., Truong-Phuoc, L. et al. (2016). ACS Catal. 6: 1408–1419. 169 Gounder, R. and Iglesia, E. (2009). J. Am. Chem. Soc. 131: 1958–1971. 170 Ba, H., Luo, J., Liu, Y. et al. (2017). Appl. Catal. B, Environ. 200: 343–350. 171 Tuci, G., Pilaski, M., Ba, H. et al. (2017). Adv. Funct. Mater. 27: 1605672. 172 Tuci, G., Iemhoff, A., Ba, H. et al. (2019). Beilstein J. Nanotechnol. 10: 1217–1227. 173 Ba, H., Tuci, G., Evangelisti, C. et al. (2019). ACS Catal. 9: 9474–9484. 174 Nederlof, C., Kaptejin, F., and Makkee, M. (2012). Appl. Catal. A, Gen. 417-418: 163–173. 175 McGregor, J., Huang, Z., Parrott, E. et al. (2010). J. Catal. 269: 329–339.

245

247

13 Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment Cutting-Edge Solutions for a Sustainable World Clara Pereira, Diana M. Fernandes, Andreia F. Peixoto, Marta Nunes, Bruno Jarrais, Iwona Kuźniarska-Biernacka, and Cristina Freire REQUIMTE/LAQV, Departamento de Química e Bioquímica, Faculdade de Ciências, Universidade do Porto, Rua do Campo Alegre s/n, Porto, Portugal

13.1  Introduction In the last decades, there has been a remarkable progress on the development of carbon-based (nano)materials for catalytic applications owing to their extraordinary properties, namely the variety of morphologies ranging from 0D to 3D (e.g., carbon dots [CDs], graphene, carbon nanotubes [CNTs], activated carbon [AC], biochars [BCHs]), tunable textural, electrical and surface chemistry properties, and high thermal and mechanical stability [1, 2]. The combination or hybridization of carbon-based (nano)materials with transition metal oxides, namely those with magnetic, redox-active, and/or semiconductive properties allows developing new engineered (nano)catalysts with synergistically-enhanced functionalities, which are shaping the field of Green Catalysis toward a more Sustainable World. In particular, carbon-metal oxide hybrid and composite materials have been demonstrating their widespread potential as efficient and recyclable heterogeneous catalysts, photocatalysts, and electrocatalysts for a myriad of eco-sustainable catalytic applications, ranging from energy-related technologies and biomass valorization to wastewater treatment [3]. The carbon (nano)materials act as supports for the metal oxide (nano)particles, preventing their aggregation and thus increasing the amount of exposed catalytically active sites. During the catalytic processes, due to its adsorption properties, the carbon support allows the adsorption and concentration of the target molecules near the catalytically active sites of the immobilized metal oxides, improving the interfacial contact between the substrates/oxidant/reducing agent and the active centers. Moreover, its electrical conductivity and/or photosensitizing properties can, in some types of reactions, facilitate the charge transfer phenomena. On the other hand, the grafted metal oxides endow or upgrade the (photo)(electro)catalytic properties of the hybrid/composite materials [3–7]. Thus, the conjugation of both building blocks can lead to advanced multifunctional materials that act both as adsorbents and catalysts and present enhanced performance relative to that of their individual components. Within the context of energy, the current high energy demand and the continuous exhaustion of fossil fuels have been highly contributing to environmental problems due to Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 1, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

248

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

the emission of dangerous gases to the atmosphere. These challenges have prompted the development of efficient electrocatalysts for energy conversion/storage technologies, including fuel cells (FCs), water splitting (WS) devices, batteries/supercapacitors, and, more recently, the electrochemical reduction of CO2 into value-added products [8, 9]. Still, the high cost of several components of such technologies, namely of the electrocatalysts, and the sluggish kinetics of the reactions occurring at the cathode (or anode) in many of these systems are major cornerstone issues that hinder their large-scale implementation. These drawbacks have been stimulating the search for alternative, cost-effective, stable and more eco-sustainable electrocatalysts, in which hybrid carbon-metal oxides stand out as promising solutions for oxygen reduction, oxygen evolution, hydrogen evolution, and CO2 reduction reactions (CO2RRs) [10, 11]. Moreover, the production of energy and chemicals in biorefineries has gained considerable attention to minimize petroleum dependence and the environmental issues associated with the massive utilization of fossil resources. The biorefinery concept has emerged as a strategy to manufacture a new generation of value-added chemicals, materials, and fuels with limited environmental footprints with the aim to sustainably meet the energy future requirements on a bio-based economy [12]. The design of novel catalysts plays a key role, with bio-based carbon catalysts representing one of the most promising class of materials within the biorefinery context for biomass valorization toward valuable chemicals/fuels [13, 14]. In parallel, the continuous growth of the worldwide population has fomented a wide demand for resources and amenities and has led to the release of a diversity of persistent recalcitrant organic chemicals into water/wastewater daily (e.g. industrial dyes, pharmaceuticals, personal care products, and pesticides) [15]. Some of these products, when at the end of their life-times, are inefficiently handled by the common water/wastewater treatment technologies and inevitably contaminate water and soils, constituting a major threat to ecosystems and to health. To overcome this bottleneck, advanced oxidation processes (AOPs), namely Fenton-based technologies (heterogeneous Fenton, photo-Fenton, and electro-Fenton) and photocatalysis have deserved particular attention [16]. A distinct strategy for wastewater treatment consists on the catalytic reduction of hazardous organic pollutants found in water and wastewater, such as nitroaromatic compounds, into compounds with high added-value and lower toxicity to be used in the chemical industry for the synthesis of dyes, agrochemicals, pharmaceuticals, polymers, and other fine chemicals [17, 18]. The development of efficient, sustainable, and low-cost hybrid carbon-metal oxide catalysts for such advanced (photo)catalytic oxidation and reduction processes is an asset to promote the widespread application of these wastewater treatment technologies. This chapter provides an overview of the fascinating advances on the development and application of engineered carbon-metal oxide hybrid and composite (nano)catalysts in state-of-the-art and emerging sustainable catalytic processes (Figure 13.1): (i) as electrocatalysts in energyrelated reactions (Section 13.2), (ii) in biomass valorization for the production of various value-added biochemicals, biomaterials, and biofuels in the context of modern biorefineries (Section 13.3), and (iii) in advanced (photo)(electro)catalytic oxidation and reduction processes for water and wastewater treatment (Sections 13.4 and 13.5, respectively). The chapter will end with the future challenges and directions to further advance the rational design of high-­ performance eco-sustainable carbon-metal oxide (nano)catalysts for the different processes explored throughout the chapter.

13.2  Hybrid Carbon-metal Oxide Electrocatalysts for Energy-related Applications

Figure 13.1  Overview of the applications of hybrid carbon-metal oxide materials in electrocatalysis, biomass valorization, and wastewater treatment.

13.2  Hybrid Carbon-metal Oxide Electrocatalysts for Energy-related Applications In the last decade, hybrid carbon-metal oxide (nano)materials have gained particular relevance as electrocatalysts since they play an essential role in many energy conversion and storage technologies, such as FCs, WS devices, batteries, supercapacitors, and CO2 conversion into value-added products. Electrocatalysts play a key role on several electrochemical reactions, including oxygen reduction and oxygen evolution reactions, the hydrogen evolution reaction, and the CO2RR. Carbon-based materials, besides their numerous advantageous properties, provide a support to metal oxide (nano)particles, preventing their aggregation and leading to more exposed active sites, which makes them more accessible to the (electro)catalytic medium; in this sense, the resulting carbon-metal oxide electrocatalysts better perform their functions when compared to the unsupported metal oxides [7].

13.2.1  Oxygen Reduction Reaction (ORR) The oxygen reduction reaction (ORR) is merely the electrochemical interconversion between O2, H2O2 and H2O. However, it involves many steps and proceeds through two different pathways in both acidic and alkaline media, as summarized in Table 13.1. A direct O2 reduction to H2O is a 4-electron process, whereas the indirect O2 reduction occurs in two steps and produces H2O2. The direct process is comprehended for efficient ORR. The nature of the electrocatalyst has a great impact on the ORR routes, affecting the O2 adsorption and the interactions of O2− with the catalyst surface. Pt is considered the best ORR

249

250

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

Table 13.1  ORR pathways (adapted from Ref [19]).

Electrolyte

Reduction potential (V vs. NHE)

Reaction

Alkaline (aqueous)4-electron pathway: O2 + 2H2O + 4e− → 4HO−

+0.401

2-electron pathway: O2 + H2O + 2e− → HO2− + OH−

−0.065

HO2− + H2O + 2e− → 3HO−

+0.867

or decomposition: HO2− → 2HO− + O2 Acidic (aqueous)4-electron pathway: O2 + 4H+ + 4e− → 2H2O

+1.229

2-electron pathway: O2 + 2H+ + 2e− → H2O2 +

−

H2O2 + 2H  + 2e → 2H2O

+0.67 +1.77

or decomposition: 2H2O2 → 2H2O + O2

electrocatalyst. The quantitative analysis of a catalyst is mainly established from the onset potential (Eonset), the diffusion-limiting current density (jL), the half-wave potential (E1/2), the electrochemically active surface area (ECSA), and the Tafel slope from the plot of E vs. log jk, where jk is the kinetic current density. A catalyst with more positive Eonset and E1/2 values and a lower Tafel slope than those of standard Pt/C is considered to be a good ORR catalyst. In practice, the high cost and poor stability of Pt and of other noble metal catalysts limit their large application as ORR catalysts. Thus, the development of cost-effective materials with promising electrochemical parameters is the current main challenge [19, 20]. Low-cost transition metal oxide-based composites exhibit better ORR stability than Pt catalyst, holding great potential over precious metal-based catalysts. Thus, the combination of nanostructured transition metal oxides and carbon materials is a promising approach to obtain inexpensive, highly efficient, and stable ORR electrocatalysts, some of them even with bifunctional electrocatalytic activity in ORR and oxygen evolution reaction (OER). (Nano)composites and (nano)hybrids of carbon materials and metal oxides (such as, Fe-, Co-, and Mn-based oxides) or polymetallic oxides, as well as those based on biomass-derived carbons and the materials derived from metal-organic frameworks (MOFs) will be focused in this section. Considering that several reviews related to this topic have already been published [19–23], only the most recent studies (since 2019) will be considered. As a first example, Feng et al. reported the preparation of a hybrid composed of maghemite (γ-Fe2O3) nanoparticles (NPs, 10 nm, spherical morphology) and reduced graphene oxide (rGO) by a simple and environmentally friendly one-pot hydrothermal strategy [24]. The γ-Fe2O3/rGO outperformed the ORR electrocatalytic activity of pristine γ-Fe2O3 and rGO, presenting larger reduction current density (−7.9 mA cm−2 vs.  ≈ −2.0 and −3.0 mA cm−2 for γ-Fe2O3 and rGO, respectively, at 3000 rpm) and more positive reduction potential (Eonset = 0.78 V, 0.60 V and 0.65 V for Fe2O3/rGO, γ-Fe2O3 and rGO, respectively, vs. the reversible hydrogen electrode [RHE]); moreover, the Eonset value was comparable to that of the commercial Pt/C electrode (0.85 V vs. RHE).

13.2  Hybrid Carbon-metal Oxide Electrocatalysts for Energy-related Applications

The oxygen reduction catalyzed by γ-Fe2O3/rGO followed a combination of 4-electron and 2-electron reduction pathways, with a number of electrons transferred in the ORR process near 3; moreover, the catalyst showed good electrocatalytic stability after 1000 cycles. The excellent ORR properties of the nanohybrid were ascribed to the strong coupling and synergistic effect between the γ-Fe2O3 NPs and the rGO sheets. Mn-based oxides are other typical class of electrocatalysts studied for ORR. Freire et al. studied the ORR electrocatalytic activity of new Mn3O4@oxidized graphene flake (GF) nanocomposites [25]. The materials were prepared by an in situ one-step coprecipitation route, using GFs that were previously selectively oxidized with HNO3 (preferable introduction of carboxylic groups and quinones), KMnO4 (introduction of epoxy, hydroxyl, carboxylic and carbonyl groups) and O3 (insertion of epoxy groups and quinones) as oxidizing agents, resulting in C/O atomic ratios of 21.2, 4.9, and 9.8, respectively. The loading of Mn3O4 NPs (30–38 nm, quasi-spherical morphology, spinel structure) was between 39.4 and 49.6 wt% in the nanocomposites. All the nanocomposites showed ORR electrocatalytic activity in alkaline medium (catalyst loading of 0.77 mg cm−2 on a glassy carbon rotating disk electrode [RDE]), with the Mn3O4@GF_O3 nanocomposite (i.e., using GF oxidized with O3 as support) presenting the least negative Eonset of −0.14 V vs. Ag/AgCl. Higher jL values were achieved for Mn3O4@GF_O3 and Mn3O4@GF_HNO3 nanocomposites (−2.8 mA cm−2). Mechanistically, Mn3O4@GF_O3 nanocomposite stood out, with a number of electrons involved in the process (nO2) close to 4 and the dominance of the one-step 4-electron transfer mechanism. All the nanocomposites showed robust electrocatalytic performance over 20000 s, with current retention values in the range of 87.0–90.3%, and excellent tolerance to methanol. Overall, the best ORR electrocatalytic performance of the Mn3O4@GF_O3 nanocomposite was explained by: (i) an adequate loading of Mn3O4 onto GF_O3 flakes, (ii) the highest relative content of Mn2+ ions, and (iii) the predominance of quinone and epoxy groups on GF_O3 support, which appear to have a key role on the overall electrocatalytic activity of this family of nanocomposites. Co-based oxides and their composites with carbon materials have been also widely tested as ORR electrocatalysts. Freire et al. explored the ORR (and OER) electrocatalytic activity of Co3O4 NPs anchored on GFs selectively oxidized (with KMnO4 and O3) by a coprecipitation route [26]. The characterization of the nanocomposites indicated a cubic phase Co3O4 spinel structure, cobalt contents of 6.11 mmol g−1 and 6.50 mmol g−1 for Co3O4@GF_KMnO4 and Co3O4@GF_O3, respectively, and a morphology based in flat sheets of oxidized graphene covered by agglomerates of Co3O4 NPs. For ORR (catalyst loading of 0.77 mg cm−2), Eonset values of 0.79 V and 0.82 V vs. RHE were achieved for Co3O4@GF_KMnO4 and Co3O4@GF_O3, respectively. Both nanocomposites presented excellent tolerance to methanol and, in the case of Co3O4@GF_O3, a superior long-term stability with a current retention of 96.8% after 20000 s. In general, mixed transition metal oxides with spinel-type (AB2O4) or perovskite-type (ABO3) structures (A and B represent different transition metal cations) show more promising ORR electrocatalytic activity than the corresponding single transition metal oxides, due to their higher electrical conductivity and richer redox properties. Thus, the incorporation of these metal oxides onto carbon materials with large specific surface area has been attempted to overcome the issues related with the activity and stability of the metal oxides [27]. Zhou et al. reported the first example on the ORR catalyzed by an N-doped porous carbon (NPC)-supported MnFeO2 catalyst (denoted as MnFeO2/NPC) with Pt-like activity [27]. The composite (BET surface area of 242.8 m2 g−2, pore volume of 0.32 cm3 g−1) was prepared by a simple agar-hydrogel strategy and was composed of ≈ 15 nm MnFeO2 NPs (Mn0.503Fe0.497O2 phase) uniformly immobilized in the porous NPC aerogel (which had a high graphitization degree, beneficial to improve the electronic conductivity).

251

252

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

Regarding the ORR catalytic activity, the MnFeO2/NPC catalyst exhibited an Eonset similar to that of commercial Pt/C (Eonset = 0.98 V vs. RHE for MnFeO2/NPC and Pt/C), E1/2 = 0.86 V (10 mV more positive than that of Pt/C), large limiting diffusion current density (5.9 mA cm−2), a small Tafel plot (68 mV dec−1 vs. 71 mV dec−1 for Pt/C) and an average electron-transfer number of 4.0 (HO2− yield 70 min); moreover, the reduction of Mn3+ by Fe2+ during the catalytic reaction benefited the redox cycles of Fe3+/Fe2+ and Mn3+/Mn2+. Zheng and co-authors prepared a Fe3O4-usushiol-rGO composite (Fe3O4-U-rGO) by a one-step solvothermal reaction (Figure 13.3). Urushiol monomer was used as reducing agent and crosslinker to promote the reduction of GO to rGO, the strong grafting of the Fe3O4 NPs to the support and to tune the size and improve the stability of the anchored Fe3O4 NPs (~8 nm, by XRD) [147]. The Fe3O4-U-rGO catalyst led to ~98% and ~99% of Rhodamine B (RhB) removal after 20 and 50 min, respectively, and to 60% and complete degradation of MB after 10 and 60 min, respectively (pH 3, room temperature, |RhB|0 = 60 ppm or |MB|0 = 20 ppm, Vdye solution = 100 mL, 20 mg of catalyst 2 mL of 30% H2O2). Moreover, it was highly stable upon reuse in six further cycles, preserving its performance due to the strong interaction between the Fe3O4 NPs and the urushiol-functionalized support, with negligible sludge formation. In fact, urushiol prevented the formation of catalytically inactive Fe3+ sludge and the decomposition of H2O2 (to form H2O and O2) during the Fenton-type reactions (Figure 13.3). In contrast, the catalyst prepared in the absence of urushiol lost ~ 60% of activity after seven catalytic cycles. Nevertheless, no mineralization studies were performed.

269

270

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

Figure 13.3  Preparation of Fe3O4-U-rGO heterogeneous Fenton-like catalyst for the degradation of Rhodamine B (RhB) and methylene blue (MB) dyes. Reproduced with permission from Ref [116].

The emerging graphene analogue, g-C3N4, has been attracting increased interest for catalytic applications. For instance, Ding and co-authors prepared g-C3N4-iron oxide composites by in situ synthesis of g-C3N4 in the presence of iron oxide NPs (α-Fe2O3 with traces of Fe3O4, 90–100 nm by transmission electron microscopy (TEM), previously prepared by hydrothermal process) and tested their performance in the dark Fenton oxidative degradation of ciprofloxacin (CIP) [148]. The catalyst with the best performance (prepared using iron oxide:dicyandiamide ratio of 1:5) led to full degradation of CIP after 45 minutes and 48.5% mineralization after 2 h in the optimum reaction conditions (pH 3.0, |CIP|0  =  20 mg L−1, VCIP solution  =  50 mL, |catalyst|  =  1 g L−1 corresponding to 50 mg of catalyst, |H2O2| = 0.0056 M). Moreover, leaching was minimized due to the confinement of the iron oxide NPs between the g-C3N4 layers, albeit the low pH. In contrast, the catalyst prepared by physical mixture of g-C3N4 and iron oxide presented poor catalytic performance. The high catalytic activity of g-C3N4-iron oxide composite was assigned to three factors: (i) the higher surface area of the catalyst (54.042 m2 g−1 vs. 14.537 and 8.705 m2 g−1 for g-C3N4 and unsupported iron oxide, respectively); (ii) the amount of catalytically active sites within the composite, owing to the intercalation of iron oxide NPs in the interlayer spacing of g-C3N4; and (iii) the enhanced electron transfer in Fe3+/Fe2+ redox couple. Nevertheless no recycling tests were performed. Other magnetically-recyclable g-C3N4-based composites were prepared by Lu et al. by in situ formation and growth of copper ferrite NPs on mesoporous carbon nitride (MCN) via a hydrothermal route (CuFe2O4:MCN weight ratios: 0.1, 0.2, 0.3, and 0.4) [149]. The optimized catalyst (prepared with 0.3 CuFe2O4:MCN weight ratio; BET surface area: 245 m2 g−1; CuFe2O4 particle size: 11.6 nm by XRD) led to almost total 4-chlorophenol (4-CP) degradation within 60 min and 59% total organic content (TOC) removal (pH 4, 55 ºC, |4-CP|0 = 100 mg L−1, |catalyst| = 1.0 g L−1, |H2O2| = 2 g L−1), revealing the lack of total oxidation of organic compounds, namely chlorinated intermediates, hydroquinone, formic acid, fumaric acid, acetic acid, and malonic acid. The catalyst presented 3 × higher catalytic activity than a physical mixture of CuFe2O4 and g-C3N4 components (pseudo-first-order rate constant, k  =  0.076 min−1 vs. 0.026 min−1) and 6×  higher than that of CuFe2O4 (k = 0.012 min−1). Finally, the performance and stability of the composite were preserved after two additional cycles. The high catalytic performance was assigned to the good dispersion of CuFe2O4 NPs over the MCN support and to the strong interaction between both components. The

13.4  Advanced (Photo)catalytic Oxidation Processes for Wastewater Treatment

authors proposed a Fenton-type mechanism for the catalytic process based on Fe3+/Fe2+ and Cu2+/ Cu+ redox cycles: the Fe3+ cations within CuFe2O4 activated H2O2 generating Fe2+, which then reacted with H2O2 to yield Fe3+ cations and HO• radicals; the resulting HO• radicals oxidized and degraded 4-CP; identically, the Cu2+/Cu+ pair could induce such redox cycle. Magnetic core-shell carbon-based nanomaterials have been also reported as efficient heterogeneous Fenton-type catalysts in the degradation of organic pollutants. For example, Qin et al. prepared core–shell MnFe2O4@C and MnFe2O4@C-NH2 (core diameter: 40 nm; carbon shell thickness: 2 nm) magnetic nanomaterials and tested their catalytic activity in the heterogeneous Fenton oxidation of three antibiotics (ofloxacin [OFX], amoxicillin [AMX], and tetracycline [TC]) in aqueous solution (30 ºC, |antibiotic|0 = 30 mg L−1, Vantibiotic solution = 100 mL, 100 mg of catalyst, 0.30 mL of 30 wt% H2O2) [150]. Firstly, the magnetic cores were produced by solvothermal process, followed by carbon coating (glucose precursor) through a hydrothermal route. MnFe2O4@C-NH2 was the most efficient catalyst, with enhanced catalytic activity relative to MnFe2O4@C and to the parent MnFe2O4. In the case of OFX degradation, MnFe2O4@C-NH2 led to 97.4% substrate removal and 62.5% TOC removal after 180 min at pH 3, with an apparent rate constant k of 1.46 h−1 vs. 0.29 h−1 and 0.47 h−1 for MnFe2O4 and MnFe2O4@C, respectively. The catalyst could also degrade OTX in neutral and alkaline media (pH 9), although the OTX removal efficiency decreased with the increase of the pH of the reaction medium, from 97.4% to 54.8% ongoing from pH 3 to pH 9; this decrease was assigned to the fast decomposition of H2O2 into O2 molecules instead of HO• radicals under basic conditions. The heterogeneous catalyst presented negligible metal leaching due to the presence of the carbon coating. MnFe2O4@C-NH2 was also tested in the removal of AMX and TC, leading to 99.1% and 97.4% antibiotic removal, respectively, with TOC removal values of 78.8% and 48.4%, respectively; thus, TC was the most refractory pollutant. The remarkable catalytic efficiency of MnFe2O4@C-NH2 was assigned to two main factors: the carbon shell and the presence of amine groups. The carbon shell increased the specific surface area of the material (from 25 m2 g−1 for the uncoated material to 62 m2 g−1 and 55 m2 g−1 for MnFe2O4@C and MnFe2O4@C-NH2, respectively), which increased the amount of exposed active sites for the decomposition of H2O2 into HO• radicals; moreover, it decreased the aggregation between the MNPs and prevented metal leaching. On the other hand, the presence of amine groups at the surface of the material increased the alkaline character of the material (point of zero charge, pHPZC = 8.45 vs. 6.75 and 6.98 for MnFe2O4 and MnFe2O4@C, respectively), which favored the decomposition of H2O2 and formation of HO• radicals; it also improved the electron density of the carbon shell, leading to higher electron transfer from the carbon shell to the metal ferrite, yielding more Fe2+ and Mn2+ cations, which were considered the primary active sites. Finally, MnFe2O4@C-NH2 was efficient in the treatment of a solution of mixed antibiotics (OFX:AMX:TC weight ratio of 1:1:1), leading to 63.8% of chemical oxygen demand (COD) removal efficiency after 180 min and to an increase of the 5-day oxygen demand (BOD5)/COD ratio from 0.012 to 0.36, indicating higher biodegradability (BOD5/ COD > 0.3 considered adequate for posterior biological treatment). Moreover, the composite preserved the catalytic efficiency in six consecutive cycles, with only a decrease of the COD removal efficiency from 63.8% to 58.4% and negligible metal cations leaching.

13.4.2  Heterogeneous photo-Fenton Process Photo-Fenton process is similar to the Fenton treatment but co-catalyzed by light [4, 16, 130, 131, 140]. The great advantage of the photo-Fenton treatment is that the irradiation may promote the recovery of Fe2+ (Eq. 13.7), and the newly generated Fe2+ ions react again with H2O2 yielding HO• radicals, thus accelerating the decomposition of H2O2.

271

272

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment



Fe3+ + hυ →  Fe2+ + HO⋅

(13.7)

Several studies have reported the use of heterogeneous photo-Fenton catalysts based on active iron species and other Fenton active metals or cations (e.g., Cu and Co, Fenton-like processes) supported on carbonaceous materials [4, 131, 151]. Photo-assisted Fenton process holds great promise in overcoming the identified challenges of conventional Fenton reaction. In addition, the treatment process can be scaled up for practical application and achieve highly efficient removal of various organic pollutants. For instance, the photo-Fenton degradation of the norfloxacin antibiotic, using Fe3O4/MWCNT as catalyst in aqueous solution, was studied by Shi et al. [152]. Fe3O4/MWCNT was prepared by an in situ hydrothermal method, using carboxylated MWCNT as support for Fe3O4 NPs (200 nm, spherical), and the adsorption capacity and catalytic degradation efficiency of norfloxacin in aqueous medium under simulated solar light was studied (|norfloxacin|0 = 0.5 mg mL−1, Vnorfloxacin −1 solution = 50 mL, |catalyst| = 0.6–1.6 g L , |H2O2| = 0.039–0.137 M, 300 W xenon lamp). The Fe3O4/ MWCNT composite exhibited high norfloxacin adsorption ability (10.31 μg g−1 and 25.68 μg g−1 for Fe3O4 and composite, respectively, when using an initial concentration of catalyst of 1.2 g L−1). Moreover, it was catalytically active in the dark and under light irradiation: under optimal conditions (|catalyst| = 1.2 g L−1, |H2O2| = 0.098 M), the Fenton process (dark) led to 60.99% norfloxacin degradation in 180 min, whereas the photo-Fenton process under light irradiation degraded 91.36% of norfloxacin after 180 min. The enhanced photo-Fenton activity was due to a synergistic effect between Fe3O4/MWCNT and the Xenon light on the decomposition of H2O2 to produce reactive species for norfloxacin degradation. A Fe3O4/MWCNT nanocomposite (specific surface area: 96.7 m2 g−1) prepared by coprecipitation method was also tested as Fenton-like and photo-Fenton catalyst under visible light irradiation for MB dye degradation in the presence of H2O2 (|MB|0 = 30 mg L−1, VMB aqueous solution = 150 mL, |catalyst| = 1.0 g L−1, |H2O2| = 1.5 g L−1, 400 W halide lamp) [153]. The catalyst was active in photo-Fenton (almost complete MB degradation in 20 min) and in Fenton (91% degradation of dye in 20 min). Magnetically-recyclable composites of MWCNT incorporating NiFe2O4 ferrite (NiFe-CNT), with different CNT contents (10–50 wt%), were synthesized by a one-step hydrothermal method and applied as catalysts in the photo-Fenton degradation of the antibiotic sulfamethoxazole (SMX) [154]. The NiFe-CNT catalysts (0.025 g L−1) were tested in the photocatalytic degradation of SMX (|SMX|0 = 5 ppm, VSMX solution = 40 mL) in the presence and absence of H2O2 (1 μL mL−1) under UV-A (medium-pressure mercury lamp, 100 W) and simulated solar light irradiation (Xe Arc lamp, 150 W). The best photo-Fenton activity was found for the NiFe-CNT composite with 25 wt% MWCNT, leading to the total SMX removal and 68% TOC removal after 120 min irradiation with UV-A light vs. 80% removal under UV-A light in the absence of H2O2. The pristine NiFe2O4 NPs were also active and degraded 81% of SMX (39% TOC removal) after 120 min, with or without UV-A irradiation. Other MWCNT-MFe2O4 composites, where M is Zn2+, Co2+, and Mn2+, were also tested under similar experimental conditions, showing slightly lower photo-Fenton activity than NiFe-CNT (total SMX removal, k = 0.041 min−1) and leading to the degradation of 99% (k = 0.036 min−1), 92% (k  =  0.023 min−1) and 89% (k  =  0.019 min−1) of SMX, respectively. The enhanced photo-Fenton properties of the composites relative to the pristine ferrite NPs were attributed to a synergetic effect between the NPs and MWCNT, as the MWCNT have high electronic conductivity and could work as photosensitizer. By employing a Stöber-like method, ultra-dispersed Fe3O4 NPs (3–8 nm) supported on rGO were used as photo-Fenton catalyst for the degradation of common dyes: methyl orange (MO), MB, and

13.4  Advanced (Photo)catalytic Oxidation Processes for Wastewater Treatment

RhB [155]. The hybrid material (specific surface area of ∼199.8 m2 g−1) showed high and stable catalytic performance in the dyes degradation in acidic medium in the presence of H2O2 (pH 3.5, |dye|0 = 10 mg L−1, Vdye solution = 50 mL, 1.2 mL of 30 wt% H2O2) under simulated solar light irradiation (300 W Xe lamp). Regarding the photo-Fenton performance for MO removal, 98% degradation was observed after 30 min, and the catalyst maintained its activity during four cycles (90% degradation after the forth cycle). In contrast, the photo-Fenton activity of the pristine Fe3O4 dropped after the first cycle, and after the forth cycle was able to degrade only 20% of MO vs. ca. 85% in the first cycle. The hybrid material also exhibited significant capacity for MO absorption, of 55%, in the dark for 30 min. The Fe3O4/rGO composite also showed excellent photo-Fenton activity in the degradation of MB and RhB dyes, with nearly total degradation in less than 10 min and 32 min, respectively. The excellent photo-Fenton ability and high stability of the hybrid catalyst, namely when compared to Fe3O4, were explained by the photo-generated electrons from the dye and Fe3O4, which could be transferred to rGO. Afterwards, Fe3+ could capture the electrons to form Fe2+, and then the Fe2+ cations could continue reacting with H2O2 to form Fe3+ and HO• on the surface of rGO, which then degraded the dye molecules. The generated Fe3+ was reduced to Fe2+ again by the electrons concentrated on the surface of graphene, to maintain the Fe3+/Fe2+ cycle. Enhanced photo-Fenton activity was observed on the removal of MB dye catalyzed by hematite NPs (α-Fe2O3, 45 ± 10 nm) anchored on GO nanosheets, prepared by an in situ method, in the presence of H2O2 under UV light irradiation (|MB|0 = 40 mg L−1, VMB solution = 400 mL, 100 mg of catalyst, |H2O2| = 1.10 mm, 100 W high-pressure mercury lamp, λ = 365 nm) [156]. The estimated MB decolorization rate in the presence of α-Fe2O3@GO nanocomposite (49.6% Fe loading) was k = 0.1953 min−1 (decolorization ratio of 99% within 80 min for a wide pH range of 3–12), which was approximately 2.4 × higher than that obtained for α-Fe2O3. The improved performance in the presence of α-Fe2O3@GO was attributed not only to the low agglomeration of α-Fe2O3 NPs on the composite surface, but also to the high electron conductivity and electrostatic interaction between the negatively-charged GO and positively-charged MB. The α-Fe2O3@GO was also tested as adsorbent (without H2O2, in the dark) and catalyst in the photo-Fenton degradation of other organic pollutants: cationic (RhB), anionic (Orange II and Orange G), neutral (phenol, and 2-nitrophenol), and endocrine disrupting compounds (17-Estradiol). α-Fe2O3@GO showed high MB adsorption (34% in 80 min) and degradation rate (k = 0.1953 min−1, total decolorization in 40 min), which was similar to that achieved for RhB, probably due to the existence of a strong electrostatic interaction between the negatively-charged GO surface and the RhB and MB cationic dyes. In contrast, moderate adsorption values (15% in 80 min) and degradation rates (k = 0.0638 min−1 and k = 0.060 min−1) were estimated for Orange G and Orange II, probably due to a strong electrostatic repulsion between the negatively-charged GO and these anionic dyes. The adsorption values of phenol, 2-nitrophenol and 17-estradiol on the surface of α-Fe2O3@GO were relatively low (below 13% after 20 min), but their degradation rates were high: k  =  0.1231 min−1, 0.1042 min−1 and 0.0608 min−1 for phenol, 17-estradiol and 2-nitrophenol, respectively. Thus, the combination of GO with α-Fe2O3 NPs greatly improved the catalytic efficiency of α-Fe2O3 toward cationic dyes and phenol-like compounds. The degradation of synzol red reactive dye (|dye|0 = 5–60 mg L−1, Vdye solution = 100 mL, pH 2–9) using a GO-ZnFe2O4 composite (average grain size: 22.9 nm), prepared by hydrothermal method, as heterogeneous photo-Fenton catalyst (20–200 mg L−1) was studied in the presence of H2O2 (9.8– 87 mm) [157]. The GO-ZnFe2O4 composite led to significantly higher dye degradation (94% at pH 3, |GO-ZnFe2O4| = 50 mg L−1, |H2O2| = 27 mm) than the pristine ZnFe2O4 (57%, |ZnFe2O4| = 75 mg L−1), considering 60 min of irradiation time by six UV lamps (each of 18 W). The enhanced photocatalytic efficiency of the composite was attributed to multiple factors, such as the more efficient

273

274

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

generation of HO• radicals and charge transfer ability, the lower recombination of excited electrons and holes, and the smaller band gap (2.2 eV vs. 2.9 eV for composite and ZnFe2O4 NPs, respectively). Both catalysts showed a progressive decrease in the degradation efficiency from 94% to 68% from the first to the sixth reaction cycle. Graphite-based composites obtained by thermal treatment of biomass tar and iron salt (Figure 13.4) and presenting different Fe loadings (theoretical loadings: 10, 20 and 40 wt%) were tested as photo-Fenton catalysts in the degradation of RhB dye under visible light irradiation (|RhB|0 = 40 mg L−1, VRhB solution = 50 mL, 50 mg of catalyst, |H2O2| = 20 × 10−3 M, 300 W high-pressure Xenon lamp with a 420 nm UV cut filter) [158]. The catalysts, denoted as Fe10@C, Fe20@C, and Fe40@C (corresponding to 10, 20, and 40 wt% loading, respectively), were prepared via a self-reduction and solvent-free strategy, being decorated with various iron-based species (Fe3O4, Fe3C, and Fe0); moreover, they were active in a wide pH range (2–8). The catalyst with 20 wt% iron content (Fe20@C, containing Fe0 and Fe3O4 NPs with approximately 5 nm and 13 nm, respectively, by TEM), which presented the highest specific surface area (228.96 m2 g−1), graphitization degree, and amount of Fe0, showed excellent activity at pH 6, leading to total RhB degradation in 2 h vs. 81.8% and 99.8% for the catalysts containing 10 and 40 wt% of iron, respectively. The Fe20@C composite was slightly active as photocatalyst (reaction performed in the absence of H2O2 under visible light irradiation), degrading less than 20% of RhB in 120 min, which confirmed the essential role of H2O2. Moreover, it was able to degrade only 89% of RhB after 120 min under Fenton conditions (in the presence of H2O2 without light radiation), demonstrating that the catalyst could initiate the photocatalytic process. Additionally, Fe20@C showed excellent stability and reusability, with only 7.3% activity loss after five reaction cycles. The excellent catalytic performance was attributed to the presence of a defective graphitic structure and Fe0, which promoted the transformation of Fe3+ to Fe2+, and to the strong adsorption capacity of the carbonaceous material (Figure 13.4). Overall, owing to the high catalytic activity, excellent reusability and stability, as well as a green and facile synthesis process, the prepared catalyst was appointed as a valuable alternative for organic pollutants removal from aqueous solutions. Amorphous Fe-Zn-oxide/hydrochar (Fe-Zn/HC) prepared by a hydrothermal method (average particle size: 310 nm, specific surface area: 76.8 m2 g−1) was used as a heterogeneous photo-Fenton catalyst for the degradation of various organic pollutants, namely dyes (RhB and MB), an anti-inflammatory drug (antipyrine, nonsteroidal drug), and phenol (highly toxic compound) [159]. The Fe-Zn/HC catalyst was able to almost completely degrade phenol after 50 min under

Figure 13.4  Schematic representation of the synthesis of graphite-supported iron-based photo-Fenton nanocatalysts and photo-Fenton mechanism of Rhodamine B (RhB) degradation under visible light irradiation. Reproduced with permission from Ref [158].

13.4  Advanced (Photo)catalytic Oxidation Processes for Wastewater Treatment

visible light (pH 6.5, |phenol|0 = 10 mg L−1, 50 mg of catalyst, |H2O2| = 1 × 10−2 M, 300 W Xe lamp with a UV cut-off filter at λ ≥ 420 nm). The removal efficiencies of the other pollutants were 96.2%, 95.2% and 84.1% for MB, RhB and antipyrine, respectively, after 80 min. It was shown that the catalyst was significantly less active in the Fenton process (in the presence of H2O2 but in the dark), with pollutant removals of only 34.5%, 7.6%, and 36.6% for MB, antipyrine, and RhB, respectively, for the same reaction time. This provided strong evidence about the synergistic effect between heterogeneous Fenton catalytic and photocatalytic reactions, which may promote the generation of HO• for nonselective oxidation of organic pollutants. Perovskite LaFeO3 NPs were immobilized on the surface of monodisperse carbon spheres through an environmentally friendly ultrasonic-assisted surface ions adsorption method [160]. The obtained LaFeO3/C nanocomposite was used as heterogeneous photo-Fenton-like catalyst for the RhB degradation under visible light irradiation (|RhB|0 = 15 mg L−1, VRhB solution = 50 mL, 50 mg of catalyst, 1 mL of 30 wt% H2O2, 300 W Xe lamp with a 420 nm cut-off filter). The nanocomposite exhibited much higher photo-Fenton like catalytic activity (99.4% degradation after 120 min) than LaFeO3 (85.5% degradation after 120 min). The immobilization of LaFeO3 (spherical morphology, 5.4 nm by XRD) on carbon spheres (diameter of ~150 nm) led to better dispersion of the NPs on the carbon surface, which prevented their aggregation. The presence of carbon spheres in the nanocomposite was also responsible for the enhancement of the adsorption ability of the LaFeO3/C composite (BET surface area: 254.6 m2 g−1) relative to the bare LaFeO3 NPs (BET surface area: 24.7 m2 g−1) and suppressed the recombination rate of photo-generated electron-hole pairs. Three-component composites of iron(II,III) oxide / zinc oxide / graphene (Fe3O4/ZnO/graphene, graphene contents of 1 and 3 wt%) were obtained by sol–gel followed by hydrothermal method, and used as catalysts in the Fenton and photo-Fenton (40 W UV-C lamp or 40 W Xe lamp) removal of both cationic (MB) and anionic (Congo red [CR]) dyes (|dye|0 = 40 mg L−1) from aqueous solutions in a wide pH range (3–13) in the presence of H2O2 (4 mL) [161]. The graphene incorporation significantly enhanced the Fenton-like dyes degradation (MB dye: from 79% for Fe3O4/ZnO to 85% for Fe3O4/ZnO/graphene in 120 min, pH 13; CR dye: from 80% for Fe3O4/ZnO to 90% for Fe3O4/ ZnO/graphene in 120 min, pH 3), which was further improved by the introduction of light irradiation. Under UV light irradiation, the total degradation of MB and CR was reached within 120 min, and in the presence of visible light, the degradation of MB and CR reached 93% and 97%, respectively, within 120 min. The addition of graphene was very beneficial to increase the specific surface area of the catalysts (from 10 m2 g−1 for Fe3O4/ZnO to 16 m2 g−1 and 21 m2 g−1 for Fe3O4/ZnO/ graphene composites with 1 wt% and 3 wt% graphene loading, respectively) and thus to enhance the dyes mass transfer toward the active sites (Fe2+/Fe3+) during the reaction. Under UV and visible light irradiation, the regeneration of Fe2+ ions would be facilitated during the reaction owing to the photo-reduction of Fe3+ to Fe2+, which would speed up the redox cycle between the active sites (Fe2+/Fe3+). The H2O2 molecules could also be decomposed into HO• by the irradiation with light (reaction with photogenerated h+ by ZnO semiconductor). Another three-component composite containing sulfur-doped GF (S-GF), CuS, and Fe3O4 was prepared by the in situ growth of metal-containing species on the S-GF scaffold [162]. The average crystallite size of the Fe3O4 NPs decreased from 9.9 nm (Fe3O4) to 5.6 nm (S-GF@Fe3O4) and 5.1 nm (S-GF@CuS-Fe3O4), whereas that of CuS decreased from 7.4 nm for the bare NPs to 6.7 nm for S-GF@ CuS. The composite was used as photo-Fenton-like catalyst for the 4-nitrophenol (4-NP) degradation under UV-A radiation (|4-NP|0 = 0.05 mm, V4-NP solution = 100 mL, 20 mg of catalyst, 2 mL of H2O2, 15 W mercury lamp, Figure 13.5). Similarly to Fe3O4/ZnO/graphene in the previous example, the S-GF@ CuS/Fe3O4 composite showed better catalytic activity than the individual components, leading to 95.2% 4-NP degradation vs. 40%, 54%, 63%, and 77% for S-G, Fe3O4, S-GF@Fe3O4, and S-GF@CuS,

275

276

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

Figure 13.5  Illustration of the structure of S-GF@CuS-Fe3O4 photo-Fenton catalyst and catalytic performance on the degradation of 4-NP under UV-A radiation in the presence of H2O2. Reproduced with permission from Ref [162].

respectively. The photo-Fenton degradation of 4-NP followed a pseudo-first order kinetics, with k = 0.016 min−1 for S-GF@CuS-Fe3O4, which was 25.7 × higher than that of S-GF and 3.2 × higher than that of S-GF@Fe3O4. All these achievements showed a positive synergetic effect between the nanocomposite components. The presence of CuS NPs, similarly to ZnO in the Fe3O4/ZnO/graphene nanocomposite, showed a main role in the 4-NP degradation, leading to a higher improvement of the catalyst performance when compared to the Fe3O4 NPs (as revealed by the comparison between the S-GF@CuS and S-GF@Fe3O4 photocatalytic activities: 77% vs. 63%, respectively). Magnetic ternary structures (MIL-101(Fe) MOF/CoFe2O4/GO, GO contents of 3 wt% or 7 wt%) were used as photo and photo-Fenton-like catalysts for effective visible light-driven (100 W LED lamp) degradation of the azo dyes Direct Red 23 (DtR-23) and Reactive Red 198 (ReR-198), and tetracycline hydrochloride (TC-H) antibiotic, Figure 13.6 [163]. During the photo-Fenton tests, the organic compounds (pH 3 or 8–8.5 for the dye and TC-H degradation, respectively; |dye| = 60–100 mg L−1 or |TC-H| = 30 mg L−1; Vorganic compound solution = 100 mL) was degraded in the presence of H2O2 (50 μL) and 2 mg of catalyst. The ternary magnetic composite with 3 wt% of GO was tested as photocatalyst in the degradation of DtR-23 dye and demonstrated the best performance (without addition of H2O2) for DtR-23 dye concentration of 60 mg L−1, leading to 93.20% of degradation in 70 min vs. ~80% for the pure MOF, ~85% for MIL/GO, and ~60% for CoFe2O4/GO. The photocatalytic activity of the composite was significantly reduced upon the increase of the dye concentration to 100 mg L−1, leading to 56.26% DtR-23 degradation after the same reaction time. Under photoFenton conditions, the same catalyst presented the best degradation performance after 70 min of irradiation, leading to DtR-23 and ReR-198 degradation percentages of 99.93% and 99.65%, respectively. Moreover, the MIL_101(Fe)/CoFe2O4/GO catalyst (3 wt% GO) demonstrated good durability in terms of stability and reusability, being used five times with a reduction of the photocatalytic activity of only 8%. The magnetic ternary composite also showed very good performance in the photo-Fenton like degradation of the colorless TC-H organic contaminant (92% in 50 minutes), whereas only 22% of TC-H was degraded in the presence of only H2O2. The excellent photo-Fenton properties of the composite were related to the rapid formation of the electron-hole pairs inside the structure of MIL-101(Fe) and CoFe2O4, and to the strong electron acceptor properties of GO

13.4  Advanced (Photo)catalytic Oxidation Processes for Wastewater Treatment

Figure 13.6  Schematic illustration of the mechanism of photo-Fenton-like and photocatalytic degradation of organic pollutants using MIL-101(Fe)/CoFe2O4/GO catalyst (3 wt% GO) as catalyst. Reproduced with permission from Ref [163] / Elsevier.

13.4.3  Heterogeneous electro-Fenton Process The electro-Fenton (E-Fenton) process is an electrochemical AOP, which is popularly known for the in situ generation of H2O2 via the two-electron ORR at the cathode (i.e., through the O2 reduction by the transference of two electrons, with formation of H2O2 as intermediate, as explained in more detail in Section 13.2.1). Hence, this process reduces the risk associated with H2O2 transport and storage. There are three major steps involved in the E-Fenton process: (i) in situ generation of H2O2, (ii) effective generation of HO• via the decomposition of H2O2 catalyzed by Fe2+ cations (Fenton’s reaction), and (iii) regeneration of Fe2+ at the cathode surface, as represented in Eqs. (13.8–13.10) [6, 164]:

O2 + 2H+ + 2e− → H2O2

Fe2+ + H2O2 → Fe3+ +  HO− + HO⋅ Fe

3+

−

+ e → Fe

2+

(13.8) (13.9) (13.10)

Similarly to the conventional classical Fenton process described in Section 13.4.1, the generated HO• radicals react with organic pollutants, leading to the formation of intermediate compounds or to their complete mineralization [6, 127, 164]. Depending on the reactive phase of the catalyst, the process can be classified as homogeneous or heterogeneous E-Fenton. The use of homogeneous E-Fenton catalysts (e.g., Fe(II) chloride, Fe(II) sulfate) typically suffers from serious drawbacks related to the non-reusability of the catalyst (formation of iron sludge) and acidic medium requirement [6]. Heterogeneous E-Fenton catalysts allow circumventing these issues, but the efficiency of the E-Fenton reaction always relies on the

277

278

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

nature of the cathode material (typically, a Pt sheet or boron-doped diamond (BDD) thin film were used as the anode). An effective cathode material should possess high overpotential for HER, low activity for H2O2 decomposition, high stability, and electrical conductivity [164]. The use of carbon-based materials combined with transition metal-based species is a major tool considering their excellent characteristics of adsorption, electrical conductivity, and catalysis. Li et al. reported the fabrication of core-shell Fe@Fe2O3 (spherical morphology, 10–100 nm) loaded on ACF by in situ coprecipitation method and its application as oxygen diffusion cathode in the heterogeneous E-Fenton degradation of RhB [165]. Considering an initial concentration of RhB of 5 mg L−1 in 0.05 mol L−1 Na2SO4 aqueous solution (pH 6.2), 74% of RhB was degraded after 120 min of reaction with Fe@Fe2O3/ACF hybrid, which was better than the activity observed with zerovalent iron and Fe2+ cations (47.6% and 25.5%, respectively). The H2O2 concentration increased with the reaction time and reached a steady state at ~80 μmol L−1 after 90 min of reaction at neutral pH (vs. ~8 μmol L−1 at pH 2–3). This showed that the H2O2 electro-generation was not favored at low pH values, although typically the Fenton reactions take place more easily at pH 2, since this is the optimum value for H2O2 decomposition catalyzed by Fe2+ ions. Wang et al. prepared a α-Fe2O3/CA hybrid by an innovative one-pot synthetic route, through the impregnation of iron precursor on CA followed by heat treatment and using different iron contents (in the range of 0.05–6.0 wt%, added as Fe2SO4) [166]. The optimized E-Fenton catalyst (hybrid prepared with an iron content of 5 wt%, spherical NPs composed of α-Fe2O3 and Fe with 80–120 nm embedded in the carbon matrix just like plums in pudding, BET surface area: 421 m2 g−1) was tested in the degradation of metalaxyl substrate (at 25 ºC, pH from 3 to 9, |metalaxyl|0 = 500 mg L−1 in 0.1 mol L−1 Na2SO4 (100 mL), BDD as anode), which is an active ingredient of fungicides. The catalyst worked over a wide potential range and resulted in a TOC removal of 98% in 240 min, which was 1.5 × higher than that obtained using a commonly supported Fe@Fe2O3@CA electrode. Such performance was attributed to the mixture of Fe and α-Fe2O3 phases, which ensured Fe2+ cations as the mediator and maintained high catalytic activity via reversible redox reactions involving the electron transfer among iron species in different oxidation states. In another work, an electrocatalyst of α-Fe2O3 NPs wrapped in graphene aerogel (Fe loading: 49.73 wt%) was prepared by a hydrothermal method followed by natural-drying [167]. The material exhibited efficient electrocatalytic performance for the RhB degradation (at room temperature, |RhB|0 = 10 mg L−1 in aqueous 1 mol L−1 Na2SO4, Pt sheet as anode), reaching 99% after 30 min, with low iron leaching, almost no decrease of catalytic activity after six cycles, and good efficiency over a wide pH range (from 3 to 8). A highly ordered mesoporous Fe3O4@carbon composite was grafted onto monolithic CA by a soft-template method [168]. The obtained Fe3O4@OMC/CA composite presented stripe-like morphology and hexagonally-arranged pores (OMC/CA support) with the Fe3O4 NPs (cubic spinel structure, average size of 12 nm) highly dispersed in the carbon matrix. The composite was tested as E-Fenton cathode for the degradation of the endocrine disruptor compound dimethyl phthalate (pH 3.0–7.0, |dimethyl phthalate|0 = 50 mg L−1 in 50 mm Na2SO4, BDD thin-film as anode), leading to 95% removal and 65% demineralization (measured by TOC) after 120 min. The total dissolved iron after 120 min of reaction was 0.66–0.93 mg L−1, indicating the excellent chemical stability of the catalyst. Regarding its reusability, 78–86% of dimethyl phthalate removal efficiency was attained after five cycles, but the TOC removal efficiency significantly decreased to 10%. The highly dispersed Fe3O4 NPs increased the number of catalytically active sites, whereas the OMC/CA support provided accessible pathway to the active centers, reduced the resistance to mass transport, and limited the E-Fenton reaction to a confined space, effectively maintaining the Fe3O4 activity and reducing the iron leaching.

13.4  Advanced (Photo)catalytic Oxidation Processes for Wastewater Treatment

Fe3O4 NPs (~80 nm) were also incorporated in graphene/CNT-containing CA (GMCA) by a sol-gel method, followed by carbonization (iron oxide content: 5 wt%) [169]. The characterization results showed that the coexistence of graphene and CNT was mandatory to enhance the hybrid micro/ mesoporosity (0.736 cm3 g−1 vs. 0.167 and 0.416 cm3 g−1 for CNT- and graphene-containing CA, respectively), specific surface area (479.8 cm2 g−1 vs. 216.1 and 521.6 cm2 g−1 for CNT- and graphene-­ containing CA, respectively), and robustness, which are good indicators for a valuable catalytic activity. The hybrid was tested as E-Fenton catalyst for the methyl blue degradation (room temperature, |methyl blue|0 = 10 mg L−1 in 0.05 mol L−1 Na2SO4, Pt sheet as anode), leading to 99% methyl blue removal efficiency after 60 min at pH 3. A slight decrease in the dye removal efficiency was observed when the pH was lower or higher than 3, since the E-Fenton system was governed by the HO• radicals, which were generated by the reaction between Fe2+ and H2O2 and, thus, depended on the H2O2 generation; this process was favored at pH 3, since at lower or higher pH values other side reactions occurred, which reduced the H2O2 yield. The catalyst presented a reaction rate constant of 0.072 min−1 and high catalytic activity (up to 91%) after prolonged recycling/reuse (10 total cycles). More recently, Fe3O4 NPs were entrapped in a MWCNT network (carbon microtubes) by a onepot synthesis method [170], and in N-doped rGO by a one-pot solvothermal method [171]; the obtained composites were effective in the removal of carbamazepine (pH 7, |carbamazepine|0 = 4.7 mg L−1 in 0.05 mol L−1 Na2SO4, ~100% removal after 40 min) and bisphenol A (pH 3, |bisphenol A |0 = 20 mg L−1 in 0.05 mol L−1 Na2SO4, 93.0% within 90 min), respectively. Once again, the presence of the carbon support facilitated the mass transfer of pollutants to the electrode; moreover, it accelerated the regeneration of Fe2+ due to faster electron transfer, thereby enhancing the efficiency of the E-Fenton process. Recently, NC nanofiber (N-CNF) electrodes incorporating Co/CoOx NPs (prepared by electrospinning cobalt(II) acetate and polyacrylonitrile (PAN) solution, cobalt(II) acetate:PAN weight ratios between 0–20%, followed by thermal carbonization of the resulting cobalt acetate/PAN nanofibers under N2 atmosphere) were developed as new cathode materials to remove the Acid Orange 7 dye [172]. The Co/CoOx NPs (32.1–37.0 nm, content between 8.9–48.2%) were homogeneously distributed on N-CNF (diameter between 300–400 nm). The highly crystalline graphitic layers of N-CNF coated the Co/CoOx NPs and protected them from dissolution. The hybrids (BET surface area: 7.1–240.2 m2 g−1; mean pore diameter: 5.6–13.9 nm) revealed excellent electrocatalytic ability for the degradation of Acid Orange 7 (pH 3–6, | Acid Orange 7|0 = 0.1 mmol L−1 in 0.05 mol L−1 Na2SO4, Pt mesh as anode), which was clearly influenced by the Co/CoOx NPs content. At pH 3, the residual concentration of Acid Orange 7 after 40 min of reaction varied between 8.8% and 11.1%. The lowest residual concentration and highest TOC of 92.4% were achieved for the hybrid prepared with a cobalt acetate:PAN weight ratio of 10% (optimal mass content of Co NPs of 25%). The apparent kinetic constant varied between 0.071 min−1 and 0.089 min−1. On the other hand, the residual Acid Orange 7 concentration using the same hybrid but at pH 6 was 12.4% after 40 min (TOC removal: 93.3%), and the prepared electrode showed to be chemically stable over ten electrolysis cycles. Very recently, CeO2 encapsulated on N,P-doped carbon was synthesized via pyrolysis process at several temperatures (700–1100 ºC, Figure 13.7) [173]. The TEM and XRD characterization of the composite pyrolyzed at 1000 ºC indicated the coexistence of CeO2 and CePO4 phases with irregular morphology. The material showed to be more active for the CIP degradation by E-Fenton process (pH 2.0–7.0, |CIP|0 = 30 mg L−1 in 0.05 mol L−1 Na2SO4, carbon felt (CF) as cathode substrate, and Pt sheet as anode), leading to 95.4% degradation after 180 min at pH 3.0 (k = 0.0200 min−1), a TOC of 57.3%, and a mineralization current efficiency of 16.4% after 1 h. The CIP degradation was only 62.7% after 180 min using the pristine CeO2 and 78.6% using the analogous hybrid based on simple

279

280

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

Figure 13.7  Schematic illustration of the synthesis of CeO2 encapsulated on N,P-doped carbon and of the mechanism of heterogeneous E-Fenton degradation of ciprofloxacin (CIP). Reproduced with permission from Ref [173].

NC, revealing the advantage of the P-doping. After three consecutive runs, the CIP degradation efficiency promoted by the composite catalyst almost did not decline, being 94.8% with a leaching concentration of Ce ions of 5.89 mg L−1. The composite showed increased ability for the electron transport and higher concentration of oxygen vacancies (in comparison with the undoped analogues), which were beneficial for the CIP degradation, Figure 13.7. Additionally, the degradation activity of the catalyst was evaluated for other antibiotics, namely TC, enrofloxacin, and SMX, with degradation efficiencies in the range of 90.6–98.8% after 180 min, showing the wide action spectrum of this catalyst. Multi-metallic oxides have also been widely explored as E-Fenton cathodes. CoFe2O4 self-supported on CFP (CFP@CoFe2O4) [174] or on CF (CoFe2O4/CF) [175] were prepared by solvothermal method (followed by thermal treatment in the first case) and tested as E-Fenton catalysts for the 4-NP and tartrazine degradation, respectively. The suitable conditions for the E-Fenton process (e.g., pH, initial pollutant concentration and electrolyte) were explored: globally, pH 3, initial pollutant concentration of 50 mg L−1 and Na2SO4 as electrolyte were found to be the optimal conditions. In both cases, the catalysts induced high decolorization efficiency (~100% after 120 min for 4-NP using CFP@CoFe2O4 and 97.05% after 40 min for tartrazine using CoFe2O4/CF) and high stability (4-NP degradation efficiency of ~95% after five cycles for CFP@CoFe2O4), showing that the E-Fenton processes benefited from the redox pairs Co3+/Co2+ and Fe3+/Fe2+ on the CoFe2O4containing cathode surfaces. The work performed with the CoFe2O4/CF electrocatalyst demonstrated yet that the use of NaNO3, NaHCO3, and Na2HPO4 electrolytes significantly inhibited the E-Fenton process.

13.4  Advanced (Photo)catalytic Oxidation Processes for Wastewater Treatment

Cu/CuFe2O4 modified graphite felt was tested as E-Fenton cathode for TC degradation [176]. Cu/ CuFe2O4 NPs (spherical morphology, cubic spinel phase of CuFe2O4, and metallic copper, Cu0 content in the range of 36.3–61.7%) were synthesized by a one-step solvothermal approach and, subsequently, used to modify the graphite felt through the polytetrafluoroethylene (PTFE)-bonding technique. The PTFE acted as a hydrophobic protective agent, inhibiting metal leaching and ensuring the long-term operational stability of the catalyst. The degradation efficiency of TC (pH 3–7, |TC|0 = 50 mg L−1 in 0.05 mol L−1 Na2SO4, modified graphite felt and Pt electrode as cathode and anode, respectively) was found to increase from 69.3% to 96.3% (after 2 h) with the Cu0 content, as well as the mineralization rate (TOC values from 31.8% to 83.6% after 2 h). Cu0 was pointed out as the main responsible for the high catalytic activity by enhancing the 2-electron ORR selectivity and accelerating the Fe2+ regeneration. The most effective cathode (with 61.7% of Cu0 content) displayed high stability and led to nearly 80% TC degradation after five cycles. Recently, a self-supporting CFP electrode modified with MnO2 and Fe3O4 (4.96 wt% of Fe and 0.025 wt% of Mn) was used as catalyst for the E-Fenton treatment of shale gas fracturing flowback wastewater [177]. The modified electrode was prepared by electrodeposition method followed by thermal treatment, using MIL-101 as precursor. Regarding the E-Fenton activity in the treatment of real fracturing flowback wastewater containing RhB, 4-NP, phenol, bisphenol A, and polyacrylamide (PAM) as the main substrates (100 mL of total volume, BDD electrode as anode), the catalyst led to 65% of TOC and 75% of COD removal after 4 h. The residual COD level met the emission requirement of the integrated wastewater discharge standard (COD  TiO2@CNT (k = 0.027 min−1) > TiO2@C (k = 0.0068 min−1). The results of electrochemical impedance spectroscopy showed a decrease of the separation efficiency of the photogenerated e− – h+ pairs in the same order of the photocatalysts activity. Thus, the significant improvement in the photocatalytic degradation of RhB observed for the hybrid materials was attributed to the efficient separation between the injected electrons and excited RhB (photosensitization pathway). Similarly, the photoactivity of TiO2 over dyes photodegradation and bacteria inactivation was improved in the presence of polyhydroxy fullerene (PHF) deposited onto TiO2 anatase phase (named PHF-TiO2) prepared by impregnation method [187, 188]. Krishna et al. found that the degradation rate of Procion red dye (3 mg L−1) in the presence of the PHF-TiO2 composite (single crystal size of 5 nm, 30 mg L−1) was 2.6×  higher than that obtained with TiO2 anatase (0.0128 min−1 vs. 0.0048 min−1) and 213 × higher than that promoted by PHF alone (6 × 10−5 min−1) under 60 min irradiation with UV-A light (intensity 86 W m−2) [188]. Moreover, the inactivation of Escherichia coli by PHF-TiO2 (k  =  0.177 min−1) was 1.9×  faster than that obtained with TiO2 (commercial P25, Degussa, k = 0.094 min−1) under similar experimental conditions. The enhancement in the photocatalytic activity of PHF-TiO2 composite was assigned to the high electron affinity of PHF, leading to the reduction of the recombination rate of e− – h+ pairs within titania.

283

284

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

The remarkable synergy between adsorption and photocatalysis for ciprofloxacin antibiotic elimination from water was achieved using a graphitized mesoporous carbon (GMC)-TiO2 anatase nanocomposite (GMC-TiO2) as adsorbent and photocatalyst [189]. The composite was prepared by low temperature hydrothermal method, containing TiO2 anatase NPs of 12 nm crystallite size well dispersed on the GMC matrix and a carbon content of around 14.5%. The GMC-TiO2 (70 mg) as well as commercial TiO2 (P25 Degussa, 70 mg) degraded ciprofloxacin (15 mg L−1, 200 mL) in 120 min of UV light irradiation, with almost the same reaction rates of 0.102 min−1 and 0.107 min−1, respectively. However, the photocatalytic activity of GMC-TiO2 benefited from its excellent adsorption capability (5× higher specific surface area than that of commercial TiO2, 286 m2 g−1 vs. 54 m2 g−1, and mesoporous structure) and suppression of recombination of photogenerated electrons and holes induced by the carbon component. The same effects were observed for molecularlyimprinted carbon nanosheets supported TiO2 (pure anatase phase) tested in antibiotics removal (ciprofloxacin and SMX) under UV irradiation, Figure 13.9 [190]. The enhanced photocatalytic degradation of RhB dye and phenol was also reported using CDs with average diameter of around 4–5 nm anchored onto TiO2 nanotube arrays (NTAs) as photocatalyst [191]. The CDs/TiO2 material was obtained by a two-step method of electrochemical anodization to produce the TiO2 nanotube arrays, followed by electrochemical deposition of the CDs. The CDs/TiO2 with 18.22 At% surface carbon content and 27.87 At% Ti content led to 72.5% of RhB dye removal (|RhB|0 = 10 mg L−1, VRhB solution = 10 mL, photocatalyst with 1.5 × 3 cm2 dimensions) vs. 50.7% for TiO2 for 180 min irradiation with simulated solar light. The enhanced photocatalytic results were assigned to the synergy between the CDs and TiO2, leading to narrower bandgap and better separation of e− – h+ pairs in TiO2. A similar tendency was observed in the degradation of the aromatic organic pollutant (similar experimental conditions to those used in RhB degradation): 73.3% of phenol was eliminated by CDs/TiO2 composite after 180 min, whereas only 35.5% was removed when using TiO2 in the same irradiation time. Carbon materials derived from earth-abundant, renewable, and biodegradable residues have been also combined with TiO2 to develop efficient photocatalysts for water treatment [192]. Dalto et al. highlighted the synergy between TiO2 and AC derived from spent coffee grains (10 mg) on the adsorption and photodegradation of MB dye (|MB|0 = 25 mg L−1, VMB solution = 20 mL) [193]. The

Figure 13.9  Schematic illustration of the mechanism of adsorption and photocatalytic degradation of ciprofloxacin (CIP) by molecularly-imprinted carbon nanosheets supported TiO2. Reproduced with permission from Ref [190].

13.4  Advanced (Photo)catalytic Oxidation Processes for Wastewater Treatment

composites were produced by in situ immobilization of TiO2 NPs (anatase and brookite phases) over AC, using different TiO2:AC weight ratios (50:50 and 10:90). The highest MB adsorption was observed for 50:50 TiO2:AC composite, leading to 80% dye removal in 90 min, whereas the 10:50 TiO2:AC composite promoted 78% dye removal after 90 min. In contrast, the parent AC led to 68% removal after 90 min contact time and TiO2 practically did not adsorb the dye ( 420 nm, 0.8 W m−2 nm−1), leading to 98% MB degradation and to an apparent rate constant of 1.0 × 10−1 min−1, which was 1.7 ×, 2.3 × and 10 × higher than those obtained for 10:90 composite (5.7 × 10−2 min−1), pristine AC (4.4 × 10−1 min−1) and TiO2 (1.0 × 10−2 min−1), respectively. The presence of AC in the composites increased the photoactivity of TiO2 due to the high surface area of AC (adsorption), which effectively concentrated the MB molecules around the deposited TiO2. The higher combined adsorption and photocatalytic activity of the 50:50 composite was due to the synergetic effect between AC and TiO2. When the assays were performed using a secondary effluent from an urban wastewater treatment plant as water matrix, all the materials maintained the photoactivity and led to MB discoloration percentages similar to those obtained in the assays with ultrapure water. Antonopoulou et al. reported the improved photocatalytic activity of a composite of pyrolytic char (obtained from the pyrolysis of used rubber tires) and N-F-doped TiO2 (anatase) obtained by sol-gel method (char:TiO2 weight ratio of 0.2:2), in the degradation of phenol (|phenol|0 = 5 mg L−1, |catalyst| = 100 mg L−1, Vphenol solution = 50 mL) under simulated solar light irradiation (average irradiation intensity of 350 W m−2) [194]. The composite (TiO2 crystallite size = 9.5 nm) led to 70% of photocatalytic phenol removal after 240 min, whereas TiO2 only promoted 5% removal. The enhanced photocatalytic activity of the composite was attributed to increased phenol adsorption (char matrix) and improved HO• production due to the effective separation of photogenerated charges in TiO2 since the char could act as a sink for the photogenerated electrons, thus preventing e— – h+ recombination (Figure 13.10). A similar enhancement in the photocatalytic activity was reported by Lisowski et al. in phenol degradation (50 mg L−1) under UV light (medium-pressure 125 W mercury lamp, λmax = 365 nm) at 30  °C using biomass-derived carbon/TiO2(Degussa P25) hybrids (25 wt% TiO2/secondary char + lignin and 25 wt% TiO2/secondary char + SWP700 [softwood pellets BCH]) prepared by controlled pyrolysis [195]. The hybrid materials led to higher phenol degradation than the pristine TiO2 (Degussa P25): 52.5% and 35.8% for 25 wt% TiO2/secondary char + SWP700 and 25 wt% TiO2/secondary char + lignin, respectively, vs. 32.4% for the pristine TiO2 after 240 min of UV light irradiation. The photocatalytic activity of C-doped hollow anatase TiO2 spheres (50 mg) was evaluated in the degradation of RhB dye in an aqueous solution (|RhB|0 = 4.7 mg L−1, VRhB solution = 50 mL) under visible light irradiation (Xe arc lamp, 350 W with UV cut-off filter λ = 400 nm) [196]. After 30 min of reaction, the C-doped TiO2 degraded 95% of the RhB dye, whereas TiO2 (Degussa P25) only led to 60% removal. Clearly, the doped carbon replacing oxygen in the TiO2 lattice played an optical absorption role in extending the response of hollow TiO2 spheres into the visible light range of the solar spectrum. Similarly, pine cone derived C-doped TiO2 (both anatase and rutile phases with crystallite sizes of 26.67 nm and 29.51 nm, respectively) was used in the photodegradation of TC-H (|TC-H|0 = 5 mg L−1, Vpollutant solution = 200 mL; |catalyst| = 300 mg L−1) under visible-LED light (25 W) [197]. The composite exhibited better photocatalytic degradation efficiency when compared to the undoped TiO2 and to bare carbon under visible-LED light. The obtained TC-H removal

285

286

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

Figure 13.10  Schematic illustration of the structure and performance of pyrolytic char/N-F-doped TiO2 (anatase) composite in the photocatalytic degradation of phenol under simulated solar light (SSL). Reproduced with permission from Ref [194].

percentages after 120 min of irradiation were 70%, 54%, and 14% for C-doped TiO2, TiO2 and bare carbon, respectively. The enhanced photocatalytic activity of the C-doped TiO2 was assigned to the carbon doping of TiO2, which led to the formation of O–Ti–C bonds, enhancing the visible light absorbance of the doped TiO2 when compared to TiO2. Recently, bismuth oxybromide (BiOBr) has gained attention as alternative photocatalyst to TiO2 due to its narrower band gap (2.61–2.90 eV vs. 3.0–3.3 eV for TiO2), leading to superior visible light absorption and, consequently, superior catalytic activity under sunlight; moreover, it is cheap, non-toxic, and chemically stable [198]. BiOBr has been coupled with carbonaceous materials to design effective photocatalysts for wastewater treatment. For instance, a nanocomposite of graphene/bismuth oxybromide obtained by coprecipitation method was used as photocatalyst for RhB dye degradation in water (|RhB|0 = 10 mg L−1, VRhB solution = 100 mL, 50 mg of catalyst) under visible light irradiation (300 W Xenon with a UV cut-off filter λ 99% 4-methoxylaniline (4-MBA) selectivity after 40 min (30 ºC, n4-MNB = 6 mmol, 10 mg catalyst, 6 equiv. N2H4·H2O, solvent: 6 mL ethanol). The excellent catalytic performance of the Co-MoO3/NC@SBA-15 was attributed to the Co-Nx(C)-Mo active sites generated through the interaction between the surface Co-Nx(C) and MoO3 species, promoting the dissociation of hydrazine molecule into the active H* species for the reduction of nitro groups. Recycling tests also revealed the high stability and recyclability of 2.0%Co‒MoO3/NC@SBA-15 in the 4-MNB reduction, leading to comparable substrate conversion (70−75% for seven cycles) and 4-MBA selectivity (>99%) without any metal leaching. Moreover, the textural properties of the catalyst were preserved, as well as the size and dispersion of the metal species and the valence states of the elements. Ai et al. presented a novel strategy for the scalable fabrication of nanocomposites of nano-γFe2O3 anchored onto a 3D porous carbon framework (3DPCF) [213]. The nanocomposites (denoted as γ-Fe2O3@3DPCF) were prepared through condensation of iron(III) acetylacetonate with acetylacetonate at room temperature to yield a polymer precursor, which was then carbonized at different temperatures (400  ºC, 600  ºC, and 800  °C), with 800 ºC being considered the optimum temperature. The homogeneous aldol condensation promoted the uniform distribution of the γ-Fe2O3 NPs, with an average size of approximate 20 nm, over the 3DPCF support, leading to a BET

289

290

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

surface area of 125 m2 g−1. The authors claimed that the iron oxide NPs maintained its γ-phase instead of the more stable α-phase due to the capping with carbon. The nanocomposite performed excellently in the reduction of nitroarenes to the corresponding anilines at 100 °C, using N2H4·H2O as the hydrogen source. For instance, the catalyst led to >99% p-nitrophenol with >99% selectivity and 95% product yield after 3 h (100 ºC, nsubstrate = 1 mmol, 14 mg catalyst, nN2 H4 ⋅H2O  = 3 mmol, solvent: 2 mL ethanol). Moreover, it preserved its performance and structure after five subsequent catalytic cycles. This new facile synthetic method provided a new design strategy for the fabrication of 3D porous carbon anchored metal oxides or metal nanocomposites that could be further utilized in other areas including sewage treatment and energy storage. Using a similar approach, Lv et al. described the preparation of porous organic polymers (POPs) through the facile condensation of p-phenylenediamine with ferrocene carboxaldehyde, followed by the carbonization of the ferrocene-functionalized POP material under inert atmosphere at different temperatures (600 ºC, 700 ºC, 800 ºC, 900 ºC, and 1000 ºC), obtaining γ-Fe2O3 supported on NPC catalysts (γ-Fe2O3/NPC) [214]. The γ-Fe2O3/NPC material with the best catalytic performance was prepared using a carbonization temperature of 800 ºC, presenting a surface area of 415 m2 g−1, the highest Fe loading (34.9%), and containing 18 nm γ-Fe2O3 NPs uniformly anchored in the carbon framework. That catalyst showed excellent activity in the reduction of nitroarenes with N2H4·H2O as reducing agent. For example, it promoted 100% of nitrobenzene with >99% selectivity after 80 min at 80 ºC in ethanol (nnitrobenzene = 1 mmol, 20 mg catalyst, 500 μL N2H4·H2O, solvent, 10 mL ethanol). The enhanced performance was attributed to the synergistic effect between the γ-Fe2O3 NPs and NPC, which led to an improvement of the hydrazine hydrate adsorption and activation for hydrogen atoms production. Furthermore, the γ-Fe2O3/NPC catalyst was easily recycled through magnetic separation and reused in nine further cycles without loss of catalytic activity. Later on, in 2021, Shaikh et al. described the preparation of a Co3O4/N-Gr/Fe3O4 magnetic heterostructure composed of Co3O4 NPs (15−20 nm) surrounded by nitrogen-doped graphitic carbon derived from ZIF-67 on a Fe3O4 support (cube-like microcrystals of ∼450 nm) [215]. An N- and Co-containing MOF, ZIF-67, was added to a well dispersed suspension of magnetite and allowed to be impregnated with Fe3O4. After separation of excess ZIF-67, the resulting solid was further thermally treated at 800 °C under N2 atmosphere. The resulting hybrid catalyst showed an impressive catalytic activity, with a very high chemoselectivity toward the hydrogenation of several substrates: (i) N-heteroarenes (e.g., 3-methylquinoline hydrogenation in methanol at 120 ºC and 40 bar H2: 94% conversion and >99% selectivity after 24 h; reaction conditions: 0.5 mmol substrate, 10 mg catalyst, 5  mL anhydrous methanol); (ii) cinnamaldehyde (quantitative conversion and >99% hydrocinnamaldehyde selectivity after 24 h; reaction conditions: 110 ºC and 30 bar H2, 0.25 mmol substrate, 10 mg catalyst, 5 mL toluene); and (iii) nitroarenes (e.g., >99% conversion of nitrobenzene with >99% selectivity; reaction conditions: 100 ºC and 25 bar H2, 0.5 mmol substrate, 10 mg catalyst, 5 mL THF). The authors also assessed the catalyst reusability and concluded that it varied according to the reaction substrate. The catalyst showed a relatively stable performance in the hydrogenation of quinoline for up to five consecutive cycles and then decreased to 72% in the sixth cycle, whereas for the cinnamaldehyde hydrogenation, no significant drop of conversion was observed for up to eight cycles. However, the performance of the hybrid catalyst in the reduction of the nitroarenes significantly decreased after the forth cycle, which the authors attributed to the easier penetration of the graphitic shell by comparatively smaller and stronger coordinating amines that may have led to the deactivation of the catalyst. Nevertheless, a cost-effective nonnoble-metal-based fabrication strategy that provided an efficient, reusable, and magnetically-­ separable Co3O4/N-Gr/Fe3O4 catalyst was demonstrated.

13.6  Conclusions and Future Perspectives

13.6  Conclusions and Future Perspectives In summary, this chapter reviewed the extraordinary progress on carbon-metal oxide hybrid and composite nano(catalysts) for energy-related reactions, biomass valorization, and wastewater treatment. In all processes, the notorious potentialities of multicomponent carbon-metal oxide hybrids and composites were demonstrated, revealing improved (photo)(electro)catalytic efficiency relative to that of the individual counterparts as they act simultaneously as adsorbents and as (photo)(electro)catalysts. Despite the tremendous advances, there are still several challenges to pursue. In the field of energy-related processes, the developed electrocatalysts for WS (HER and OER) do not meet or exceed the performance of precious metals. Another drawback is related to the fact that most HER electrocatalysts perform better in acidic medium, whereas almost all OER are better adapted to alkaline conditions. Hence, it is essential to develop versatile electrocatalysts that work in a wider range of industrial electrolysis conditions. A possible solution would probably be an electrocatalyst operating under neutral conditions to avoid corrosion issues and increase its durability. Regarding the ORR, besides the previously mentioned challenges, the performance of the electrocatalysts based on carbon materials in acidic medium should be improved (typically they work better in alkaline medium) considering a large-scale application in proton-exchange membrane FCs. Within the context of biomass valorization, the use of hybrid carbon-metal oxide catalysts in biorefinery processes is quite well reported, emphasizing the potential of BCH as an alternative to the conventional catalyst supports in biorefineries. The provided examples demonstrated the role of the BCH support and how its properties could be readily modified by changing the activation conditions (chemical or physical activation). Moreover, the metal oxide/BCH hybrids can potentially be an alternative to replace conventional catalysts used for the synthesis of renewable fuels and chemicals in the biorefinery context. In fact, metal oxide supported-BCH catalysts can be employed not only in glucose conversion, but also in lignin depolymerization/hydrogenolysis and many other important reactions to transform biomass in value-added products. It remains a challenge to gain a more precise understanding of the BCH structure and the synergy/interactions between the metal/metal oxide phase and the BCH matrix in order to understand the dominant reaction mechanisms. Regarding the wastewater treatment processes, the overall (photo)(electro)catalytic activity was enhanced by a synergetic effect between the carbon support and the metal oxide component, which mainly resulted from the combination of the catalytic ability induced by the metal oxide with the high available surface area, chemical stability, electrical conductivity, adsorption ability, and/or photosensitizer properties typically provided by the carbon materials. In the future, the development of multicomponent catalysts, namely combining binary or ternary metal oxides and different carbon-based materials can be an added value for a safe and sustainable water supply to take advantage of their synergetic and complementary properties. Also importantly, more attention should be given to the byproducts formed in the pollutants degradation by the AOPs. As example, some pollutants can contain inorganic anions (e.g., chloride ions) in their composition that can be transformed into toxic and carcinogenic byproducts during the AOPs, whereby toxicity analysis of the effluents should be considered. In all applications, most of the studies have been conducted at laboratory scale with scarce or no information on reproducibility. The scale-up production of such catalysts through cost-effective and eco-sustainable routes and the assessment of their performance and chemical stability under real conditions should be pursued. Moreover, a comprehensive cost analysis of the scale-up and prototype installation considering a specific process will be also important to raise the Technology Readiness Level (TRL) of these technologies.

291

292

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

Much work remains to be undertaken on the understanding of the complex structure-function relationships and mechanistic aspects of these hybrid carbon-metal oxide catalyzed reactions. One key point is to unravel what are the actual active sites in all these catalytic reactions. Even though DFT calculations have been used in some cases to predict the reaction intermediates and active sites of the catalysts, the true mechanism of the multicomponent systems is not entirely understood. Computational and theoretical studies must be applied for a systematic investigation that will help to better understand the intermediates, providing an in-depth understanding of the reaction mechanism, which is extremely important for the prediction of newly developed catalysts properties. In situ characterization techniques are also required to further apprehend structural and electronic behavior during the catalytic processes. Furthermore, the use of artificial intelligence tools, namely machine learning, combined with high-throughput experimentation, can allow the refinement of the strategies to engineer and optimize the properties and performance of these multifunctional catalysts and accelerate the understanding of the underlying reaction mechanisms. Hence, although there are still several challenges to pursue, it is with great enthusiasm that we look at these recent advances merging the fields of materials chemistry and catalysis, which are paving the way to a cleaner and more sustainable world.

Acknowledgments This work was funded by Portuguese funds through Fundação para a Ciência e a Tecnologia (FCT)/ MCTES in the framework of the projects UIDB/50006/2020 and UIDP/50006/2020. C.P., D.M.F. and A.F.P. thank FCT for funding through the Individual Call to Scientific Employment Stimulus (Refs. 2021.04120.CEECIND/CP1662/CT0008, 2021.00771.CEECIND/CP1662/CT0007, and 2020.01614.CEECIND/CP1596/CT0007, respectively). M.N. thanks FCT for her working contract in the framework of project BoostEnergy4Tex (Ref. PTDC/CTM-TEX/4126/2021).

References 1 Serp, P. and Machado, B. (2015). Nanostructured Carbon Materials for Catalysis. The Royal Society of Chemistry. 2 Serp, P. and Figueiredo, J.L. (2008). Carbon Materials for Catalysis. John Wiley & Sons. 3 Chaudhry, M.A., Hussain, R., and Butt, F.K. (eds.) (2022). Metal Oxide-Carbon Hybrid Materials: Synthesis, Properties and Applications. Elsevier. 4 Thomas, N., Dionysiou, D.D., and Pillai, S.C. (2021). J. Hazard. Mater. 404: 124082. 5 Liu, J., Peng, C., and Shi, X. (2022). Environ. Pollut. 293: 118565. 6 Gopinath, A., Pisharody, L., Popat, A., and Nidheesh, P.V. (2022). Curr. Opin. Solid State Mater. Sci. 26 (2): 100981. 7 Jorge, A.B., Jervis, R., Periasamy, A.P. et al. (2020). Adv. Energy Mater. 10: 11. 8 Stacy, J., Regmi, Y.N., Leonard, B., and Fan, M. (2017). Renew. Sustain. Energy Rev. 69: 401–414. 9 Freire, C., Fernandes, D.M., Nunes, M., and Abdelkader, V.K. (2018). ChemCatChem. 10 (8): 1703–1730. 10 Shah, S.S.A., Sufyan Javed, M., Najam, T. et al. (2022). Coord. Chem. Rev. 471: 214716. 11 Xu, Y., Fan, K., Zou, Y. et al. (2021). Nanoscale 13 (48): 20324–20353. 12 Shahid, M.K., Batool, A., Kashif, A. et al. (2021). J. Environ. Manage. 297: 113268.

References

13 Ahorsu, R., Constanti, M., and Medina, F. (2021). Ind. Eng. Chem. Res. 60 (51): 18612–18626. 14 Giannakoudakis, D.A., Zormpa, F.F., Margellou, A.G. et al. (2022). Nanomaterials 12: 10. 15 Richardson, S.D. and Kimura, S.Y. (2020). Anal. Chem. 92 (1): 473–505. 16 Ganiyu, S.O., Sable, S., and Gamal El-Din, M. (2022). Chem. Eng. J. 429: 132492. 17 Song, J., Huang, Z.-F., Pan, L. et al. (2018). Appl. Catal. B Environ. 227: 386–408. 18 Formenti, D., Ferretti, F., Scharnagl, F.K., and Beller, M. (2019). Chem. Rev. 119 (4): 2611–2680. 19 Patowary, S., Chetry, R., Goswami, C. et al. (2022). ChemCatChem. 14 (7): e202101472. 20 Ren, S., Duan, X., Liang, S. et al. (2020). J. Mater. Chem. A 8 (13): 6144–6182. 21 Lang, P., Yuan, N., Jiang, Q. et al. (2020). Energy Technol. 8 (3): 1900984. 22 He, Y., Yin, Z., Wang, Z. et al. (2022). J. Mater. Chem. A 10 (18): 9788–9820. 23 Li, Z., Gao, R., Feng, M. et al. (2021). Adv. Energy Mater. 11 (16): 2003291. 24 Feng, Q., Chen, Z., Zhou, K. et al. (2021). ChemistrySelect 6 (31): 8177–8181. 25 Araújo, M.P., Nunes, M., Rocha, I.M. et al. (2019). J. Mater. Sci. 54 (12): 8919–8940. 26 Araújo, M.P., Nunes, M., Rocha, I.M. et al. (2018). ChemistrySelect 3 (35): 10064–10076. 27 Zhou, Q., Su, Z., Tang, Y. et al. (2019) Chem. – A Eur. J. 25 (24): 6226–6232. 28 Gangadharan, P.K., Bhange, S.N., Kabeer, N. et al. (2019). Nanoscale Adv. 1 (8): 3243–3251. 29 Alegre, C., Busacca, C., Di Blasi, A. et al. (2020). ChemElectroChem. 7 (1): 124–130. 30 He, Y., Aasen, D., McDougall, A. et al. (2021). ChemElectroChem 8 (8): 1455–1463. 31 Zhao, J., Liu, J., Jin, C. et al. (2020). Chem. – A Eur. J. 26 (55): 12606–12614. 32 Yuasa, M., Koga, Y., Ueda, H., and Zayasu, T. (2022). J. Appl. Electrochem. 52 (8): 1173–1186. 33 Kéranguéven, G., Bouillet, C., Papaefthymiou, V. et al. (2020). Electrochim. Acta 353: 136557. 34 Mefford, J.T., Kurilovich, A.A., Saunders, J. et al. (2019). Phys. Chem. Chem. Phys. 21 (6): 3327–3338. 35 Zhuang, S., Wang, Z., He, J. et al. (2021). Sustain. Mater. Technol. 29: e00282. 36 Hu, S., Tan, Y., Feng, C. et al. (2019). J. Solid State Electrochem. 23 (8): 2291–2299. 37 Abdelkader-Fernández, V.K., Fernandes, D.M., Balula, S.S. et al. (2019). ACS Appl. Energy Mater. 2 (3): 1854–1867. 38 Fernandes, D.M., Araujo, M.P., Haider, A. et al. (2018). Chemelectrochem. 5 (2): 273–283. 39 Ji, Y.C., Huang, L.J., Hu, J. et al. (2015). Energy Environ. Sci. 8 (3): 776–789. 40 Liu, S.Q. and Tang, Z.Y. (2010). Nano Today 5 (4): 267–281. 41 Ingavale, S., Patil, I., Prabakaran, K., and Swami, A. (2021). Int. J. Energy Res. 45 (5): 7366–7379. 42 Sanij, F.D., Balakrishnan, P., Su, H.N. et al. (2021). Rsc Adv. 11 (62): 39118–39129. 43 Li, L.Q., Yang, J., Yang, H.B. et al. (2018). ACS Appl. Energy Mater. 1 (3): 963–969. 44 Han, X.P., He, G.W., He, Y. et al. (2018). Adv. Energy Mater. 8 (10): 1702222. 45 Niu, Y.L., Huang, X.Q., Zhao, L. et al. (2018). ACS Sustain. Chem. Eng. 6 (3): 3556–3564. 46 Zhang, Y., Wang, X.X., Luo, F.Q. et al. (2019). Appl. Catal. B-Environmental 256: 117852. 47 Ouyang, T., Wang, X.T., Mai, X.Q. et al. (2020). Angew. Chemie-International Ed. 59 (29): 11948–11957. 48 Lee, C., Jeon, D., Park, J. et al. (2020). ACS Appl. Mater. Interfaces 12 (29): 32689–32697. 49 Nagaiah, T.C., Gupta, D., Das Adhikary, S. et al. (2021). J. Mater. Chem. A 9 (14): 9228–9237. 50 Limani, N., Marques, I.S., Jarrais, B. et al. (2022). Catalysts 12 (4): 357. 51 Abdelkader-Fernandez, V.K., Fernandes, D.M., Cunha-Silva, L. et al. (2021). Electrochim. Acta 389: 138719. 52 Marques, I.S., Jarrais, B., Mbomekalle, I.M. et al. (2022). Catalysts 12 (4): 400. 53 Seh, Z.W., Kibsgaard, J., Dickens, C.F. et al. (2017). Science 355 (6321): eaad4998. 54 Liu, C.H., Wang, K., Zhang, J. et al. (2018). J. Mater. Sci. Electron. 29 (13): 10744–10752. 55 Navarro-Pardo, F., Liu, J.B., Abdelkarim, O. et al. (2020). Adv. Funct. Mater. 30 (14): 1908467.

293

294

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

56 Srinivas, K., Chen, Y.F., Su, Z. et al. (2022). Electrochim. Acta 404: 139745. 57 Jayaseelan, S.S., Bhuvanendran, N., Xu, Q., and Su, H.N. (2020). Int. J. Hydrogen Energy 45 (7): 4587–4595. 58 Wu, J.Q., Zhao, J.W., and Li, G.R. (2020). Energy & Fuels 34 (7): 9050–9057. 59 Qin, Q., Chen, L.L., Wei, T. et al. (2019). Catal. Sci. Technol. 9 (7): 1595–1601. 60 Xiao, Y., Zhang, P.F., Zhang, X. et al. (2017). J. Mater. Chem. A 5 (30): 15901–15912. 61 Liu, S.L., Chen, C., Zhang, Y.F. et al. (2019). J. Mater. Chem. A 7 (24): 14466–14472. 62 Begum, H., Ahmed, M.S., and Jeon, S. (2019). Electrochim. Acta 296: 235–242. 63 Jiang, J.B., Zhu, L.Y., Sun, Y.X. et al. (2019). J. Power Sources 426: 74–83. 64 Guo, H.L., Zhou, J., Li, Q.Q. et al. (2020). Adv. Funct. Mater. 30 (15): 2000024. 65 Liu, X.X., Zang, J.B., Chen, L. et al. (2017). J. Mater. Chem. A 5 (12): 5865–5872. 66 Karuppasamy, K., Jothi, V.R., Vikraman, D. et al. (2019). Appl. Surf. Sci. 478: 916–923. 67 Li, C.Y., Zhao, S.Y., Zhu, K.L. et al. (2020). J. Mater. Chem. A 8 (30): 14944–14954. 68 Jing, S.Y., Lu, J.J., Yu, G.T. et al. (2018). Adv. Mater. 30 (28): 1705979. 69 Chen, J.D., Yu, D.N., Liao, W.S. et al. (2016). ACS Appl. Mater. Interfaces 8 (28): 18132–18139. 70 Hu, G.J., Li, J., Liu, P. et al. (2019). Appl. Surf. Sci. 463: 275–282. 71 Askari, M.B. and Salarizadeh, P. (2019). J. Mol. Liq. 291: 111306. 72 Galal, A., Hassan, H.K., Atta, N.F., and Jacob, T. (2017). Chemistryselect 2 (31): 10261–10270. 73 Mukherjee, A., Chakrabarty, S., Su, W.N., and Basu, S. (2018). Mater. Today Energy 8: 118–124. 74 Li, N., Liu, J., Dong, B.X., and Lan, Y.Q. (2020). Angew. Chemie-International Ed. 59 (47): 20779–20793. 75 Horn, M.R., Singh, A., Alomari, S. et al. (2021). Energy Environ. Sci. 14 (4): 1652–1700. 76 Ge, J.X., Hu, J., Zhu, Y.T. et al. (2020). Acta Physico-Chimica Sin. 36 (1): 1906063. 77 Wang, Z.H., Wang, X.F., Tan, Z., and Song, X.Z. (2021). Mater. Today Energy 19: 100618. 78 Zhang, Y., Liu, J., Li, S.L. et al. (2019). Energychem 1 (3): 100021. 79 Jawale, D.V., Fossard, F., Miserque, F. et al. (2022). Carbon N. Y. 188: 523–532. 80 Fernandes, D.M., Peixoto, A.F., and Freire, C. (2019). Dalt. Trans. 48 (36): 13508–13528. 81 Abdelkader-Fernandez, V.K., Fernandes, D.M., and Freire, C. (2020). 42: 101350. 82 Fan, Q., Zhang, M.L., Jia, M.W. et al. (2018). Mater. Today Energy 10: 280–301. 83 Sekar, P., Calvillo, L., Tubaro, C. et al. (2017). ACS Catal. 7 (11): 7695–7703. 84 Zhang, Q., Du, J., He, A.B. et al. (2019). J. Solid State Chem. 279: 120946. 85 Miao, Z.P., Hu, P., Nie, C.Y. et al. (2019). J. Energy Chem. 38: 114–118. 86 Yan, Y., Ke, L.W., Ding, Y. et al. (2021). Mater. Chem. Front. 5 (6): 2668–2683. 87 Zhang, B.H. and Zhang, J.T. (2017). J. Energy Chem. 26 (6): 1050–1066. 88 Wang, Y.H., Liu, J.L., and Zheng, G.F. (2021). Designing copper-based catalysts for efficient carbon dioxide electroreduction. Adv. Mater. 33 (46): 2005798. 89 Xie, H., Wang, T.Y., Liang, J.S. et al. (2018). Nano Today 21: 41–54. 90 Zhao, J., Xue, S., Barber, J. et al. (2020). J. Mater. Chem. A 8 (9): 4700–4734. 91 Ning, H., Wang, X.S., Wang, W.H. et al. (2019). Carbon N. Y. 146: 218–223. 92 Rashid, N., Bhat, M.A., Das, A., and Ingole, P.P. (2020). 39: 101178. 93 Altaf, N., Liang, S.Y., Iqbal, R. et al. (2020). 40: 101205. 94 Zhang, Y.Y., Li, K., Chen, M.M. et al. (2020). ACS Appl. Nano Mater. 3 (1): 257–263. 95 Zhi, W.Y., Liu, Y.T., Shan, S.L. et al. (2021). J. CO2 Util. 50: 101594. 96 Li, D., Liu, T.T., Huang, L.L. et al. (2020). J. Mater. Chem. A 8 (35): 18302–18309. 97 Fu, Y.Y., Wang, T.T., Zheng, W.Z. et al. (2020). ACS Appl. Mater. Interfaces 12 (14): 16178–16185. 98 Lee, M.Y., Han, S., Lim, H. et al. (2020). ACS Sustain. Chem. Eng. 8 (5): 2117–2121. 99 Jing, Y.X., Guo, Y., Xia, Q.N. et al. (2019). Chem 5 (10): 2520–2546.

References

100 Siddiki, S.M.A.H. and Touchy, A.S. (2020). Chapter 10 - Challenges and future prospects in heterogeneous catalysis for biorefinery technologies. In: Advanced Functional Solid Catalysts for Biomass Valorization (eds. C. Mustansar Hussain and P. Sudarsanam), 225–250. Elsevier. 101 Sudarsanam, P., Zhong, R.Y., Van den Bosch, S. et al. (2018). Chem. Soc. Rev. 47 (22): 8349–8402. 102 Kashyap, P., Bhardwaj, S., Chodimella, V.P., and Sinha, A.K. (2022). Biomass Convers. Biorefinery. https://doi.org/10.1007/s13399-022-02675-y 103 Koley, P., Chandra Shit, S., Joseph, B. et al. (2020). ACS Appl. Mater. Interfaces 12 (19): 21682–21700. 104 Liu, B. and Zhang, Z. (2016). ACS Catal. 6 (1): 326–338. 105 Lam, E. and Luong, J.H.T. (2014). ACS Catal. 4 (10): 3393–3410. 106 Khodafarin, R., Tavasoli, A., and Rashidi, A. (2020). Biomass Convers. Biorefinery 12 (1): 5813–5824. 107 Siddiqui, M.T.H., Baloch, H.A., Nizamuddin, S. et al. (2021). Renew. Energy 172: 1103–1119. 108 Yu, I.K.M., Xiong, X., Tsang, D.C.W. et al. (2019). Green Chem. 21 (16): 4341–4353. 109 Rusanen, A., Kupila, R., Lappalainen, K. et al. (2020). Catalysts 10 (8): 821. 110 Sudarsanam, P., Gupta, N.K., Mallesham, B. et al. (2021). ACS Catal. 11 (21): 13603–13648. 111 Ramos, R., Abdelkader-Fernández, V.K., Matos, R. et al. (2022). Catalysts 12 (2): 207. 112 Peixoto, A.F., Ramos, R., Moreira, M.M. et al. (2021). Fuel 303: 121227. 113 Lee, Y., Lee, S.W., Tsang, Y.F. et al. (2020). Chem. Eng. J. 387: 124194. 114 Lee, Y., Kim, Y.T., Kwon, E.E., and Lee, J. (2020). Environ. Res. 184: 109325. 115 Liu, J.L., Yang, L.H., Shuang, E. et al. (2022). Fuel 315: 123172. 116 Marianou, A.A., Michailof, C.M., Pineda, A. et al. (2018). Appl. Catal. A Gen. 555: 75–87. 117 Xiong, X.N., Yu, I.K.M., Tsang, D.C.W. et al. (2020). J. Clean. Prod. 268: 122378. 118 Yu, I.K.M., Xiong, X., Tsang, D.C.W. et al. (2019). Green Chem. 21 (6): 1267–1281. 119 Yang, X., Yu, I.K.M., Cho, D.-W. et al. (2019). ACS Sustain. Chem. Eng. 7 (5): 4851–4860. 120 Liu, J.L., Yang, M., Gong, C.X. et al. (2021). J. Environ. Chem. Eng. 9 (6): 106721. 121 Zhang, Y., Wang, J.G., Wang, J.H. et al. (2019). Chemistryselect 4 (19): 5724–5731. 122 Kupila, R., Lappalainen, K., Hu, T. et al. (2021). Appl. Catal. A Gen. 612: 118011. 123 Kupila, R., Lappalainen, K., Hu, T. et al. (2021). Appl. Catal. A Gen 624: 118327. 124 Gurrala, L., Kumar, M.M., Sharma, S. et al. (2022). Fuel 308: 121818. 125 Cheng, C., Shen, D., Gu, S., and Luo, K.H. (2018). Catal. Sci. Technol. 8 (24): 6275–6296. 126 Wang, -Y.-Y., Ling, -L.-L., and Jiang, H. (2016). Green Chem. 18 (14): 4032–4041. 127 Xin, L., Hu, J., Xiang, Y. et al. (2021). Materials. 14 (10): 2643. 128 Vagi, M.C. and Petsas, A.S. (2020). J. Environ. Chem. Eng. 8 (1): 102940. 129 Ma, D., Yi, H., Lai, C. et al. (2021). Chemosphere 275: 130104. 130 Brillas, E. (2022). Sep. Purif. Technol. 284: 120290. 131 Ramos, M.D.N., Santana, C.S., Velloso, C.C.V. et al. (2021). Process Saf. Environ. Prot. 155: 366–386. 132 Pesqueira, J.F.J.R., Pereira, M.F.R., and Silva, A.M.T. (2020). J. Clean. Prod. 261: 121078. 133 Vorontsov, A.V. (2019). J. Hazard. Mater. 372: 103–112. 134 Fu, W., Yi, J., Cheng, M. et al. (2022). J. Hazard. Mater. 424: 127419. 135 Lai, C., Shi, X., Li, L. et al. (2021). Sci. Total Environ. 775: 145850. 136 Luo, H., Zeng, Y., He, D., and Pan, X. (2021). Chem. Eng. J. 407: 127191. 137 Ribeiro, J.P. and Nunes, M.I. (2021). Environ. Res. 197: 110957. 138 Dihingia, H. and Tiwari, D. (2022). J. Water Process Eng. 45: 102500. 139 Scaria, J., Gopinath, A., and Nidheesh, P.V. (2021). J. Clean. Prod. 278: 124014. 140 Leonel, A.G., Mansur, A.A.P., and Mansur, H.S. (2021). Water Res. 190: 116693.

295

296

13  Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and Wastewater Treatment

141 Wang, J. and Tang, J. (2021). J. Mol. Liq. 332: 115755. 142 Ribeiro, R.S., Silva, A.M.T., Figueiredo, J.L. et al. (2016). Appl. Catal. B Environ. 187: 428–460. 143 Hu, X., Liu, B., Deng, Y. et al. (2011). Appl. Catal. B Environ. 107 (3): 274–283. 144 Hu, X., Deng, Y., Gao, Z. et al. (2012). Appl. Catal. B Environ. 127: 167–174. 145 Zhou, L., Zhang, H., Ji, L. et al. (2014). RSC Adv. 4 (47): 24900–24908. 146 Peng, X., Qu, J., Tian, S. et al. (2016). RSC Adv. 6 (106): 104549–104555. 147 Zheng, X., Cheng, H., Yang, J. et al. (2018). ACS Appl. Nano Mater. 1 (6): 2754–2762. 148 Ding, Q., Lam, F.L.Y., and Hu, X. (2019). J. Environ. Manage. 244: 23–32. 149 Lu, K., Yang, F., Lin, W. et al. (2018). ChemistrySelect 3 (16): 4207–4216. 150 Qin, H., Cheng, H., Li, H., and Wang, Y. (2020). Chem. Eng. J. 396: 125304. 151 Brillas, E. and Garcia-Segura, S. (2020). Sep. Purif. Technol. 237: 116337. 152 Shi, T., Peng, J., Chen, J. et al. (2017). Catal. Letters 147 (6): 1598–1607. 153 Tolba, A., Gar Alalm, M., Elsamadony, M. et al. (2019). Process Saf. Environ. Prot. 128: 273–283. 154 Nawaz, M., Shahzad, A., Tahir, K. et al. (2020). Chem. Eng. J. 382: 123053. 155 Qiu, B., Li, Q., Shen, B. et al. (2016). Appl. Catal. B Environ. 183: 216–223. 156 Liu, Y., Jin, W., Zhao, Y. et al. (2017). Appl. Catal. B Environ. 206: 642–652. 157 Nadeem, N., Zahid, M., Tabasum, A. et al. (2020). Mater. Res. Express 7 (1): 015519. 158 Li, D., Yang, T., Liu, Z. et al. (2022). Sci. Total Environ. 824: 153772. 159 Liang, C., Liu, Y., Li, K. et al. (2017). Sep. Purif. Technol. 188: 105–111. 160 Wang, K., Niu, H., Chen, J. et al. (2017). Appl. Surf. Sci. 404: 138–145. 161 Saleh, R. and Taufik, A. (2019). Sep. Purif. Technol. 210: 563–573. 162 Matos, R., Nunes, M.S., Kuźniarska-Biernacka, I. et al. (2021). Eur. J. Inorg. Chem. 2021 (47): 4915–4928. 163 Bagherzadeh, S.B., Kazemeini, M., and Mahmoodi, N.M. (2021). J. Colloid Interface Sci. 602: 73–94. 164 Nair, K.M., Kumaravel, V., and Pillai, S.C. (2021). Chemosphere 269: 129325. 165 Li, J., Ai, Z., and Zhang, L. (2009). J. Hazard. Mater. 164 (1): 18–25. 166 Wang, Y., Zhao, G., Chai, S. et al. (2013). ACS Appl. Mater. Interfaces 5 (3): 842–852. 167 Cao, X., Jiang, D., Huang, M. et al. (2020). Colloids Surfaces A Physicochem. Eng. Asp. 587: 124269. 168 Wang, Y., Zhao, H., and Zhao, G. (2016). Electroanalysis 28 (1): 169–176. 169 Chen, W., Yang, X., Huang, J. et al. (2016). Electrochim. Acta 200: 75–83. 170 Mohseni, M., Demeestere, K., Du Laing, G. et al. (2021). Adv. Sustain. Syst. 5 (4): 2100001. 171 Zhang, Y., Chen, Z., Wu, P. et al. (2020). J. Hazard. Mater. 393: 120448. 172 Barhoum, A., Favre, T., Sayegh, S. et al. (2021). Nanomater. 11 (10): 2686. 173 Han, Z., Li, Z., Li, Y. et al. (2022). Chemosphere 287: 132154. 174 Guo, M., Lu, M., Zhao, H. et al. (2022). J. Hazard. Mater. 423: 127033. 175 Dung, N.T., Duong, L.T., Hoa, N.T. et al. (2022). Chemosphere 287: 132141. 176 Cui, L., Li, Z., Li, Q. et al. (2021). Chem. Eng. J. 420: 127666. 177 Dong, P., Chen, X., Guo, M. et al. (2021). J. Hazard. Mater. 412: 125208. 178 Liu, X., Xie, L., Liu, Y. et al. (2020). Catal. Today 355: 458–465. 179 Xie, L., Mi, X., Liu, Y. et al. (2019). ACS Appl. Mater. Interfaces 11 (34): 30703–30712. 180 Das, R., Vecitis, C.D., Schulze, A. et al. (2017). Chem. Soc. Rev. 46 (22): 6946–7020. 181 Ganguly, P., Panneri, S., Hareesh, U.S. et al. (2019). Chapter 23 - Recent advances in photocatalytic detoxification of water. In: Micro and Nano Technologies (eds. S. Thomas, D. Pasquini, S.-Y. Leu, and D.A.B.T.-N.M. Gopakumar in W.P.), 653–688. Elsevier. 182 Nakata, K. and Fujishima, A. (2012). J. Photochem. Photobiol. C Photochem. Rev. 13 (3): 169–189.

References

183 Nagaraj, G., Dhayal Raj, A., Albert Irudayaraj, A., and Josephine, R.L. (2019). Optik (Stuttg). 179: 889–894. 184 Zhang, J., Zhou, P., Liu, J., and Yu, J. (2014). Phys. Chem. Chem. Phys. 16 (38): 20382–20386. 185 Al-Mamun, M.R., Kader, S., Islam, M.S., and Khan, M.Z.H. (2019). J. Environ. Chem. Eng. 7 (5): 103248. 186 Cheng, G., Xu, F., Xiong, J. et al. (2016). Adv. Powder Technol. 27 (5): 1949–1962. 187 Youssef, Z., Colombeau, L., Yesmurzayeva, N. et al. (2018). Dye. Pigment. 159: 49–71. 188 Krishna, V., Yanes, D., Imaram, W. et al. (2008). Appl. Catal. B Environ. 79 (4): 376–381. 189 Zheng, X., Xu, S., Wang, Y. et al. (2018). J. Colloid Interface Sci. 527: 202–213. 190 Li, L., Zheng, X., Chi, Y. et al. (2020). J. Hazard. Mater. 383: 121211. 191 Wang, Q., Huang, J., Sun, H. et al. (2017). Nanoscale 9 (41): 16046–16058. 192 Colmenares, J.C., Varma, R.S., and Lisowski, P. (2016). Green Chem. 18 (21): 5736–5750. 193 Dalto, F., Kuźniarska-Biernacka, I., Pereira, C. et al. (2021). Nanomaterials 11 (11): 3016. 194 Antonopoulou, M., Karagianni, P., Giannakas, A. et al. (2017). Catal. Today 280: 114–121. 195 Lisowski, P., Colmenares, J.C., Mašek, O. et al. (2018). J. Anal. Appl. Pyrolysis 131: 35–41. 196 Zhang, Y., Zhao, Z., Chen, J. et al. (2015). Appl. Catal. B Environ. 165: 715–722. 197 Oseghe, E.O. and Ofomaja, A.E. (2018). J. Environ. Manage. 223: 860–867. 198 Imam, S.S., Adnan, R., and Mohd Kaus, N.H. (2021). J. Environ. Chem. Eng. 9 (4): 105404. 199 Alansi, A.M., Al-Qunaibit, M., Alade, I.O. et al. (2018). J. Mol. Liq. 253: 297–304. 200 Jiang, T., Li, J., Sun, Z. et al. (2016). Ceram. Int. 42 (15): 16463–16468. 201 Liu, D., Xie, J., and Xia, Y. (2019). Chem. Phys. Lett. 729: 42–48. 202 Fu, L., Xia, T., Zheng, Y. et al. (2015). Ceram. Int. 41 (4): 5903–5908. 203 Raizada, P., Sudhaik, A., and Singh, P. (2019). Mater. Sci. Energy Technol. 2 (3): 509–525. 204 Shalahuddin Al Ja’farawy, M., Kusumandari, P.A., and Widiyandari, H. (2022). Environ. Nanotechnology, Monit. Manag. 18: 100681. 205 Downing, R.S., Kunkeler, P.J., and Van Bekkum, H. (1997). Catal. Today 37 (2): 121–136. 206 Ono, N. (2001). The Nitro Group in Organic Synthesis, John Wiley–VCH. 207 Zhang, L., Zhou, M., Wang, A., and Zhang, T. (2020). Chem. Rev. 120 (2): 683–733. 208 Corma, A. and Serna, P. (2006). Science (80-.). 313 (5785): 332–334. 209 Westerhaus, F.A., Jagadeesh, R.V., Wienhöfer, G. et al. (2013). Nat. Chem. 5 (6): 537–543. 210 Chen, B., Li, F., Huang, Z., and Yuan, G. (2016). ChemCatChem. 8 (6): 1132–1138. 211 Pan, X., Gao, X., Chen, X. et al. (2017). ACS Catal. 7 (10): 6991–6998. 212 Huang, H., Liang, X., Wang, X. et al. (2018). Appl. Catal. A Gen. 559: 127–137. 213 Ai, Y., He, M., Lv, Q. et al. (2018). Chem. - An Asian J. 13 (1): 89–98. 214 Lv, J., Liu, Z., and Dong, Z. (2020). Mol. Catal. 498: 111249. 215 Shaikh, M.N., Abdelnaby, M.M., Hakeem, A.S. et al. (2021). ACS Appl. Nano Mater. 4 (4): 3508–3518.

297

Catalysis for a Sustainable Environment

Catalysis for a Sustainable Environment Reactions, Processes and Applied Technologies Volume 2

Edited by Professor Armando J. L. Pombeiro Instituto Superior técnico Lisboa, Portugal

Dr. Manas Sutradhar

Universidade Lusófona de Humanidades e Tecnologias Faculdade de Engenharia Lisboa, Portugal

Professor Elisabete C. B. A. Alegria Instituto Politécnico de Lisboa Departamento de Engenharia Química Lisboa, Portugal

This edition first published 2024 © 2024 John Wiley and Sons Ltd All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/ permissions. The right of Armando J.L. Pombeiro, Manas Sutradhar, and Elisabete C.B.A. Alegria to be identified as the author of the editorial material in this work has been asserted in accordance with law. Registered Offices John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Editorial Office The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com. Wiley also publishes its books in a variety of electronic formats and by print-on-demand. Some content that appears in standard print versions of this book may not be available in other formats. Trademarks: Wiley and the Wiley logo are trademarks or registered trademarks of John Wiley & Sons, Inc. and/ or its affiliates in the United States and other countries and may not be used without written permission. All other trademarks are the property of their respective owners. John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book. Limit of Liability/Disclaimer of Warranty In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. A catalogue record for this book is available from the Library of Congress Hardback ISBN: 9781119870524; ePub ISBN: 9781119870630; ePDF ISBN: 9781119870623; oBook ISBN: 9781119870647 Cover image: © Sasha Fenix/Shutterstock Cover design by Wiley Set in 9.5/12.5pt STIXTwoText by Integra Software Services Pvt. Ltd, Pondicherry, India

v

Contents VOLUME 1 About the Editors  xiii Preface  xv

1

Introduction  1

Armando J.L. Pombeiro, Manas Sutradhar, and Elisabete C.B.A. Alegria

Part I Carbon Dioxide Utilization  5 2

Transition from Fossil-C to Renewable-C (Biomass and CO2) Driven by Hybrid Catalysis  7 Michele Aresta and Angela Dibenedetto

3

Synthesis of Acetic Acid Using Carbon Dioxide  25

Philippe Kalck

4

New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources: A New Catalytic Challenge  35

Ana B. Paninho, Malgorzata E. Zakrzewska, Leticia R.C. Correa, Fátima Guedes da Silva, Luís C. Branco, and Ana V.M. Nunes

5

Sustainable Technologies in CO2 Utilization: The Production of Synthetic Natural Gas  55 M. Carmen Bacariza, José M. Lopes, and Carlos Henriques

6

Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis  73

Denzil Moodley, Thys Botha, Renier Crous, Jana Potgieter, Jacobus Visagie, Ryan Walmsley, and Cathy Dwyer

7

Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives  117

Maurizio Peruzzini, Fabrizio Mani, and Francesco Barzagli

vi

Contents

Part II 8

Transformation of Volatile Organic Compounds (VOCs)  139

Catalysis Abatement of NOx/VOCs Assisted by Ozone  141 Zhihua Wang and Fawei Lin

9

Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions  161

Elisabete C.B.A. Alegria, Manas Sutradhar, and Tannistha R. Barman

10

Catalytic Cyclohexane Oxyfunctionalization  181

Manas Sutradhar, Elisabete C.B.A. Alegria, M. Fátima C. Guedes da Silva, and Armando J.L. Pombeiro

Part III 11

Carbon-based Catalysis  207

Carbon-based Catalysts for Sustainable Chemical Processes  209 Katarzyna Morawa Eblagon, Raquel P. Rocha, M. Fernando R. Pereira, and José Luís Figueiredo

12

Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes  225

Giulia Tuci, Andrea Rossin, Matteo Pugliesi, Housseinou Ba, Cuong Duong-Viet, Yuefeng Liu, Cuong Pham-Huu, and Giuliano Giambastiani

13

Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and, Wastewater Treatment: Cutting-Edge Solutions for a Sustainable World  247 Clara Pereira, Diana M. Fernandes, Andreia F. Peixoto, Marta Nunes, Bruno Jarrais, Iwona Kuźniarska-Biernacka, and Cristina Freire

VOLUME 2 About the Editors  xiii Preface  xv Part IV

14 14.1 14.2 14.2.1 14.2.2 14.2.3 14.3

Coordination, Inorganic, and Bioinspired Catalysis  299

Hydroformylation Catalysts for the Synthesis of Fine Chemicals  301 Mariette M. Pereira, Rui M.B. Carrilho, Fábio M.S. Rodrigues, Lucas D. Dias, and Mário J.F. Calvete Introduction  301 Homogeneous Catalytic Systems  303 Development of Phosphorus Ligands  303 Hydroformylation of Biologically Relevant Substrates  307 Hydroformylation-based Sequential Reactions  309 Heterogeneized Catalytic Systems  313

Contents

14.4

Conclusions  320 References  320

15

Synthesis of New Polyolefins by Incorporation of New Comonomers  323

15.5

Kotohiro Nomura and Suphitchaya Kitphaitun Introduction  323 Synthesis of New Ethylene Copolymers by Incorporation of Sterically Encumbered Olefins, Cyclic Olefins  325 Ethylene Copolymerization with Sterically Encumbered Olefins  325 Ethylene Copolymerization with Cyclic Olefins  327 Ethylene Copolymerization with Alken-1-ol for Introduction of Hydroxy Groups into Polylefins  330 Synthesis of Biobased Ethylene Copolymers by the Incorporation of Linear and Cyclic Terpenes  332 Concluding Remarks and Outlook  334 Acknowledgements  335 References  335

16

Catalytic Depolymerization of Plastic Waste  339

15.1 15.2 15.2.1 15.2.2 15.3 15.4

16.1 16.2 16.3 16.4 16.4.1 16.4.2 16.4.3 16.5 16.5.1 16.5.2 16.5.3 16.5.4 16.6 16.7 16.8 16.9

17 17.1 17.2 17.3 17.4 17.5 17.6

Noel Angel Espinosa-Jalapa and Amit Kumar Introduction  339 Pyrolysis  340 Gasification  340 Solvolysis  342 Hydrolysis  342 Glycolysis and Methanolysis  344 Aminolysis  344 Hydrogenation  345 Hydrogenative Depolymerisation of Polyesters  345 Hydrogenative Depolymerisation of Polycarbonates  348 Hydrogenative Depolymerisation of Nylons and Polyurethanes  350 Hydrogenative Depolymerisation of Polyureas  354 Depolymerisation of Polyethylene Using Alkane Metathesis  356 Hydrogenolysis  357 Hydrosilylation and Hydroboration  359 Summary  363 References  363

Bioinspired Selective Catalytic C-H Oxygenation, Halogenation, and Azidation of Steroids  369

Konstantin P. Bryliakov Introduction  369 The Mechanistic Frame of Bioinspired Oxidative C-H Functionalizations  370 Transition Metal Catalyzed C-H Oxofunctionalizations: Early Examples  372 Transition Metal Catalyzed C-H Oxofunctionalizations with H2O2  375 Transition Metal Catalyzed Halogenations and Azidations  381 Conclusions and Outlook  383 Acknowldegment  384 References  384

vii

viii

Contents

18 18.1 18.2 18.3 18.4 18.5

19 19.1 19.2 19.3 19.4 19.5 19.6

20 20.1 20.2 20.2.1 20.2.2 20.3 20.3.1 20.3.2 20.3.3 20.3.4 20.3.5 20.3.6 20.3.7 20.3.8 20.3.9 20.4

21 21.1

Catalysis by Pincer Compounds and Their Contribution to Environmental and Sustainable Processes  389 Hugo Valdés and David Morales-Morales Introduction  389 Hydrogenation of CO2 to Methanol and Related Processes  390 Hydrogen Production from Methanol and Water  399 Biomass Transformation  402 Conclusions  403 Acknowledgments  404 References  404

Heterometallic Complexes: Novel Catalysts for Sophisticated Chemical Synthesis  409 Franco Scalambra, Ismael Francisco Díaz-Ortega, and Antonio Romerosa Introduction  409 C-X Formation (X = C, H, N, O, Metal)  410 Oxidation Processes  414 CO2 Activation  416 O2 and H2 Generation from Water  420 Conclusions  424 Note  424 Acknowledgments  424 References  425

Metal-Organic Frameworks in Tandem Catalysis  429 Anirban Karmakar and Armando J.L. Pombeiro Introduction  429 MOFs as Catalysts for Tandem Reactions  430 Active Sites in MOFs  430 Advantages and Limitations of MOFs  431 Examples of MOFs Used as Catalysts for Tandem Reactions  431 Deacetalization-Knoevenagel Condensation  431 Deacetalization-Henry Reaction  433 Meinwald rearrangement-Knoevenagel Condensation  434 Reductive Amination of Aldehydes with Nitroarenes  435 Epoxidation–Ring Opening of Epoxide  436 Oxidation-Esterification  437 Oxidation-Hemiacetal Reaction  438 Asymmetric Tandem Reactions  439 Photocatalytic Tandem Reactions  440 Conclusions  441 Acknowledgements  442 References  442 (Tetracarboxylate)bridged-di-transition Metal Complexes and Factors Impacting Their Carbene Transfer Reactivity  445

LiPing Xu, Adrian Varela-Alvarez, and Djamaladdin G. Musaev Introduction  445

Contents

21.2 21.3 21.3.1 21.3.2

Computational Procedure  447 Results and Discussion  447 Geometrical and Electronic Structures of the (RCOO)4-[M2] Complexes  447 Diazocarbene Decomposition by the Reported [RCOO]4–[M2] Complexes  455 Acknowledgement  457 Note  457 References  457

22 22.1 22.2 22.2.1 22.2.1.1 22.2.1.2 22.2.1.3 22.2.1.4 22.2.2 22.2.2.1 22.2.2.2 22.2.2.3 22.2.2.4 22.2.2.5 22.2.2.6 22.2.2.7 22.2.3 22.2.3.1 22.2.3.1.1 22.2.3.1.2 22.2.3.1.3 22.2.3.2 22.3

Sustainable Cu-based Methods for Valuable Organic Scaffolds  461 Argyro Dolla, Dimitrios Andreou, Ethan Essenfeld, Jonathan Farhi, Ioannis N. Lykakis, and George E. Kostakis Introduction  461 Synthetic Aspects  462 Propargylamines  462 Seminal Work  462 Propargylamines from in Situ Generated Cu Catalytic Species  463 Propargylamines from Well-Characterised Cu Catalytic Species  464 Propargylamines as Intermediates  465 Pyrroles  466 General Synthetic Aspects of Pyrroles  466 Cyclization/annulation Reactions of Enamino Compounds  467 Condensation/Cyclization with Diazo Compounds or Oximes  468 Condensation/cyclization Reactions with Alkynes or Isonitriles  468 Reaction of Nitroalkenes  469 Reaction of Dicarbonyl Compounds  470 Miscellaneous Reactions  470 Dihydropyridines  471 Synthesis of 1,4-DHPs with Copper Catalysts  472 Hantzsch-Type Reactions  472 Nucleophilic Addition to Iminium Salts  473 Cycloaddition and C–C/C–N Coupling  474 Synthesis of 1,2-DHPs with Copper Catalysts  475 Conclusions  476 References  476

23

Environmental Catalysis by Gold Nanoparticles  481

23.1 23.2 23.2.1 23.2.2 23.2.3 23.2.4 23.2.5 23.2.6 23.2.7 23.2.8

Sónia Alexandra Correia Carabineiro Introduction  481 Preparation Methods  482 Sol-immobilisation (COL)  482 Impregnation (IMP) and Double Impregnation (DIM)  483 Co-Precipitation (CP)  483 Deposition Precipitation (DP)  484 Liquid-phase Reductive Deposition (LPRD)  484 Ion-Exchange  485 Photochemical Deposition (PD)  485 Ultrasonication (US)  486

ix

x

Contents

23.2.9 23.2.10 23.2.11 23.3 23.3.1 23.3.2 23.3.3 23.3.4 23.4 23.4.1 23.4.2 23.4.3 23.4.4 23.5

Vapor-Phase and Grafting  486 Bi- and Tri-metallic Au Catalysts  487 Post-treatment and Storage  487 Properties of Gold Nanoparticles  487 Activity  487 Selectivity  488 Durability  489 Poison Resistance  490 Reactions Efficiently Catalyzed by Gold Nanoparticles  490 CO Oxidation  490 Preferential Oxidation of CO in the Presence of H2 (PROX)  496 Water-gas Shift  499 Total Oxidation of Volatile Organic Compounds (VOCs)  502 Conclusions and Outlook  505 Acknowledgements  505 References  505

24

Platinum Complexes for Selective Oxidations in Water  515

24.1 24.2 24.3 24.4 24.5 24.6 24.7 24.8

Alessandro Scarso, Paolo Sgarbossa, Roberta Bertani, and Giorgio Strukul Hydrogen Peroxide and Its Activation  515 Platinum Complexes  516 Enantioselective Oxidations  520 Water as the Reaction Medium  522 Catalytic Oxidation Reactions in Water  524 The Catalyst/micelle Interaction  527 Environmental Acceptability Evaluation and Possible Industrial Applications  530 Conclusions  532 References  532

25

The Role of Water in Reactions Catalyzed by Transition Metals  537

25.1 25.2 25.2.1 25.2.2 25.2.3 25.2.4 25.2.5 25.2.6 25.2.6.1 25.2.6.2 25.2.6.3 25.3 25.3.1 25.3.2 25.3.3 25.3.4 25.3.5

A.W. Augustyniak and A.M. Trzeciak Water as a Solvent in Organic Reactions  537 The Role of Water in Heterogeneous Catalytic Systems  539 The Transformations of Furfuryl Derivatives  539 Oxidation and Deoxygenation  540 Arylcyanation and C–C Cross-Coupling  542 Hydrogenation  544 Hydroformylation  545 Catalytic Reactions with MOF-based Catalysts in an Aqueous Medium  546 Cross-Coupling Reactions  546 Hydrogenation Reactions  548 Hydroamination  549 The Contribution of Water in Homogeneous Catalytic Systems  550 Oxidation and Epoxidation  550 The Hydrogenation of Carbonyl Compounds and CO2  550 The Cyclotrimerization of Alkynes  552 The Isomerization of Allylic Alcohols  553 Hydroarylation with Boron Compounds  554

Contents

25.3.6 25.4

Hydroformylation  555 Conclusions  556 References  556

26

Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts  559

26.1 26.2 26.3 26.4 26.5 26.6 26.7 26.7.1 26.7.2 26.8 26.9

27

27.1 27.2 27.3 27.4 27.5

Skyler Markham, Debbie C. Crans, and Bruce Atwater Introduction  559 The First Cross-coupling Reaction  559 Phosphine Ligands for Catalysts of Cross-Coupling Reactions  560 Speciation  564 Palladium Nanoparticle Catalysts  564 Speciation of Palladium (Pd) Catalysts  567 Alternative Metal Catalysts  570 Nickel  570 Cobalt  570 Speciation of Nickel and Cobalt Catalysts  572 Cross-coupling Reactions and Sustainability: Summary and the Future  574 References  574

Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions  577

Ana P. Carvalho, Angela Martins, Filomena Martins, Nelson Nunes, and Rúben Elvas-Leitão Introduction  577 Zeolites and Hierarchical Zeolites  580 Zeolites and Hierarchical Zeolites as Catalysts for Friedel Crafts Acylation Reactions  585 Understanding Friedel-Crafts Acylation Reactions through Quantitative Structure-property Relationships  597 Final Remarks  604 Acknowledgements  605 References  605 VOLUME 3 About the Editors  xiii Preface  xv Part V

28

Organocatalysis  609

Sustainable Drug Substance Processes Enabled by Catalysis: Case Studies from the Roche Pipeline  611

Kurt Püntener, Stefan Hildbrand, Helmut Stahr, Andreas Schuster, Hans Iding, and Stephan Bachmann

29

Supported Chiral Organocatalysts for Accessing Fine Chemicals  639 Ana C. Amorim and Anthony J. Burke

xi

xii

Contents

30

Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis  659 Kotohiro Nomura and Nor Wahida Binti Awang

31

Modern Strategies for Electron Injection by Means of Organic Photocatalysts: Beyond Metallic Reagents  675

Takashi Koike

32

Visible Light as an Alternative Energy Source in Enantioselective Catalysis  687 Ana Maria Faisca Phillips and Armando J.L. Pombeiro

Part VI Catalysis for the Purification of Water and Liquid Fuels  717 33

Heterogeneous Photocatalysis for Wastewater Treatment: A Major Step Towards Environmental Sustainability  719

Shima Rahim Pouran and Aziz Habibi-Yangjeh

34

Sustainable Homogeneous Catalytic Oxidative Processes for the Desulfurization of Fuels  743

Federica Sabuzi, Giuseppe Pomarico, Pierluca Galloni, and Valeria Conte

35

Heterogeneous Catalytic Desulfurization of Liquid Fuels: The Present and the Future  757

Rui G. Faria, Alexandre Viana, Carlos M. Granadeiro, Luís Cunha-Silva, and Salete S. Balula

Part VII Hydrogen Formation, Storage, and Utilization  783 36

Paraformaldehyde: Opportunities as a C1-Building Block and H2 Source for Sustainable Organic Synthesis  785

Ana Maria Faísca Phillips, Maximilian N. Kopylovich, Leandro Helgueira de Andrade, and Martin H.G. Prechtl

37

Hydrogen Storage and Recovery with the Use of Chemical Batteries  819 Henrietta Horváth, Gábor Papp, Ágnes Kathó, and Ferenc Joó

38

Low-cost Co and Ni MOFs/CPs as Electrocatalysts for Water Splitting Toward Clean Energy-Technology  847

Anup Paul, Biljana Šljukić, and Armando J.L. Pombeiro



Index  871

xiii

About the Editors Armando Pombeiro is a Full Professor Jubilado at Instituto Superior Técnico, Universidade de Lisboa (ULisboa), former Distant Director at the People’s Friendship University of Russia (RUDN University), a Full Member of the Academy of Sciences of Lisbon (ASL), the President of the Scientific Council of the ASL, a Fellow of the European Academy of Sciences (EURASC), a Member of the Academia Europaea, founding President of the College of Chemistry of ULisboa, a former Coordinator of the Centro de Química Estrutural at ULisboa, Coordinator of the Coordination Chemistry and Catalysis group at ULisboa, and the founding Director of the doctoral Program in Catalysis and Sustainability at ULisboa. He has chaired major international conferences. His research addresses the activation of small molecules with industrial, environmental, or biological significance (including alkane functionalization, oxidation catalysis, and catalysis in unconventional conditions) as well as crystal engineering of coordination compounds, polynuclear and supramolecular structures (including MOFs), non-covalent interactions in synthesis, coordination compounds with bioactivity, molecular electrochemistry, and theoretical studies. He has authored or edited 10 books, (co-)authored ca. 1000 research publications, and registered ca. 40 patents. His work received over. 30,000 citations (over 12,000 citing articles), h-index ca. 80 (Web of Science). Among his honors, he was awarded an Honorary Professorship by St. Petersburg State University (Institute of Chemistry), an Invited Chair Professorship by National Taiwan University of Science & Technology, the inaugural SCF French-Portuguese Prize by the French Chemical Society, the Madinabeitia-Lourenço Prize by the Spanish Royal Chemical Society, and the Prizes of the Portuguese Chemical and Electrochemical Societies, the Scientific Prizes of ULisboa and Technical ULisboa, and the Vanadis Prize. Special issues of Coordination Chemistry Reviews and the Journal of Organometallic Chemistry were published in his honor. https://fenix.tecnico.ulisboa.pt/homepage/ist10897

xiv

About the Editors

Manas Sutradhar is an Assistant Professor at the Universidade Lusófona, Lisbon, Portugal and an integrated member at the Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa, Portugal. He was a post‐doctoral fellow at the Institute of Inorganic and Analytical Chemistry of Johannes Gutenberg University of Mainz, Germany and a researcher at the Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa. He has published 72 papers in international peer review journals (including three reviews + 1 reference module), giving him an h-index 28 (ISI Web of Knowledge) and more than 2250 citations. In addition, he has 11 book chapters in books with international circulation and one patent. He is one of the editors of the book Vanadium Catalysis, published by the Royal Society of Chemistry. His main areas of work include metal complexes with aroylhydrazones, oxidation catalysis of industrial importance and sustainable environmental significance, magnetic properties of metal complexes, and bio-active molecules. The major contributions of his research work are in the areas of vanadium chemistry and oxidation catalysis. He received the 2006 Young Scientist Award from the Indian Chemical Society, India and the Sir P. C. Ray Research Award (2006) from the University of Calcutta, India. https://orcid.org/0000-0003-3349-9154 Elisabete C.B.A. Alegria is an Adjunct Professor at the Chemical Engineering Department of the Instituto Superior de Engenharia de Lisboa (ISEL) of the Polytechnic Institute of Lisbon, Portugal. She is a researcher (Core Member) at the Centro de Química Estrutural (Coordination Chemistry and Catalysis Group). She has authored 86 papers in international peered review journals and has an h-index of 23 with over 1600 citations, four patents, five book chapters, and over 180 presentations at national and international scientific meetings. She was awarded an Honorary Distinction (2017–2020) for the Areas of Technology and Engineering (Scientific Prize IPL-CGD). She is an editorial board member, and has acted as a guest editor and reviewer for several scientific journals. Her main research interests include coordination and sustainable chemistry, homogeneous and supported catalysis, stimuli‐responsive catalytic systems, green synthesis of metallic nanoparticles for catalysis, and biomedical applications. She is also interested in mechanochemistry (synthesis and catalysis) and molecular electrochemistry. https://orcid.org/0000-0003-4060-1057

xv

Preface Aiming to change the world for the better, 17 Sustainable Development Goals (SDGs) were adopted by the United Nations (UN) Member States in 2015, as part of the UN 2030 Agenda for Sustainable Development that concerns social, economic, and environmental sustainability. Hence, a 15-year plan was set up to achieve these Goals and it is already into its second half. However, the world does not seem to be on a good track to reach those aims as it is immersed in the Covid-19 pandemic crisis and climate emergency, as well as economic and political uncertainties. Enormous efforts must be pursued to overcome these obstacles and chemical sciences should play a pivotal role. Catalysis is of particular importance as it constitutes the most relevant contribution of chemistry towards sustainable development. This is true even though the SDGs are integrated and action in one can affect others. For example, the importance of chemistry and particularly catalysis is evident in several SDGs. Goal 12, addresses “Responsible Consumption and Production Patterns” and is aligned with the circularity concept with sustainable loops or cycles (e.g., in recycle and reuse processes that are relevant within the UN Environmental Program). Goal 7 addresses “Affordable and Clean Energy” and relates to efforts to improve energy conversion processes, such as hydrogen evolution and oxygen evolution from water, that have a high environmental impact. Other SDGs in which chemistry and catalysis play an evident role with environmental significance include Goal 6 (“Clean Water and Sanitation”), Goal 9 (“Industry, Innovation and Infrastructure” 13 (“Climate Action”), Goal 14 (“Life Below Water”), and Goal 15 (“Life on Land”). The book is aligned with these SDGs by covering recent developments in various catalytic processes that are designed for a sustainable environment. It gathers skilful researchers from around the world to address the use of catalysis in various approaches, including homogeneous, supported, and heterogeneous catalyses as well as photo- and electrocatalysis by searching for innovative green chemistry routes from a sustainable environmental angle. It illustrates, in an authoritative way, state-of-the-art knowledge in relevant areas, presented from modern perspectives and viewpoints topics in coordination, inorganic, organic, organometallic, bioinorganic, pharmacological, and analytical chemistries as well as chemical engineering and materials science. The chapters are spread over seven main sections focused on Carbon Dioxide Utilization, Transformation of Volatile Organic Compound (VOCs), Carbon-based Catalysts, Coordination, Inorganic, and Bioinspired Catalysis, Organocatalysis, Catalysis for the Purification of Water and Liquid Fuels,and Hydrogen Formation, Storage, and Utilization. These sections are gathered together as a contribution towards the development of the challenging topic.

xvi

Preface

The book addresses topics in (i) activation of relevant small molecules with strong environmental impacts, (ii) catalytic synthesis of important added value organic compounds, and (iii) development of systems operating under environmentally benign and mild conditions toward the establishment of sustainable energy processes. This work is expected to be a reference for academic and research staff of universities and research institutions, including industrial laboratories. It is also addressed to post-doctoral, postgraduate, and undergraduate students (in the latter case as a supplemental text) working in chemical, chemical engineering, and related sciences. It should also provide inspiration for research topics for PhD and MSc theses, projects, and research lines, in addition to acting as an encouragement for the development of the overall field. The topic Catalysis for Sustainable Environment is very relevant in the context of modern research and is often implicit, although in a non-systematic and disconnected way, in many publications and in a number of initiatives such as international conferences. These include the XXII International Symposium on Homogeneous Catalysis (ISHC) that we organized (Lisbon, 2022) and that to some extent inspired some parts of this book. In contrast to the usual random inclusion of the topic in the literature and scientific events, the applications of catalytic reactions focused on a sustainable environment in a diversity of approaches are addressed in this book. The topic has also contributed to the significance of work that led to recent Nobel Prizes of Chemistry. In 2022, the Nobel Prize was awarded to Barry Sharpless, Morten Meldal, and Carolyn Bertozzi for the development of click chemistry and bioorthogonal chemistry. The set of criteria for a reaction or a process to meet in the context of click chemistry includes, among others, the operation under benign conditions such as those that are environmentally friendly (e.g., preferably under air and in water medium). In 2021, the Nobel Prize was awarded to Benjamin List and David W.C. MacMillan for the development of asymmetric organocatalysis, which relies on environmentally friendly organocatalysts. The book illustrates the connections of catalysis with a sustainable environment, as well as the richness and potential of modern catalysis and its relationships with other sciences (thus fostering interdisciplinarity) in pursuit of sustainability. At last, but not least, we should acknowledge the authors of the chapters for their relevant contributions, prepared during a particularly difficult pandemic period, as well as the publisher, John Wiley, for the support, patience, and understanding of the difficulties caused by the adverse circumstances we are experiencing nowadays and that constituted a high activation energy barrier that had to be overcome by all of us… a task that required rather active catalysts. We hope the readers will enjoy reading its chapters as much as we enjoyed editing this book. Armando Pombeiro Manas Sutradhar Elisabete Alegria

299

Part IV Coordination, Inorganic, and Bioinspired Catalysis

301

14 Hydroformylation Catalysts for the Synthesis of Fine Chemicals Mariette M. Pereira1, Rui M.B. Carrilho1, Fábio M.S. Rodrigues1, Lucas D. Dias2, and Mário J.F. Calvete1 1 2

Coimbra Chemistry Centre, Department of Chemistry, University of Coimbra, Rua Larga Coimbra, Portugal Laboratório de Novos Materiais, Universidade Evangélica de Goiás, Anápolis GO, Brazil

14.1  Introduction The hydroformylation reaction consists of the addition of CO and H2, also known as synthesis gas or syngas, across the π system of a C=C double bond, in the presence of a catalyst, to form aldehydes (Figure 14.1). As a pure addition reaction, in which all raw materials are incorporated in the aldehyde products, the catalytic hydroformylation is a paradigmatic example of a 100% atom economy chemical process [1–5]. The reaction was discovered accidentally by Otto Roellen (1897–1993) in 1934, when he worked in the German Ruhrchemie industry, aiming the optimization of the Fischer-Tropsch reaction for the preparation of fuels [6, 7]. During an experiment to prepare ethylene, Roelen surprisingly found that, in the presence of ammonia, the imine of propionaldehyde was obtained as a white solid when a mixture of cobalt, thorium, and magnesium oxides were used as catalysts [4]. In the following years, he investigated the potentialities and limitations of the new reaction and discovered that, indeed, homogeneous cobalt salts were the catalytic species responsible for the synthesis of aldehydes from olefins and syngas. Roelen recognized the high synthetic and economic potential of the oxo-synthesis and, in 1938, he submitted a patent application [6]. Soon after its discovery, the scientific community quickly found that, to turn the hydroformylation reaction into a viable synthetic alternative for the industrial production of fine chemicals, the simultaneous control of the catalyst activity and chemo-, regio-, and enantioselectivity or diastereoselectivity would be the greatest challenges to achieve (Figure 14.2). The first challenging goal was to obtain high chemoselectivity for the desired aldehydes. This was almost achieved through replacement of cobalt-carbonyl catalysts by rhodium-carbonyl complexes coordinated with phosphorus ligands. With respect to the regio-, enantio-, and diastereoselectivity issues, scientists and industrials have been dedicated to the design and synthesis of elaborate phosphorus ligands over the years, which led to the development of both homogenous and immobilized rhodium catalysts with high activity and selectivity, allowing the implementation of hydroformylation in fine chemical industries [8]. Since then, the hydroformylation has turned into one of the largest scale industrial homogeneous catalytic processes for production of Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 2, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

302

14  Hydroformylation Catalysts for the Synthesis of Fine Chemicals

Figure 14.1  General scheme of the hydroformylation reaction.

Figure 14.2  Control of chemo-, regio-, and enantioselectivity in catalytic olefin hydroformylation.

aldehydes and their derivatives (nearly 10 million tons/year) [4]. Furthermore, the aldehydes formed are key intermediates for the preparation of added-value fine chemical products such as alcohols, amines, acetals, and carboxylic acid derivatives (Figure 14.3). Thus, the catalytic hydroformylation is currently considered a central reaction for sequential synthetic processes to transform aldehydes into other functional groups under sustainable reaction conditions [9–11]. In this chapter, we present the most paradigmatic examples of homogeneous and immobilized hydroformylation catalysts including a brief description of their use in hydroformylation reactions and hydroformylation centered sequential reactions, particularly for the synthesis of fine chemicals. A special emphasis will be given to the contribution of Portuguese researchers to the development of this topic.

Figure 14.3  Examples of hydroformylation-based sequential reactions for fine-chemical synthesis.

14.2  Homogeneous Catalytic Systems

14.2  Homogeneous Catalytic Systems 14.2.1  Development of Phosphorus Ligands As mentioned, rhodium complexes are considered the catalysts of choice for hydroformylation reactions. To increase their activity and selectivity, over the years, a plethora of phosphorus ligands have been developed. Wilkinson et al. [12–14] reported the use of rhodium complexes coordinated with triphenylphosphine 1, providing the first highly chemoselective systems. Soon after Wilkinson’s work, Pruett and Smith [15, 16] introduced the use of rhodium phosphite complexes (Figure 14.4). In the 80s, van Leeuwen et al. [17–19] discovered a peculiar effect of bulky aryl monophosphites, such as tris(o-tertbutylphenyl)phosphite 2 (Figure 14.4), with a large cone angle (c. 175°), whose rhodium complexes led to highly active chemo- and regioselective catalysts in the hydroformylation of substituted olefins under mild reaction conditions. Regarding the development of highly active rhodium-phosphite catalysts, Pereira et al. and Bayón et al. designed and prepared a family of chiral bulky monophosphites with C3-symmetry [20–22]. Their synthesis involved the mono-etherification of (R) or (S)-BINOL with primary, secondary or tertiary alcohols, via Mitsunobu reaction, followed by reaction of the BINOL monoethers with 1/3 equivalent of PCl3 in triethylamine, used simultaneously as base and reaction solvent (Figure 14.5).

Figure 14.4  Some of the first phosphorus ligands used in Rh-catalyzed hydroformylation.

Figure 14.5  Synthesis of tris-BINOL based monophosphites.

303

304

14  Hydroformylation Catalysts for the Synthesis of Fine Chemicals

The cone angles of 4a-d were calculated by semi-empirical computational methods, using the PM6 Hamiltonian, and the values obtained were within the range of 240 – 270º. In addition, their rhodium complexes, formed in situ, showed remarkable catalytic activity in the hydroformylation of disubstituted styrene derivatives under relatively mild reaction conditions (80 °C, 30 bar CO/H2), achieving 100% chemoselectivity for aldehydes and regioselectivities between 90–99%; however, low enantioselectivity was obtained (up to 20% ee). More recently [23], the family of ­tris-binaphthyl phosphites was expanded, with the preparation of four stereoisomeric monophosphites 4e, based on axially chiral (R)- or (S)-BINOL bearing a chiral (+)- or (–)-neomenthyloxy group. These new tris-binaphthyl phosphite ligands were characterized by DFT computational methods, which allowed calculatation of an extremely large cone angle of 345°. Furthermore, trisBINOL neomenthol mononophosphites were applied in rhodium-catalyzed hydroformylation of styrene, leading to complete conversions in 4h with 100% chemoselectivity for aldehydes and up to 98% iso-regioselectivity. The Rh(I)/phosphite catalytic system was also highly active and selective in the hydroformylation of disubstituted olefins, including (E)-prop-1-en-1-ylbenzene and prop1-en-2-ylbenzene, yielding the corresponding branched aldehyde in 80% regioselectivity and the linear aldehyde in 99% regioselectivity, respectively. The effect of the phosphine ligands bite angles on the reaction’s regioselectivity was reported by several research groups [24], including van Leeuven et al., who reported the use of Xantphos ligand 3, with a bite angle of 111°, which led to exclusive formation of linear aldehydes from alkyl terminal olefins [25–27]. The most remarkable enantioselectivities have been accomplished using rhodium complexes coordinated with bidentate phosphorus ligands. Babin and Whiteker described the use of bulky diphosphites, derived from homochiral (2R,4R)-pentane-2,4-diol, including Chiraphite 5, in the Rh-catalyzed hydroformylation of styrene (Figure 14.6) [28]. This led to 99% regioselectivity for the branched aldehyde and up to 90% ee, which was attributed to formation of eight-membered chelates with rhodium(I), through a bis-equatorial coordination mode. Another family of chiral diphosphite ligands, based on carbohydrate backbones, was developed by Claver et al. [29, 30]. Among them, the rhodium(I) complex of ligand 6, containing a three-carbon bridge, achieved 93% ee, along with 98% of iso-regioselectivity, in styrene hydroformylation (Figure 14.6). Klosin et al. reported the use of bis-diazaphospholane ligand 7 in the asymmetric hydroformylation of styrene, vinyl acetate and allyl cyanide, achieving for these substrates with 89%, 96%, and 87% ee, respectively (Figure 14.6) [31, 32]. The major breakthrough on enantioselectivity came from Takaya’s laboratory with the synthesis of (R,S)-BINAPHOS 8 [33, 34], which led to up to 95% ee with 91% regioselectivity toward the branched product in the hydroformylation of styrene, and more than 90% ee for a range of other functionalized substrates, including internal alkenes [35–37]. HP-NMR studies showed that the C1symmetric ligand 8 coordinate to Rh(I) in equatorial-axial fashion, in which the more σ-donor phosphine P atom sits in the plane with the CO ligands, while the more π-accepting phosphite P atom binds apical to the hydride (complex 9). This unique dissymmetric environment in a single catalytically active species was found to be a decisive factor for the high enantiodiscrimination. Later, Zhang et al. reported the synthesis of the mixed phosphine-phosphoramidite ligand (R,S)-YANPHOS 10 [38], derived from 2-amino-2ʹ-hydroxy-1,1ʹ-binaphthyl (NOBIN). The N-alkyl substituent was intended to supply a rigid conformation able to provide a more closed chiral pocket than the BINAPHOS complex and, as expected, it led to a significant improvement on asymmetric induction. Indeed, the Rh/(R,S)-YANPHOS complex provided up to 99% ee along with 90% iso-regioselectivity, in the Rh-catalyzed hydroformylation of styrene and a number of other olefins [39]. Freixa and Bayón reported a BINOL-based diphosphite ligand 11 (Figure 14.8), linked through an isophthalate bridge, forming a large chelating ring [40]. Their rhodium complexes provided

14.2  Homogeneous Catalytic Systems

Figure 14.6  Examples of bidentate phosphites and diazaphospholane ligands.

Figure 14.7  Mixed phosphine-phosphite (8) and phosphine-phosphoramidite ligands (10).

active and regioselective catalysts in the hydroformylation of vinyl arenes and the appropriate combination of stereogenic centers provided by chiral binaphthyl and propan-1,2-diol backbones allowed to achieve up to 76% enantiomeric excess. Following these studies, Pereira et al. [41] developed a family of related ditopic (R)-BINOL-based phosphites 12–14, containing pyridine bridges linking two binaphthyl fragments (Figure 14.8). The same authors developed the synthesis of diphosphites 15, containing alkyl ether spacers linking the two binaphthyl fragments. These were synthesized by reaction of (1R,1ʹ’R)-2ʹ,2ʹ’-[propane-1,3-diyl-bis(oxy)]di-1,1ʹ-binaphthyl-2-ol or

305

306

14  Hydroformylation Catalysts for the Synthesis of Fine Chemicals

Figure 14.8  Examples of large chelating ring diphosphites developed by Bayón et al. (11) and Pereira et al. (12–14).

(1R,1ʹ’R)-2ʹ,2ʹ’-[2,2-dimethylpropane-1,3-diyl-bis(oxy)]di-1,1ʹ-binaphthyl-2-ol, with pyrocathecol phosphochloridite, using triethylamine as base (Figure 14.9). Their rhodium complexes achieved up to 95% regioselectivity for the formation of branched aldehydes, but enantiomeric excesses were only ca. 10% in all cases. For the Rh/14 catalytic system, the authors further studied the effect of different reaction variables (pressure, temperature, and the addition of a lithium salt) on the regioselectivity by 23 factorial design computational methods, which revealed that temperature was the most decisive reaction parameter[41].

Figure 14.9  Synthesis of bis-BINOL-pyrocathecol disphosphite ligands.

14.2  Homogeneous Catalytic Systems

14.2.2  Hydroformylation of Biologically Relevant Substrates Steroid and terpene molecules are important classes of biologically relevant compounds that are extensively applied in the synthesis of fine chemicals in medicinal chemistry [42]. However, the hydroformylation of these substrates is usually very difficult to perform because they contain highly substituted and internal double C=C bonds, which are less reactive due to steric hindrance that prevents the approach of the metal catalyst. The only exceptions are the use of Rh/bulky aryl phosphite catalytic systems, whose exceptional activity results from both electronic and stereo effects. On one hand, the π-acidic character of the phosphite weakens the metal-CO bond, thereby allowing a faster CO dissociation; on the other hand, the phosphite ligand’s large cone angle allows the coordination of only one phosphite to the metal center, even when used in large excess, which results in a low global steric hindrance around the metal center [19]. Therefore, Rh-phosphite catalysts have opened new perspectives for application of hydroformylation as key tool for functionalization of natural products such as steroids and terpene molecules. In this regard, the use of Rh/tris(o-tert-butylphenyl)phosphite catalyst (Rh(I)/2a) in the hydroformylation of highly substituted steroid olefins was also reported, namely for 17β-acetoxyandrost4-ene and 3β,17β-diacetoxyandrost-4-ene (Figure 14.10). In both cases, the major reaction product (68%) was 4β-formyl-17β-acetoxy-5β-androstane. The authors concluded that the hydroformylation reaction occurred preferentially at the β face of the steroid nucleus. This reaction was the first example of a catalytic carbonylation to the β face of a steroid backbone [43]. The studies were expanded further, to the hydroformylation of cholest-4-ene and 3β-acetoxycholest4-ene, using the same catalytic system. Under the reaction conditions assayed, these steroids were hydroformylated, producing the corresponding 4-formyl derivatives with 100% regioselectivity and up to 70% β-diastereoselectivity. Both formyl steroids were isolated in their acetal forms [44]. More recently, the application of Rh(I)/tris-binaphthyl monophosphite catalyst Rh(I)/(R)-4b was also reported in the hydroformylation of 17β-acetoxyandrost-4-ene, leading to 95% conversion in 48 h, with 86% chemoselectivity for aldehydes and 70% diastereoselectivity for the β-formyl aldehyde [45]. Terpenes are another relevant example of natural molecules that have been extensively used as substrate for hydroformylation reactions aiming the preparation of value-added fine chemicals [46, 47]. In this context, Pereira et al. reported the hydroformylation of monoterpenes, such as (1R)-(−)-myrtenyl acetate using and (−)-isopulegol benzyl ether, using highly active Rh(I)/2a or Rh(I)/4b bulky aryl monophosphite catalytic systems (Figure 14.11). The hydroformylation of diterpenes with exocyclic methylenic double bonds was also reported [48], namely the methyl esters of kaurenic and grandiflorenic acids, and the trimethylsilylether of kaurenol. In this reactions, the authors used Rh/triphenulphosphine (1) and Rh/tris-(o-

Figure 14.10  Hydroformylation of steroid derivatives using Rh(I)/bulky aryl monophosphite catalytic systems.

307

308

14  Hydroformylation Catalysts for the Synthesis of Fine Chemicals

Figure 14.11  Hydroformylation of monoterpenes: a) (1R)-(−)-myrtenyl acetate; and b) (−)-isopulegol benzyl ether.

Figure 14.12  Hydroformylation of kaurenic acid methyl ester and kaurenol trimethylsilylether.

t-butylphenyl)phosphite (2a) as catalysts, with P(CO:H2) = 20 bar and T = 100 °C (Figure 14.12). High yields, chemo- and regioselectivity toward the formation of the linear aldehydes were achieved with both catalytic systems. The major isolated linear aldehyde diastereoisomer resulted from the coordination of rhodium and consequent introduction of CO/H2 in the less hindered face of the exocyclic double bond. Additionally, the authors observed that the isomerization to the internal ring position is a competitive process, which is strongly dependent of the catalyst used (% of isomerization: PPh3-modified system 99% yield after six hours at 80 °C. The proposed reaction mechanism starts with the insertion of CO2 into the Mn-H bond, then RSiH3 attacks the coordinated O atom, followed by a hydrogen transfer, which results in the regeneration of the initial pincer hydride complex. Next, the hydride attacks the carbonyl carbon to form a metal-coordinated silyl hemiacetal, which reacts with a second molecule of silane, forming a silyl ether and a methoxide-Mn complex. Finally, the methoxide is released in the form of methoxy-silane, and the catalyst is regenerated by reacting with another silane molecule.

Scheme 18.7  CO2 reduction with boranes catalyzed by pincer complexes.

18.2  Hydrogenation of CO2 to Methanol and Related Processes

Analogously, the reduction of CO2 to methanol can be also performed using boranes instead of silanes (Scheme 18.7) [111]. POCOP-Ni [112–117] and -Pd [21, 118] pincer complexes have been used to catalyze such transformations. When the POCOP-Pd pincer complex was employed, the reduction of CO2 to methanol was achieved using 1 atm of CO2, 500 equivalents of catecholborane (HBcat), and benzene at room temperature. Under these conditions, a TOF of 1780 h–1 was observed [118]. Under similar conditions, the POCOP-Ni complex achieved a TOF of 495 h–1. The different substituents at the phosphorous atom influence the catalytic activity of the POCOP-Ni complex. The bulkier the substituent is, the more activity the catalyst shows. For example, the tBu substituent gave a TON of 100 in 45 minutes, whereas the iPr derivative needs two hours to reach a similar TON, and the less active cyclopentyl group gave a TON of 30 in 12 hours [114, 115]. Additionally, the coordination of a thiolate ligand to the Ni-POCOP complex produces a more active species, reaching a TOF of 2,400 h–1 [116]. The first example of a Mn-pincer complex (Mn2) for the reduction of CO2 to methoxy boronate was described by Leitner et al. in 2018 [119]. The reaction was carried out in the absence of solvent and using a catalytic amounts of base (NaOtBu). In this case, the borane source was pinacolborane (HBpin). After 14 hours at 100 °C, the reaction achieved a TON of 883. Very recently, two research groups have independently studied the reaction mechanism using theoretical DFT calculations. The reaction consists in three essential stages: (i) hydroboration of CO2 to HCOOBpin, (ii) hydroboration of HCOOBpin to HCHO, and (iii) hydroboration of HCHO to CH3OBpin. Lei and Cao proposed that, after the deprotonation of the pincer ligand, HBpin reacts with the complex, forming an Mn-H and N-B species, then the boron atom interacts with one oxygen atom of the CO2, favoring the approach of the carbonyl carbon atom to the hydride, and consequently the reduction of the carbonyl. The energy span of this proposal is 27.0 kcal/mol [120]. In contrast, Schaefer III et al. found that the boron is abstracted by the base (NaOtBu), forming an N-Na interaction and tBuOBpin species. The energy span of this proposal is 22.5 kcal/mol [121]. Thus, there is difference of 4.5 kcal/mol between the two proposals and it is possible that both mechanisms may be operating at the same time. By changing the pincer by a PNNNP ligand and adding a co-catalyst (B(OPh)3) as a Lewis acid, the formation of the boryl-protected methanol was achieved in high yields [122]. In addition, the use of CO as feedstock is also attractive. In principle, the hydrogenation of CO to methanol is more thermodynamically favored than the hydrogenation of CO2 to methanol (ΔH298K = –90.7 kJ/mol vs –49.5 kJ/mol), so the former should be carried out under softer reaction conditions. However, CO shows a high affinity to ­several transitions metals, forming very stable compounds that sometimes result in inactive species on some catalytic cycles. This is particularly problematic in the case of homogeneous catalysts. Nevertheless, there are reports that describe the transformation of CO to methanol (Scheme 18.8) using metal pincer complexes [123–125] and other transition metal-based catalysts [126]. MACHO-Ru-BH type complexes are able to fully convert CO into methanol in the presence of a base and dialkylamine in a sequential stepwise reaction [123]. Analogously to CO2, CO can be captured in the presence of an amine and a base, forming an amide that is subsequently hydrogenated by the catalyst. When the reaction is carried out using 10 bar CO, 70 bar H2 at 145 °C, for 168 hours, methanol is obtained in 59% yield, giving a TON of 539. Using a MACHO-Mn catalyst and indoles as nitrogen promoters, the formation of Scheme 18.8  Reduction of CO to methanol methanol was achieved in TON’s up to 3170 [124]. catalyzed by pincer metal complexes.

397

398

18  Catalysis by Pincer Compounds and Their Contribution to Environmental and Sustainable Processes

It  is worth mentioning that the reaction was performed using lower hydrogen pressures (40 bar). Some alcohols have also promoted the conversion of CO to methanol [125]. In this case, the MACHO-Mn catalyst gives a TON of 4,023, using 5 bar CO, 50 bar H2, and ethanol, at 150 °C for 12 hours. The reaction mechanism was also studied, and follows a typical ligand-assisted mechanism in which hydrogen is first activated by the presence of the N in the pincer and the metal center, forming a hydride species. Then the ester or the amide is hydrogenated, generating formaldehyde, which is hydrogenated by a hydride-Mn species. All of these achievements represent a great breakthrough if we consider that methanol synthesis from syngas (CO/H2 mixtures) requires high temperatures (200–300 °C) and pressures (50–100 atm). Thus, MACHO pincer complexes have helped to develop procedures that require less energy consumption and that are consequently considered greener. Following a similar approach, the MACHO-Mn catalyst was also active in the β-methylation of alcohols using 5 bar CO, 15 bar H2, and a two-fold excess of base (NaOtBu) [127]. This interesting reactivity can be explained by the fact that the formed formaldehyde from the hydrogenation of CO can react with the aldehyde formed by the dehydrogenation of the alcohol under these reaction conditions, in a typical aldol condensation under basic conditions. The scope of the reaction was wide, including functionalities such as aromatic rings, heterocycles, and alkenes. The synthesis of ethylene glycol from CO/H2 catalyzed by pincer complexes has also been explored in recent years (Scheme 18.9) [128–130]. The reaction proceeds in two stages. (i) The CO is captured in the form of a dione. Dione formation is catalyzed by a Pd catalyst ([Pd(acac)2] or Pd/C) and consists of the reaction of CO with piperidine (forming an oxamide), ethanol (forming an oxalate), or piperidine/ethanol mixture (forming an oxamate). This first stage was inspired by previous work regarding oxidative couplings of CO and amines or alcohols [131, 132]. (ii) The second stage consists of the hydrogenation reaction of the dione using a pincer complex. Hence, the Ru-MACHO-BH complex was able to convert an oxamide (1,2-di(piperidin-1-yl)ethane-1,2-dione) to ethylene glycol in 96%, using a 60 bar H2, KOtBu (catalytic) in toluene at 160 °C for 12 hours [128]. Milstein’s catalyst (NNP-Ru pincer ­complex) showed good activity toward the hydrogenation of oxamate (ethyl 2-oxo-2-(piperidin-1-yl) acetate), reaching 92% yield under optimized conditions (toluene, KOtBu, 60 bar H2, 160 °C, 10 hours) [129]. Additionally,Milstein’s catalyst was also active toward the hydrogenation of oxalate (diethyl oxalate). In this case, the reaction required a lower temperature (100 °C) and hydrogen pressure (40 bar) to afford >90% yield [130].

Scheme 18.9  Synthesis of ethylene glycol using CO as feedstock.

18.3  Hydrogen Production from Methanol and Water

18.3  Hydrogen Production from Methanol and Water Methanol reforming consists of the production of molecular H2 and CO2 from methanol. This reaction is achieved in three successive dehydrogenation reactions, starting with the dehydrogenation of methanol to formaldehyde, which reacts with water, to further undergo another dehydrogenation, forming formate. The latter is then dehydrogenated to produce CO2. In each dehydrogenation step, a molecule of hydrogen is produced, hence each molecule of methanol has the potential to form three molecules of hydrogen (Scheme 18.10). At present, this process is of particular interest due the high content of hydrogen in methanol, allowing the transport of hydrogen in a safer way than using the traditional available methods that involves cryogenic temperatures, high pressure, and inefficient tanks to contain hydrogen. In fact, the content of hydrogen in methanol is 12.6%, which is adequate for some mobile applications. Furthermore, methanol is a liquid that can be easily transported by pipelines, facilitating its distribution. Therefore, the simple idea of transporting chemical hydrogen in the form of methanol is highly attractive to extend the use of hydrogenbased technologies.

Scheme 18.10  Dehydrogenation reaction of methanol to CO2 and H2.

The methanol dehydrogenation reaction must be very selective to hydrogen and CO2 to feed a hydrogen fuel cell because the presence of small contaminants such as CO or NH3, poisons the catalyst, reducing the production of energy that the fuel cell is able to achieve. This reaction has been heterogeneously catalyzed and high temperatures are required (>200 °C) [133, 134]. Pincer catalysts have shown a remarkable activity toward methanol reforming processes at low temperatures. In 2013, Beller’s group described major breakthroughs of aqueous-methanol reforming ­catalyzed by pincer complexes [135–137]. Ru-MACHO pincer complexes catalyzed the production of hydrogen from methanol at 72 °C, reaching a TOF of 124 h–1 in one hour. The phosphorous substituents play an important role in the activity of the catalyst, being more active with aromatic substituents; TOF = 124 h–1 for PPh2 vs 45 h–1 for PiPr2. Remarkably, the production of CO was very low ranging on the ppm scale ( 900 h–1, TON = 2,743). The reaction also proceeds in the absence of base, reaching a TON of 59, and gas evolution of 61 mLh–1. Interestingly, a positive synergistic effect was observed by the addition of an extra Ru-based catalyst in the absence of base. When the reaction was carried out in the presence of [Ru(H)2(dppm)2], the production of hydrogen gas reached was 239 h–1 in the first hour, but then a deacceleration of the reaction was observed, decreasing the TOF to 138 h–1 after three hours, and 83 h–1 after seven hours. A more stable hydrogen evolution was achieved by using a Ru-based bi-catalyst with a bigger bite angle. Using [Ru(H)2(dppm)2], a TOF of ~90 h–1 was observed (TOF = 87 h–1 [one hour], 94 h–1 [three hours], 93 h–1 [seven hours]).

399

400

18  Catalysis by Pincer Compounds and Their Contribution to Environmental and Sustainable Processes

Because of the high activity of Ru-MACHO pincer complexes, they have been immobilized using the supported liquid phase (SLP) concept [138, 139]. This approach consists in dispersing a thin film of liquid containing the catalyst onto a large inner surface area, allowing a continuous vapor-phase process that ideally can be directly used in an hydrogen fuel cell. Hence, Ru-MACHO ­catalysts together with KOH were immobilized in two ionic liquids and dispersed onto alumina. Charging 2 wt% of Ru-MACHO in 10 wt% of KOH on porous alumina the activity was 14 molH2molRu–1h–1 at 150 °C and 1 bar and this catalytic system maintained its activity over 70 h time on stream with negligible deactivation. Further improvements of this catalytic system were later described. The use of CsOH instead of KOH and the presence of a second Ru-MACHO-Me catalyst enhanced the catalytic activity, reaching a TOF of 91 h–1. Furthermore, the reaction mechanism employing Ru-MACHO and Ru-MACHO-Me catalyst was studied in detail [140–142] (Scheme 18.11). The Ru-MACHO starts reacting with the base, forming a Ru-amido species that is the “real” catalyst. Next, an HRO− compound (R = CH2, CHOH, CO) is coordinated to the Ru-amido species. This ligand is first coordinated by the oxygen atom, and subsequently is shifted to the hydrogen atom. According to DFT calculations focusing on the frontier orbital interaction between the Ru-amido species and the methoxide, gem-diolate, or formate, coordination through O or H atom is possible and exists in a dynamic equilibrium. The coordination of the H atom favors the formation of the Ru-hydride species and the formation of the RO− compound, which can further react with −OH to capture formaldehyde, formic acid, or CO2 by-products. The high concentration of base promotes this step and drives the reaction forward. Because of the basic character of the N atom in the MACHO ligand, it can deprotonate a RHOH species, forming a neutral pincer ligand. Next, methanol promotes the formation of dihydrogen and the regeneration of the catalyst. Other Ru-based pincer catalysts have been also used for the methanol reforming reaction [143, 144]. For instance, Milstein’s catalyst is active in the hydrogen production from methanol and water [143]. In this case, the reaction was carried out in the biphasic system of methanol-toluenewater, using 0.025 mol% of catalyst and a base (NaOH) at 100–105 °C. Under these conditions, a 77 % yield was obtained after 9 days. Interestingly, the organic phase was reused two more times, reaching similar yields (82% the second time and 80% the third time). Very recently, Milstein

Scheme 18.11  Reaction mechanism of the methanol dehydrogenation catalyzed by pincer complexes.

18.3  Hydrogen Production from Methanol and Water

described a base-free procedure for reforming methanol [144]. The catalyst was a Ru pincer complex with a dearomatized acridine-based ligand. This type of pincer ligands offers a unique platform to obtain highly active catalysts [145]. After heating at 150ºC for 24 hours, a mixture of methanol and water in a 4:1 volumetric ratio, and the Ru-acridine pincer complex, a TOF of 4 h–1 was obtained. However, the addition of 1 equivalent, corresponding to the catalyst, of hexanethiol produce a tremendous increase in activity, reaching a TOF of 337 h–1. Furthermore, by increasing the amount of methanol (9:1 methanol:water), a TOF of 643 was obtained after six hours, but the TOF decreased to 480 h–1 after 12 hours. The presence of a higher amount of methanol increases the solubility of the catalyst, which has a positive effect in the catalytic performance. Fe-MACHO-BH catalyst was also active in the hydrogenation of methanol under mild conditions, but the stability of the catalyst and its activity decreases over time [137]. The TONmax was 9,834 in 46 hours, whereas the TONmax was 9184 in 111 hours using five extra-equivalents of catalyst, so the catalytic system is capable to produce hydrogen gas up to five days. In a further report, Bernskoetter, Hazari and Holthausen described that the addition of a Lewis acid produced significant enhancement in the catalytic activity of Fe-MACHO-OOCH complexes [146]. The addition of LiBF4 increases the yield and the TON of the reaction, going from 58% and 350 (20 min) to >99% and >599 (25 minutes), respectively. Under optimized conditions (4:1 ­methanol/ H2O (160 μL/ 18μL), 10 mol % LiBF4, 10 mL ethyl acetate, 0.006 mol % [Fe], reflux, 94 hours), the reaction produced a TON of 51,000 and yield of 50%. Using a higher catalyst loading (0.01 mol%), the full conversion of methanol was achieved in 52 hours. The use of manganese pincer complexes has produced attractive results with respect to dehydrogenation reactions. Hence, Beller et al. described the catalytic activity of MACHO-Mn toward the dehydrogenation of methanol under basic conditions [147]. The catalyst reached a TON > 20,000 using a mixture of methanol/water (20 mL, 9:1), triglyme (20 mL), 8 M KOH, and 10 equivalents of MACHO ligand at 92 °C for > 800 h, which means that the catalyst is stable for over a month. The catalytic activity of precious metals has been explored as well. The Ir-MACHO based catalyst was active in the dehydrogenation of methanol under basic conditions. Long-term experiments revealed that the catalyst is affected by the concentration of the base. Using 0.5 M of KOH, the conversion to hydrogen stops after 16 hours, giving a TON of 1,400, whereas the catalyst was stable for >60 hours using a KOH concentration of 8.0 M, achieving a TON of 1,900. Interestingly, the addition of a potential catalyst poison such as potassium carbonate did not affect the activity of the catalyst. Another attractive source of hydrogen must be water through a process called water splitting, which produces H2 and O2, with the advantage that the production of CO is theoretically not possible. As can be noted, this is the reverse of the reaction that occurs in a hydrogen fuel cell. The water splitting reaction has been promoted by several pincer complexes under different reaction conditions, mostly in the presence of sacrificial oxidants, reductants, or different additives [24]. In 1997, Bauer et al. described the catalytic activity of the catalyst [Co(tpp)2]Cl2 to produce hydrogen from water [148]. The reaction was induced by light and produced initial quantum yields up to 20%. Also the electrocatalytic reduction of H+ in the presence of [Co(terpy)2]-type complexes was studied [149, 150]. Other pincer metal complexes with Cu [151], Ni [152–160], Ru [161, 162] have been active in the electrocatalytic and photocatalytic reduction of water. Also, some precious transition metal-based pincer catalysts have been used to promote water splitting and related processes [163–169]. A major breakthrough in the water splitting field, was the production of hydrogen and oxygen from water catalyzed by Milstein’s catalyst [63, 170]. The reaction proceeds in two stages. Initially, Milstein’s catalyst react with water at 25 °C, yielding a Ru(II) hydride-hydroxo complex. The

401

402

18  Catalysis by Pincer Compounds and Their Contribution to Environmental and Sustainable Processes

production of hydrogen comes after the latter species was heated at 100 °C in water, together with the formation of a cis-dihydroxo complex, which by irradiation in the 320 to 420 nm range releases molecular oxygen and regenerates the Ru(II) hydride-hydroxo complex. Presumably,hydrogen peroxide (H2O2) is formed during this step through reductive elimination of two hydroxo ligands, then H2O2 disproportionates very quickly to H2O and oxygen. Remarkably,t the addition of sacrificial oxidants or reductants was not necessary in this reaction. The reaction mechanism has been studied by different research groups [171–175]. The activation of water is promoted by the dearomatized arm of the pincer, leading to an aromatized pincer ligand and a Ru(II) hydride-hydroxo species that was characterized by X-ray diffraction analysis; other Ru(II) hydride-hydroxo with different pincer ligands have also been characterized [140, 176]. Ru(II) hydride-hydroxo species may follow two routes as follows: (i) it can undergo a heterolytic coupling of the proton of the side arm with the hydride ligand, forming a dearomatized Ru pincer complex, which is actually the rate limiting step; (ii) a second molecule of water can react with the Ru(II) hydride-hydroxo species, forming hydrogen gas and a cis-dihydroxo complex. The light ­irradiation of the latter complex generates O2 and water. Fang and Chen proposed that two cis-dihydroxo complexes form a dimer, and successive hydrogen atom transfer (HAT) processes occur by photoexcitation, forming a triplet O2 molecule, water, and a carbonyl Ru(II)-complex [175]. According to their calculations, the formation of H2O2 was highly endothermic, and consequently, it is not formed during the reaction. Furthermore, all experimental findings described agree with this mechanism.

18.4  Biomass Transformation Biomass can be transformed to valuable organic compounds, but one of the most important transformations is in the generation of biofuels [45, 177–181]. The first generation of biofuels was obtained from vegetable oil, animal fats, or recycled restaurant grease [30, 31], as well as from fermentation of sugar-containing materials by yeasts. In the first case, it is possible to obtain so-called biodiesel, whereas bioethanol is produced in the second case. The second generation of biofuels was based on the decomposition of lignocellulosic biomass to produce a mixture of alkanes, alkenes, and alcohols by the Fischer-Tropsch method. At present, the use of butanol isomers as fuel has attracted much attention because these have a higher energy density compared to ethanol and are insoluble in water. In contrast to butanol isomers, ethanol forms an azeotrope with water that is difficult to purify and for direct application as a fuel. Also, butanol isomers have similar properties to gasoline; for example the caloric value of gasoline is 32.5 MJ/L and the value for butanol is 29.2 MJ/L. The research octane number of gasoline ranges from 91 to 99, compared with 96 for butanol. Thus, the use of butanol isomers is attractive. In fact, butanol is considered an “advanced biofuel” [182, 183]. Based on the above, pincer complexes appeared to be a good option for some key steps in the synthesis of biofuels. Williams et al. described a bicatalytic strategy to the conversion of vegetable-derived triglycerides (corn or soybean oil) to fatty acid methyl esters (FAME), which is the basis of the generation of biodiesel fuels [184]. The catalytic system was composed by two catalysts: a N-heterocyclic carbene Ir(I) complex and a Fe-MACHO compound. The iridium complex enables the catalytic hydrogenation of polyunsaturation in corn and soybean oils and the glyceride dehydrogenation in the presence of methanol as a hydrogen source and a base. The presence of the Fe-MACHO complex allows completion of the hydrogenation of the fatty acids, reaching a 90% yield. The synthesis of butanol isomers from ethanol has been the target to prepare advanced biofuels [185–190]. Ru-pincer complexes have shown good activity. For instance, an amide-derived

18.5  Conclusions

N,N,N-Ru(II) pincer and an acridine-based Ru(II)-pincer have been evaluated in the transformation of ethanol to butanol (Scheme 18.12). The N,N,N-Ru(II) pincer complex reached a 31% yield (TON = 420) using 10 mol% of NaOEt, 0.1 mol % of catalyst, and heating the reaction at 150 °C for two hours [185]. An improvement was observed using an acridine-based Ru(II)-pincer complex, which produced a conversion of 73.4%, giving a TON of 3671 [186]. The reaction conditions were similar: 0.03 mol% of catalyst, 20 mol % of NaOEt, at 150 °C for 40 hours. Mn-MACHO type complexes have been used for the synthesis of butanol isomers. The MACHO-Mn(I) complex reached an impressive TON of 114,120 with 92% of butanol selectivity at 160 °C and in the presence of NaOEt Scheme 18.12  Synthesis of butanol (12 mol%) [187]. Interestingly, the selectivity of the prod- isomers using pincer metal complexes. uct formation can be tuned by the temperature and the addition of others alcohols; for example, the addition of methanol generates the formation of iso-butanol at 200 °C [190]. The addition of an extra functionality in the N atom of the MACHO pincer ligand (Me-MACHO-Mn(I)) produced a negative effect in the catalytic activity.

18.5  Conclusions Pincer metal complexes have contributed enormously to the development of greener alternatives to produce energy sources and commodity products. They have been used in several reactions that involve hydrogenation and dehydrogenation reactions, allowing the usage and storage of H2. Also, they have been used to transform CO2 into valuable chemicals such as methanol that can be used as liquid organic hydrogen carrier. In addition, MACHO-type pincer complexes have shown unique reactivities toward the activation of different molecules. The presence of the central nitrogen atom in the ligand plays an active role in the activation of small molecules, such as dihydrogen, allowing the development of highly active catalysts for hydrogenation reactions. When the MACHO-type ligands are coordinated to Mn or Ru, they produce incredibly active catalysts capable of achieving challenging transformations. Among these, the hydrogenation of CO2 is currently probably one of the most important transformations and might help to palliate the effects of the climate change produced by CO2. The idea of transforming CO2 into valuable chemicals is very attractive, but it is even more relevant if it can be used to transport chemical hydrogen in the form of methanol. In a similar manner, Milstein’s catalyst has exhibited an incredibly catalytic performance in such a transformation. Finally, methanol and water are two compounds that have a high content of hydrogen, so their use as liquid hydrogen carriers is very convenient. MACHO-type and Milstein’s catalysts have shown an outstanding performance toward the production of molecular hydrogen using methanol and water as starting materials. Both types of complexes have exhibited a good selectivity to produce hydrogen and CO2, and a negligible amount of CO, during the overall process. Other alcohols such as ethanol can be transformed to advanced biofuels, which might result in novel technologies and applications toward more sustainable processes.

403

404

18  Catalysis by Pincer Compounds and Their Contribution to Environmental and Sustainable Processes

It is expected that the advances in the chemistry of pincer compounds and their uses in catalytic transformations may continue follow this lane and impact the further development of the processes discussed in this chapter, taking advantage of their well-known robustness, enhanced reactivities, and facile synthesis and structural modifications of these pincer species. This without a doubt represents a wide-open window of opportunities for future research in this field.

Acknowledgments H.V. thanks CONACYT (CVU: 410706) and Generalitat de Catalunya (Beatriu de Pinós MSCACofund 2019-BP-0080). D.M.-M. thanks the generous financial support of PAPIIT-DGAPA-UNAM (PAPIIT IN223323) and CONACYT A1-S-033933.

References 1 Moulton, C.J. and Shaw, B.L. (1976). J. Chem. Soc., Dalton Trans. 1976: 1020–1024. 2 van Koten, G., Timmer, K., Noltes, J.G., and Spek, A.L. (1978). J. Chem. Soc., Chem. Commun. 1978: 250–252. 3 van Koten, G. (1989). Pure Appl. Chem. 61: 1681–1694. 4 Roddick, D.M. (2013). Top. Organomet. Chem. 40: 49–88. 5 Valdés, H., González-Sebastián, L., and Morales-Morales, D. (2017). J. Organomet. Chem. 845: 229–257. 6 Valdés, H., Germán-Acacio, J.M., and Morales-Morales, D. (2018). In: Organic Materials as Smart Nanocarriers for Drug Delivery (ed. A.M. Grumezescu), 245–291. Elsevier - William Andrew. doi: 10.1016/B978-0-12-813663-8.00007-5. 7 Valdés, H., García-Eleno, M.A., Canseco-Gonzalez, D., and Morales-Morales, D. (2018). ChemCatChem. 10: 3136–3172. 8 Morales-Morales, D. and Jensen, C. (2007). The Chemistry of Pincer Compounds. Amsterdam: Elsevier. 9 Morales-Morales, D. (2008). Mini-Rev. Org. Chem. 5: 141–152. 10 Morales-Morales, D. (2004). Rev. Soc. Quim. Mex. 48: 338–346. 11 Serrano-Becerra, J.M. and Morales-Morales, D. (2009). Curr. Org. Synth. 6: 169–192. 12 Valdés, H., Rufino-Felipe, E., van Koten, G., and Morales-Morales, D. (2020). Eur. J. Inorg. Chem. 2020: 4418–4424. 13 Benito-Garagorri, D. and Kirchner, K. (2008). Acc. Chem. Res. 41: 201–213. 14 Zell, T. and Milstein, D. (2015). Acc. Chem. Res. 48: 1979–1994. 15 Choi, J., MacArthur, A.H.R., Brookhart, M., and Goldman, A.S. (2011). Chem. Rev. 111: 1761–1779. 16 Gunanathan, C. and Milstein, D. (2014). Chem. Rev. 114: 12024–12087. 17 Selander, N. and Szabó, K.J. (2011). Chem. Rev. 111: 2048–2076. 18 Van Der Boom, M.E. and Milstein, D. (2003). Chem. Rev. 103: 1759–1792. 19 Bertini, F., Glatz, M., Gorgas, N. et al. (2017). Chem. Sci. 8: 5024–5029. 20 Kang, P., Meyer, T.J., and Brookhart, M. (2013). Chem. Sci. 4: 3497–3502. 21 Suh, H.-W., Guard, L.M., and Hazari, N. (2014). Chem. Sci. 5: 3859–3872. 22 Peris, E. and Crabtree, R.H. (2018). Chem. Soc. Rev. 47: 1959–1968. 23 Szabó, K.J. (2013). Top. Organomet. Chem. 40: 203–242. 24 Zhang, H.-T. and Zhang, M.-T. (2021). Top. Organomet. Chem. 68: 379–449. 25 Valdés, H., Rufino-Felipe, E., and Morales-Morales, D. (2019). J. Organomet. Chem. 898: 120864.

References

26 Valdes, H., German-Acacio, J.M., van Koten, G., and Morales-Morales, D. (2022). Dalton Trans. 51: 1724–1744. 27 Kumar, A., Bhatti, T.M., and Goldman, A.S. (2017). Chem. Rev. 117: 12357–12384. 28 Alig, L., Fritz, M., and Schneider, S. (2019). Chem. Rev. 119: 2681–2751. 29 Wang, W.H., Himeda, Y., Muckerman, J.T. et al. (2015). Chem. Rev. 115: 12936–12973. 30 Zhou, Y., Remón, J., Jiang, Z. et al. (2022). Green Energy Environ. doi: 10.1016/j.gee.2022.03.001. 31 Kolet, M., Zerbib, D., Nakonechny, F., and Nisnevitch, M. (2020). Catalysts 10: 1189. 32 Falbe, J., Bahrmann, H., Lipps, W. et al. (2013). Ullmann’s Encyclopedia of Industrial Chemistry. Wiley—VCH. doi: 10.1002/14356007.a01_279.pub2. 33 Gregorio, G., Pregaglia, G.F., and Ugo, R. (1972). J. Organomet. Chem. 37: 385–387. 34 Matsu-ura, T., Sakaguchi, S., Obora, Y., and Ishii, Y. (2006). J. Org. Chem. 71: 8306–8308. 35 Gabriëls, D., Hernández, W.Y., Sels, B. et al. (2015). Cat. Sci. Tech. 5: 3876–3902. 36 Xue, W. and Tang, C. (2022). Energies 15: 2011. 37 Piccirilli, L., Lobo Justo Pinheiro, D., and Nielsen, M. (2020). Catalysts 10: 773. 38 Kar, S., Rauch, M., Kumar, A. et al. (2020). ACS Catal. 10: 5511–5515. 39 Wang, Y., Wang, M., Li, Y., and Liu, Q. (2021). Chem 7: 1180–1223. 40 Cabrero-Antonino, J.R., Adam, R., Papa, V., and Beller, M. (2020). Nat. Comm. 11: 3893. 41 Ekanayake, D.A. and Guan, H. (2020). Top. Organomet. Chem. 63: 263–320. 42 Elangovan, S., Topf, C., Fischer, S. et al. (2016). J. Am. Chem. Soc. 138: 8809–8814. 43 Csendes, Z., Brunig, J., Yigit, N. et al. (2019). Eur. J. Inorg. Chem. 2019: 3503–3510. 44 Weber, S., Brunig, J., Zeindlhofer, V. et al. (2018). ChemCatChem. 10: 4386–4394. 45 Padilla, R., Koranchalil, S., and Nielsen, M. (2020). Green Chem. 22: 6767–6772. 46 Garbe, M., Junge, K., and Beller, M. (2017). Eur. J. Org. Chem. 2017: 4344–4362. 47 Maji, B. and Barman, M. (2017). Synthesis 49: 3377–3393. 48 Mukherjee, A. and Milstein, D. (2018). ACS Catal. 8: 11435–11469. 49 Filonenko, G.A., van Putten, R., Hensen, E.J.M., and Pidko, E.A. (2018). Chem. Soc. Rev. 47: 1459–1483. 50 Kallmeier, F. and Kempe, R. (2018). Angew. Chem. Int. Ed. 57: 46–60. 51 Kaithal, A., Holscher, M., and Leitner, W. (2018). Angew. Chem. Int. Ed. 57: 13449–13453. 52 Zubar, V., Lebedev, Y., Azofra, L.M. et al. (2018). Angew. Chem. Int. Ed. 57: 13439–13443. 53 Kumar, A., Janes, T., Espinosa-Jalapa, N.A., and Milstein, D. (2018). Angew. Chem. Int. Ed. 57: 12076–12080. 54 Elangovan, S., Garbe, M., Jiao, H. et al. (2016). Angew. Chem. Int. Ed. 55: 15364–15368. 55 Espinosa-Jalapa, N.A., Nerush, A., Shimon, L.J.W. et al. (2017). Chem. Eur. J. 23: 5934–5938. 56 Widegren, M.B. and Clarke, M.L. (2018). Org. Lett. 20: 2654–2658. 57 Das, K., Waiba, S., Jana, A., and Maji, B. (2022). Chem. Soc. Rev. doi: 10.1039/d2cs00093h. 58 Zhang, J., Leitus, G., Ben-David, Y., and Milstein, D. (2006). Angew. Chem. Int. Ed. 45: 1113–1115. 59 Polukeev, A.V. and Wendt, O.F. (2018). J. Organomet. Chem. 867: 33–50. 60 Li, H., Wen, M., and Wang, Z.X. (2012). Inorg. Chem. 51: 5716–5727. 61 Gunanathan, C. and Milstein, D. (2011). Acc. Chem. Res. 44: 588–602. 62 Ben-Ari, E., Leitus, G., Shimon, L.J., and Milstein, D. (2006). J. Am. Chem. Soc. 128: 15390–15391. 63 Kohl, S.W., Weiner, L., Schwartsburd, L. et al. (2009). Science 324: 74–77. 64 Gunanathan, C., Gnanaprakasam, B., Iron, M.A. et al. (2010). J. Am. Chem. Soc. 132: 14763–14765. 65 Khaskin, E., Iron, M.A., Shimon, L.J. et al. (2010). J. Am. Chem. Soc. 132: 8542–8543. 66 Zeng, H. and Guan, Z. (2011). J. Am. Chem. Soc. 133: 1159–1161. 67 Waldie, K.M., Ostericher, A.L., Reineke, M.H. et al. (2018). ACS Catal. 8: 1313–1324. 68 Balaraman, E., Gunanathan, C., Zhang, J. et al. (2011). Nat. Chem. 3: 609–614.

405

406

18  Catalysis by Pincer Compounds and Their Contribution to Environmental and Sustainable Processes

69 Hasanayn, F., Baroudi, A., Bengali, A.A., and Goldman, A.S. (2013). Organometallics 32: 6969–6985. 70 del Pozo, C., Iglesias, M., and Sánchez, F. (2011). Organometallics 30: 2180–2188. 71 Sun, Y., Koehler, C., Tan, R. et al. (2011). Chem. Commun. 47: 8349–8351. 72 Kim, D., Le, L., Drance, M.J. et al. (2016). Organometallics 35: 982–989. 73 Le, L., Liu, J., He, T. et al. (2018). Organometallics 37: 3286–3297. 74 Kuriyama, W., Matsumoto, T., Ogata, O. et al. (2012). Org. Process Res. Dev. 16: 166–171. 75 Otsuka, T., Ishii, A., Dub, P.A., and Ikariya, T. (2013). J. Am. Chem. Soc. 135: 9600–9603. 76 Junge, K., Wendt, B., Jiao, H., and Beller, M. (2014). ChemCatChem. 6: 2810–2814. 77 Yuwen, J., Chakraborty, S., Brennessel, W.W., and Jones, W.D. (2017). ACS Catal. 7: 3735–3740. 78 Choi, J.H. and Prechtl, M.H.G. (2015). ChemCatChem. 7: 1023–1028. 79 Chakraborty, S. and Berke, H. (2014). ACS Catal. 4: 2191–2194. 80 Chakraborty, S. and Milstein, D. (2017). ACS Catal. 7: 3968–3972. 81 Chakraborty, S., Leitus, G., and Milstein, D. (2017). Angew. Chem. Int. Ed. 56: 2074–2078. 82 Vasudevan, K.V., Scott, B.L., and Hanson, S.K. (2012). Eur. J. Inorg. Chem. 2012: 4898–4906. 83 Avasare, V.D. (2022). Inorg. Chem. 61: 1851–1868. 84 Rezayee, N.M., Huff, C.A., and Sanford, M.S. (2015). J. Am. Chem. Soc. 137: 1028–1031. 85 Kar, S., Sen, R., Goeppert, A., and Prakash, G.K.S. (2018). J. Am. Chem. Soc. 140: 1580–1583. 86 Kothandaraman, J., Goeppert, A., Czaun, M. et al. (2016). J. Am. Chem. Soc. 138: 778–781. 87 Kar, S., Goeppert, A., and Prakash, G.K.S. (2019). ChemSusChem. 12: 3172–3177. 88 Kar, S., Sen, R., Kothandaraman, J. et al. (2019). J. Am. Chem. Soc. 141: 3160–3170. 89 Bai, S.-T., Zhou, C., Wu, X. et al. (2021). ACS Catal. 11: 12682–12691. 90 Zhang, L., Han, Z., Zhao, X. et al. (2015). Angew. Chem. Int. Ed. 54: 6186–6189. 91 Kar, S., Goeppert, A., Kothandaraman, J., and Prakash, G.K.S. (2017). ACS Catal. 7: 6347–6351. 92 Lane, E.M., Zhang, Y., Hazari, N., and Bernskoetter, W.H. (2019). Organometallics 38: 3084–3091. 93 Khusnutdinova, J.R., Garg, J.A., and Milstein, D. (2015). ACS Catal. 5: 2416–2422. 94 Huff, C.A. and Sanford, M.S. (2011). J. Am. Chem. Soc. 133: 18122–18125. 95 Chu, W.-Y., Culakova, Z., Wang, B.T., and Goldberg, K.I. (2019). ACS Catal. 9: 9317–9326. 96 Rayder, T.M., Adillon, E.H., Byers, J.A., and Tsung, C.-K. (2020). Chem 6: 1742–1754. 97 Li, Z., Rayder, T.M., Luo, L. et al. (2018). J. Am. Chem. Soc. 140: 8082–8085. 98 Kuß, D.A., Hölscher, M., and Leitner, W. (2021). ChemCatChem. 13: 3319–3323. 99 Sen, R., Goeppert, A., Kar, S., and Prakash, G.K.S. (2020). J. Am. Chem. Soc. 142: 4544–4549. 100 Sen, R., Koch, C.J., Goeppert, A., and Prakash, G.K.S. (2020). ChemSusChem. 13: 6318–6322. 101 Scheuermann, M.L., Semproni, S.P., Pappas, I., and Chirik, P.J. (2014). Inorg. Chem. 53: 9463–9465. 102 Mazzotta, M.G., Xiong, M., and Abu-Omar, M.M. (2017). Organometallics 36: 1688–1691. 103 Sanchez, P., Hernandez-Juarez, M., Rendon, N. et al. (2018). Dalton Trans. 47: 16766–16776. 104 Ng, C.K., Wu, J., Hor, T.S., and Luo, H.K. (2016). Chem. Commun. 52: 11842–11845. 105 Espinosa, M.R., Charboneau, D.J., Garcia de Oliveira, A., and Hazari, N. (2018). ACS Catal. 9: 301–314. 106 Li, H., Goncalves, T.P., Zhao, Q. et al. (2018). Chem. Commun. 54: 11395–11398. 107 Lv, H., Xing, Q., Yue, C. et al. (2016). Chem. Commun. 52: 6545–6548. 108 Bertini, F., Glatz, M., Stöger, B. et al. (2018). ACS Catal. 9: 632–639. 109 Mastalir, M., Glatz, M., Gorgas, N. et al. (2016). Chem. Eur. J. 22: 12316–12320. 110 Mastalir, M., Glatz, M., Pittenauer, E. et al. (2016). J. Am. Chem. Soc. 138: 15543–15546. 111 Kostera, S., Peruzzini, M., and Gonsalvi, L. (2021). Catalysts 11: 58. 112 Ma, N., Tu, C., Xu, Q. et al. (2021). Dalton Trans. 50: 2903–2914.

References

113 Chakraborty, S., Zhang, J., Krause, J.A., and Guan, H. (2010). J. Am. Chem. Soc. 132: 8872–8873. 114 Chakraborty, S., Patel, Y.J., Krause, J.A., and Guan, H. (2012). Polyhedron 32: 30–34. 115 Chakraborty, S., Zhang, J., Patel, Y.J. et al. (2013). Inorg. Chem. 52: 37–47. 116 Liu, T., Meng, W., Ma, Q.Q. et al. (2017). Dalton Trans. 46: 4504–4509. 117 Zhang, J., Chang, J., Liu, T. et al. (2018). Catalysts 8: 508. 118 Ma, Q.Q., Liu, T., Li, S. et al. (2016). Chem. Commun. 52: 14262–14265. 119 Erken, C., Kaithal, A., Sen, S. et al. (2018). Nat. Comm. 9: 4521. 120 Zhang, L., Zhao, Y., Liu, C. et al. (2022). Inorg. Chem. 61: 5616–5625. 121 Jia, Z., Li, L., Zhang, X. et al. (2022). Inorg. Chem. 61: 3970–3980. 122 Kostera, S., Peruzzini, M., Kirchner, K., and Gonsalvi, L. (2020). ChemCatChem. 12: 4625–4631. 123 Kar, S., Goeppert, A., and Prakash, G.K.S. (2019). J. Am. Chem. Soc. 141: 12518–12521. 124 Ryabchuk, P., Stier, K., Junge, K. et al. (2019). J. Am. Chem. Soc. 141: 16923–16929. 125 Kaithal, A., Werle, C., and Leitner, W. (2021). JACS Au 1: 130–136. 126 Mahajan, D. (2005). Top. Catal. 32: 209–214. 127 Kaithal, A., Holscher, M., and Leitner, W. (2020). Chem. Sci. 12: 976–982. 128 Dong, K., Elangovan, S., Sang, R. et al. (2016). Nat. Comm. 7: 12075. 129 Satapathy, A., Gadge, S.T., and Bhanage, B.M. (2017). ChemSusChem. 10: 1356–1359. 130 Satapathy, A., Gadge, S.T., and Bhanage, B.M. (2018). ACS Omega 3: 11097–11103. 131 Pri-Bar, I. and Alper, H. (1990). Can. J. Chem. 68: 1544–1547. 132 Gadge, S.T. and Bhanage, B.M. (2013). J. Org. Chem. 78: 6793–6797. 133 Palo, D.R., Dagle, R.A., and Holladay, J.D. (2007). Chem. Rev. 107: 3992–4021. 134 Salman, M.S., Rambhujun, N., Pratthana, C. et al. (2022). Ind. Eng. Chem. Res. 61: 6067–6105. 135 Nielsen, M., Alberico, E., Baumann, W. et al. (2013). Nature 495: 85–89. 136 Monney, A., Barsch, E., Sponholz, P. et al. (2014). Chem. Commun. 50: 707–709. 137 Alberico, E., Sponholz, P., Cordes, C. et al. (2013). Angew. Chem. Int. Ed. 52: 14162–14166. 138 Schwarz, C.H., Agapova, A., Junge, H., and Haumann, M. (2020). Catal. Today 342: 178–186. 139 Schwarz, C.H., Kraus, D., Alberico, E. et al. (2021). Eur. J. Inorg. Chem. 2021: 1745–1751. 140 Alberico, E., Lennox, A.J.J., Vogt, L.K. et al. (2016). J. Am. Chem. Soc. 138: 14890–14904. 141 Yang, X. (2014). ACS Catal. 4: 1129–1133. 142 Lei, M., Pan, Y., and Ma, X. (2015). Eur. J. Inorg. Chem. 2015: 794–803. 143 Hu, P., Diskin-Posner, Y., Ben-David, Y., and Milstein, D. (2014). ACS Catal. 4: 2649–2652. 144 Luo, J., Kar, S., Rauch, M. et al. (2021). J. Am. Chem. Soc. 143: 17284–17291. 145 Kar, S. and Milstein, D. (2022). Chem. Commun. 58: 3731–3746. 146 Bielinski, E.A., Förster, M., Zhang, Y. et al. (2015). ACS Catal. 5: 2404–2415. 147 Anderez-Fernandez, M., Vogt, L.K., Fischer, S. et al. (2017). Angew. Chem. Int. Ed. 56: 559–562. 148 Konigstein, C. (1997). Int. J. Hydrog. Energy 22: 471–474. 149 Abe, T. and Kaneko, M. (2001). J. Mol. Catal. A: Chem. 169: 177–183. 150 Guadalupe, A.R., Usifer, D.A., Potts, K.T. et al. (2002). J. Am. Chem. Soc. 110: 3462–3466. 151 Wang, J., Li, C., Zhou, Q. et al. (2016). Dalton Trans. 45: 5439–5443. 152 Luo, S., Siegler, M.A., and Bouwman, E. (2018). Organometallics 37: 740–747. 153 Yap, C.P., Chong, Y.Y., Chwee, T.S., and Fan, W.Y. (2018). Dalton Trans. 47: 8483–8488. 154 Ramakrishnan, S., Chakraborty, S., Brennessel, W.W. et al. (2016). Chem. Sci. 7: 117–127. 155 Mondragon, A., Flores-Alamo, M., Martinez-Alanis, P.R. et al. (2015). Inorg. Chem. 54: 619–627. 156 Jing, X., Wu, P., Liu, X. et al. (2015). New J. Chem. 39: 1051–1059. 157 Luca, O.R., Konezny, S.J., Hunsinger, G.B. et al. (2014). Polyhedron 82: 2–6. 158 Luca, O.R., Konezny, S.J., Blakemore, J.D. et al. (2012). New J. Chem. 36: 1149–1152. 159 Luca, O.R., Blakemore, J.D., Konezny, S.J. et al. (2012). Inorg. Chem. 51: 8704–8709.

407

408

18 Catalysis by Pincer Compounds and Their Contribution to Environmental and Sustainable Processes

160 Levina, V.A., Rossin, A., Belkova, N.V. et al. (2011). Angew. Chem. 123: 1403–1406. 161 Boyer, J.L., Polyansky, D.E., Szalda, D.J. et al. (2011). Angew. Chem. Int. Ed. 50: 12600–12604. 162 Chen, Z., Glasson, C.R., Holland, P.L., and Meyer, T.J. (2013). Phys. Chem. Chem. Phys. 15: 9503–9507. 163 Du, P., Schneider, J., Li, F. et al. (2008). J. Am. Chem. Soc. 130: 5056–5058. 164 Elvington, M., Brown, J., Arachchige, S.M., and Brewer, K.J. (2007). J. Am. Chem. Soc. 129: 10644–10645. 165 Yamauchi, K., Masaoka, S., and Sakai, K. (2009). J. Am. Chem. Soc. 131: 8404–8406. 166 Rabbani, R., Saeedi, S., Nazimuddin, M. et al. (2021). Chem. Sci. 12: 15347–15352. 167 Bakir, M., Lawrence, M.A.W., Ferhat, M., and Conry, R.R. (2017). J. Coord. Chem. 70: 3048–3064. 168 Lawrence, M.A.W. and Mulder, W.H. (2018). ChemistrySelect 3: 8387–8394. 169 Viertl, W., Pann, J., Pehn, R. et al. (2019). Faraday Discuss. 215: 141–161. 170 Sandhya, K.S., Remya, G.S., and Suresh, C.H. (2015). Inorg. Chem. 54: 11150–11156. 171 Li, J., Shiota, Y., and Yoshizawa, K. (2009). J. Am. Chem. Soc. 131: 13584–13585. 172 Yang, X. and Hall, M.B. (2010). J. Am. Chem. Soc. 132: 120–130. 173 Sandhya, K.S. and Suresh, C.H. (2011). Organometallics 30: 3888–3891. 174 Ma, C., Piccinin, S., and Fabris, S. (2012). ACS Catal. 2: 1500–1506. 175 Chen, Y. and Fang, W.H. (2010). J. Phys. Chem. 114: 10334–10338. 176 Gibson, D.H., Pariya, C., and Mashuta, M.S. (2004). Organometallics 23: 2510–2513. 177 Lao, D.B., Owens, A.C.E., Heinekey, D.M., and Goldberg, K.I. (2013). ACS Catal. 3: 2391–2396. 178 Ahmed Foskey, T.J., Heinekey, D.M., and Goldberg, K.I. (2012). ACS Catal. 2: 1285–1289. 179 Luque, R., Lin, C.S.K., Wilson, K. (2016). Handbook of Biofuels Production. Woodhead Publishing. 180 Li, W., Xie, J.-H., Lin, H., and Zhou, Q.-L. (2012). Green Chem. 14: 2388–2390. 181 Kirchhecker, S., Spiegelberg, B., and de Vries, J.G. (2021). Top. Organomet. Chem. 69: 341–395. 182 Durre, P. (2007). Biotechnol. J. 2: 1525–1534. 183 Harvey, B.G. and Meylemans, H.A. (2011). J. Chem. Technol. Biotechnol. 86: 2–9. 184 Lu, Z., Cherepakhin, V., Kapenstein, T., and Williams, T.J. (2018). ACS Sustain. Chem. Eng. 6: 5749–5753. 185 Tseng, K.N., Lin, S., Kampf, J.W., and Szymczak, N.K. (2016). Chem. Commun. 52: 2901–2904. 186 Xie, Y., Ben-David, Y., Shimon, L.J., and Milstein, D. (2016). J. Am. Chem. Soc. 138: 9077–9080. 187 Fu, S., Shao, Z., Wang, Y., and Liu, Q. (2017). J. Am. Chem. Soc. 139: 11941–11948. 188 Kulkarni, N.V., Brennessel, W.W., and Jones, W.D. (2018). ACS Catal. 8: 997–1002. 189 Rawat, K.S., Mandal, S.C., Bhauriyal, P. et al. (2019). Cat. Sci. Tech. 9: 2794–2805. 190 Liu, Y., Shao, Z., Wang, Y. et al. (2019). ChemSusChem. 12: 3069–3072.

409

19 Heterometallic Complexes Novel Catalysts for Sophisticated Chemical Synthesis Franco Scalambra, Ismael Francisco Díaz-Ortega, and Antonio Romerosa* Área de Química Inorgánica-CIESOL, Universidad de Almería, Almería, Spain * Corresponding author

19.1  Introduction With the aim of obtaining new more economic and eco-benign synthetic methodologies, researchers have gradually shifted toward new procedures to produce organic compounds with high yield, high stereoselectivity, and a large atom- and step-economy [1, 2]. Additionally, to meet the ­criteria suggested by green chemistry, the new synthetic processes must minimize waste and by-­products to achieve high overall efficiency. Natural enzymes, and the synthetic reactions in which they are involved, are examples of efficient catalytic systems, and have inspired a wide variety of artificial metal catalysts and cooperative organocatalysts over the last two decades ­[3–10]. In this respect, natural catalysts can transform substrates through the cumulative influence of multiple non-­covalent bonds and within tailored cavities, and these features are tentatively ­emulated by synthetic sophisticated macromolecular catalysts [11–18]. However, this goal requires laborious covalent modification of functional groups [19–21], which is one of the grand challenges of enzyme-inspired synthesis. Natural enzymes inspired also the design of heterometallic coordination complexes, which attempt to mimic the behaviour of natural catalysts. Most of the structures of these catalysts are the consequence of considering that the reaction rate will rise with increasing substrate concentration in a reduced space. In addition, it is thought that the selectivity of the reaction can be improved by placing the substrate in a suitable position close to the catalytic center. Multiple metal centers can provide the necessary microenvironment to adapt to the substrate and also act as independent or cooperative catalytic units to selectively transform a substrate by more selective procedure. Importantly, a suitable environment around the metal centers is desirable to achieve efficient dual activation of nucleophiles and electrophiles through metal cooperation. Therefore, the concept of bimetallic cooperative catalysis is an appropriate strategy, as it is also important to provide a suitable environment to the metals, achieving efficient dual activation of the target substrates. Thus, one of the key challenges in this field is to assemble the desired structures around the metal centers in a controllable way. In recent years, various synthetic procedures have been developed to construct heterometallic molecular architectures that mimic the functions of natural catalysts, and the two most important strategies are the so-called metal-ligand and self-sorting Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 2, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

410

19  Heterometallic Complexes

approaches. A metal-ligand can be described as a coordination fragment that ­features attached functional groups with the ability to coordinate other metal centers. Potential advantages of this approach include the precise control of the different metals as well as the ­structural rigidity, functionality, and location of the appended functional groups in the final ­heterometallic complex. Examples of the successful use of this strategy include the self-assembly of discrete heterometallic units and the synthesis of MOFs [22–24]. The self-sorting approach is based on the ability of different molecules to recognize their mutual homologues, working as a discrimination process to prepare a variety of binding motifs with high efficiency [25–27]. The term self-sorting is described as a network of competing recognition events defined by the binding constants between all possible pairs. Therefore, the followed strategy involves the formation of specific pairs rather than the creation of a library of all the possible non-covalent complexes that might form in the reaction mixture. This clearly distinguishes self-sorting from self-assembly strategies, in which generally identical building blocks are used repeatedly. Self-sorting systems are often based on the same type of binding motifs, which allows cross-interactions to occur ­between non-matching building blocks. Finally, there are other important factors to consider that are not directly related to the binding motifs themselves and can help to control the outcome of ­self-sorting. Of these factors, stoichiometry, spacer size and shape, or spacer-spacer interactions can be considered among the most important. In this chapter, we will describe the most interesting procedures in which heterometallic complexes have been used recently as catalysts and the advantages of these particular compounds over previously studied monometallic complexes, justifying the suitability of these catalysts despite their synthetic complexity.

19.2  C-X Formation (X = C, N, O, Metal) The synthesis of organic molecules by functionalization of the C-H bond is one of the keystones of catalysis, especially if the processes exhibit high efficiency and atom economy. A step forward in such reactions is the use of catalytic processes mediated by diheterometallic complexes, usually detected as reaction intermediates, which show new catalytic pathways [28] that differ from conventional catalytic processes due to cooperative activation of reactants, giving rise to the enhancement of reactivity and selectivity of reactions. Nevertheless, achieving precise control and understanding of the relationship between two given metals are current challenges [29, 30]. Due to their selectivity, low energy demand, chemical economy, and green approach, some of the most attractive reactions involve the photochemical generation of C-C bonds. In 2021, the modular piece-by-piece synthesis of new decorated mono- and heterobimetallic Au(I)/Ru(II) complexes containing π-conjugated [2.2]paracyclophane (PCP) (1–4) was presented [31]. This ligand is a useful rigid spacer that maintains the metals in proper spatial orientations and also the metal-to-metal distances in such a way that, depending on the ligand architecture, can impose ­different arrangements of the Ru and Au moieties (Figure 19.1). It is important to note that, beyond its interesting structural properties, this ligand is able to transfer electrons between the two metals due to the innate π-transannular communication through the cyclophane scaffold. Additionally, this ligand, both freely and coordinated to metal, is photosensitive to visible light. An interesting compound was the pseudo-para-AuRu complex (1), which shows non-additive as well as dynamic M-M interactions, as indicated by ultrafast transient absorption spectroscopy. Also, studies on isolated 1+ with ultrafast gas-phase photodissociation dynamics confirmed the intrinsic nature of the ultrafast formation of a long-lived electronically excited and reactive state.

19.2  C-X Formation (X = C, N, O, Metal)

Figure 19.1  [2.2] Paracyclophane substitution pattern for bimetallic Au(I)/Ru(II) complexes 1–4.

Another interesting feature of the heterodimetallic complexes presented in this study is their selective behaviour depending on the excitation wavelength: under irradiation at 312  nm, the pseudo-para-Au-Ru complex 1 experiences a clear influence on the dynamics of the Au moiety, but irradiation at lower energies (405 or 495 nm) only gives rise to dynamics analogous to Ru-bpy complexes, such as [Ru-(bpy)2(ppy)]. Based on the results obtained, complexes 1–4 were evaluated as visible photocatalysts for the arylative Meyer-Schuster rearrangement (Scheme 19.1). Only some of the obtained dimetallic complexes were found to be active catalysts for this reaction. The complexes display a series of steric and electronic qualities, but their most important property is probably the different Au(I)-Ru(II) distances, which define whether the desired electronic interaction between metals can take place.

Scheme 19.1  The Meyer-Schuster rearrangement reaction catalyzes the Au-Ru complexes 1–4.

It is important to note that the C-H borylation facilitates the subsequent C-H functionalization in some reactions. Additionally, boryl or s-borane transition metal intermediates play an important role in the functionalization of C-H bonds, providing, for example, distinct site selectivity for toluene and xylenes or predictable selectivity of the heterocycles ring-opening procedures. The polarity of C-M bonds provides new opportunities for organic synthesis beyond generalized application. Some recent examples have also shown how calcium hydride complexes such as 5 can promote, upon reaction with a diketominate Al species (6), the catalytic oxidative alumination of the sp2 C-H bonds of inactivated benzene, toluene, and xylene to give 7 at room temperature (Scheme 19.2) [32]. It was proposed that the cleavage of the C-H bond is accomplished by activation of

411

412

19  Heterometallic Complexes

Scheme 19.2  Ca-catalyzed alumination of arenes.

the arene by π-coordination to a Ca, acting as a Lewis acid, which is favored by simultaneous coordination to an Al center. Calculations of the relevant Al-H-Ca species showed that there is an unusual low positive charge on the Al center, similar to that found for the anionic alumanyl complexes that are able to cleave the strong C-H bond of benzene [33, 34]. Nevertheless, being very interesting from a synthetic point of view, the Ca-catalyzed alumination of toluene occurs almost selectively at the meta-position of the substrate. The proposed mechanism involves a nucleophilic attack of the aluminum reagent on an arene bearing an electron-donating substituent [35, 36]. In addition, recent studies have revealed the important role of the C-H activation step in reactions involving the cleavage of strong C-O and C-F bonds. When the palladium catalyst [Pd(PCy3)2] was introduced in the reaction, the bimetallic intermediate formed with 6 enables the formation of C-O and C-F via a C-H activation process. This intermediate increases the selectivity of the C-H bond functionalization by properly arranging the substrate and finally allowing the interaction with the metal centers (Scheme 19.3) [37, 38]. According to density functional theory (DFT) calculations the mechanism starts with a (4+1)-cycloaddition of the furan with 6, followed by a rearrangement of the intermediate. The selectivity of the reaction is correlated to the electronic influence of the substituent R′, which weakens the adjacent C-O sp2 bond. Deeper in the mechanism, the presence of the [Pd(PCy3)2] complex provides further selectivity to the reaction. At room temperature, the aluminum insertion at the -C-H in position 2 of the substrate gives rise to the kinetic product 8 (Scheme 19.3). Next, under more strained conditions and by reaction with [Pd(PCy3)2], the scission of the furan C-O bond occurs, which implies the attack of an aluminum-based metalloligand on the 2-palladate heterocycle, giving the intermediate 9. Therefore, the reactivation of 8 with palladium follows an alternative high-energy pathway that finally leads to the thermodynamic products 10a and 10b, with 10a being the major product. Nevertheless, when [Pd(PCy3)2] is not introduced in the reaction, the complex 10b is the major isomer. In the presence of [Pd(PCy3)2], also 3,4-dihydropyran and 2,3-dihydrofuran, which do not react cleanly with 6, can undergo this reaction. Finally, at room temperature and in the absence of [Pd(PCy3)2], the C-F alumination reaction of fluoroarenes with 6 (Scheme 19.3) does not occur [39], but C-F alumination of ­­mono-, di-, or trifluorobenzenes in the presence of the palladium complex takes place with high selectivity. The observed regioselectivity has been further investigated through mechanistic studies in combination with DFT calculations. These suggest a stepwise C-H to C-F functionalization process bringing the ­catalyst to a C-F bond adjacent to a reactive C-H site, giving the intermediate 12 and finally the isomers 13a and 13b.

19.2  C-X Formation (X = C, N, O, Metal)

Scheme 19.3  Pd-catalyzed C-O alumination of furans and Pd-catalyzed C-F alumination of fluoroarenes.

Finally, it is interesting to point out that the mentioned C-H functionalization was experimentally supported by determining a 100% atomic efficient palladium-­ catalyzed isomerization of the kinetic C-H alumination product 14 to the thermodyScheme 19.4  Pd-catalyzed isomerization of 14 to 15. namic C-F alumination product 15 (Scheme 19.4). The heterometallic Os-Cu complex 16 [40], which contains a pincer ligand, was presented in 2022 as a catalyst for the selective difunctionalization of unactivated aliphatic alkenes with nucleophiles, being the first example of a synthetic strategy based on cooperative heterometallic-metalloaromatic catalysis (Scheme 19.5). More than 80 substrates (including monosubstituted, 1,1-disubstituted, 1,2-disubstituted, 1,1,2-trisubstituted unactivated alkenes, dienes and trienes, and various O- or N-nucleophiles) were reacted. Results showed how good the system is for selective amino- or oxyselenation, providing a wide variety of functionalized products with up to 50:1 rr and 20:1 dr.

Scheme 19.5  Scope of the difunctionalization of unactivated aliphatic alkenes using 16.

413

414

19  Heterometallic Complexes

The experimental results supported that the system gives dual-site activation on the reaction substrates and enforces effective selectivity control, providing a robust catalytic framework and the ability to balance charges during reactions. Also, it was shown that the normally inactive osmium center enables the activation of N or O nucleophiles. (Scheme 19.6) Additionally, the metalloaromatic moiety provides further stabilization of the reaction intermediate and the Cu-Os bond and its cooperative effects determine the selectivity of the reaction by bringing the substrates close together.

Scheme 19.6  Reactivity of 16 with O- and N-nucleophiles.

In conclusion, cooperative metal-metalloaromatic catalysis significantly improves the reactivity and selectivity of the reactions in which the difunctionalization of unactivated alkenes occurs. The published results prove the important synergistic effects between two metals and show how a proper design of the bimetallic catalyst can promote new, more selective, and efficient, synthetic methods.

19.3  Oxidation Processes Interesting heterometallic complexes for oxidation reactions containing transition metal/lanthanides complexes were published recently. These complexes are receiving increasing attention due to their substantial activity as catalysts in a variety of synthetic processes, particularly in oxidation synthesis. Three isostructural tetranuclear heterometallic Co(II)−Ln(III) were described by Das et al. [41] The complexes [Co2Gd2L2(μ4-CO3)2(NO3)2] (17), [Co2Tb2L2(μ4-CO3)2(NO3)2] (18), and [Co2Dy2L2(μ4-CO3)2(NO3)2] (19), containing carbonato-bridges, were obtained by reaction of the Mannich ligand (H2L = N,N′-dimethyl l-N,N′-bis(2-hydroxy-3-methoxy-5-methylbenzyl)ethylenediamine) with Co(OAc)2·4H2O and Ln(III) nitrate salts (Ln = Gd, Tb and Dy) under atmospheric CO2 and basic conditions. The crystal structures of the obtained compounds are characterized as containing two dinuclear [(CoIIL)LnIII(NO3)] moieties linked through two μ4carbonato groups forming the tetranuclear {CoII2LnIII2} core (Scheme 19.7). The combination of Co(II) with the Ln(III) center led to heterometallic systems with moderate catalytic catecholase-like activity (Scheme 19.8) and quite high phenoxazinone-synthase-like activity (Scheme 19.9). In contrast, the analogue homometallic tetranuclear Co(II) complexes were inactive in the studied synthetic processes. The turnover frequency (TOF) values obtained for the catecholase-like reaction catalyzed by complexes 17–19 were respectively 254.5 h−1, 272.4 h−1 and 291.3 h−1, whereas for the phenoxazinone-synthase-like reaction they were 2930.6 h−1, 2965.2 h−1, and 2998.5 h−1. Results obtained from the study of the reactions with mass spectrometry suggested that the dinuclear [(CoIIL)LnIIICO3]2+ ions, which form upon dissociation of the tetranuclear complexes,

19.3  Oxidation Processes

Scheme 19.7  Synthesis of complexes 17–19. The general structure of the products is depicted from the molecular structure of 19 as found by single-crystal X-ray diffraction. Hydrogen atoms were removed for clarity. Adapted from [41].

Scheme 19.8  Oxidation of 3,5-DTBC to 3,5-DTBQ catalyzed by 17–19.

Scheme 19.9  Catalytic oxidation of o-aminophenol to phenoxazinone mediated by 17–19.

415

416

19  Heterometallic Complexes

are the catalytically active species. The authors proposed that the interaction between the substrate and catalyst is favored by the higher oxophilicity of Ln(III) ions, whereas the cobalt ion oxidizes the substrate by exchanging its oxidation state between +2 and +3, assisted by molecular oxygen in the electronic give-and-take. Therefore, the catalytic oxidase activities of the studied heterometallic complexes are favored by the presence of two metal centers.

19.4  CO2 Activation The activation of CO2 is a process of paramount importance in chemistry to obtain a large variety of important organic molecules with industrial and biological relevance, and currently of particular interest for the sequestration of atmospheric CO2. With respect to the activation of CO2 controlled by heterometallic catalysts, a pinacolborane oxidation parallel to an interesting CO2 reduction was described by Bagherzadeh and Mankad in 2015 [42]. The authors employed heterometallic Cu-M complexes (20–25) [43], consisting of a {NHC-Cu} unit bonded to the carbonyl half sandwiches [MoCp(CO)3], [WCp(CO)3], and [FeCp(CO)2]- (Figure 19.2), as catalysts to react CO2 with (pin) B-B(pin). The reaction gave CO and HCO2B(pin) (compound 23 in Figure 19.2), as well as the boryl ether (pin)B-O-B(pin) (24, Figure 19.2) (pin = pinacole), through deoxygenation of CO2. On the other hand, if the product (pin)B-OH (25) is afforded, which is also generated by fulfilling complete CO2 deoxygenation, it would react with H-B(pin) to generate 27 and H2, and therefore should be difficult to be isolated and determined as the reaction product. The lability of the [M]-(Bpin) seems related to the rate of decarbonylation, and follows the [M] leaving group ability ([MoCp(CO)3]> [WCp(CO)3]>[FeCp(CO)2]), which also correlates with their relative pKa values [44]. Despite the studies performed, the authors have not been able to determine if the decarbonylation reactivity could also come from the polarity inversion during the catalyst activation step, which gives rise to the pair (NHC)Cu(Bpin) + [M]H, where the {(NHC)Cu(Bpin)} intermediate is the catalyst for the deoxygenation of CO2 (NHC = N-heterocyclic carbene) [45]. The proposed mechanism for this process is shown in (Scheme 19.10): initially, the NHC-Cu-[M] complex undergoes self-activation and the Cu-component reacts with CO2. The resulting Cu-formate transfers CO2 to the HB(pin) upon insertion on the B-H bond, forming 26. Then, through rate-determining, electrophilic activation by [M]-E (E = (Bpin) at the first turnover and then H throughout the process) the transfer of E+ to 26 initiates the decarbonylation pathway, which proceeds through the intermediate 29. In this reaction step, the reversible activation should be conducted through the sequestration of [M]− by CO2 [46–48]. Based on the known catalytic properties of the monometallic {(NHC)Cu} moiety for CO2 reduction, the heterometallic complexes (NHC)Cu-[M] 20–25 (NHC = N-heterocyclic carbenes) were synthesized and their catalytic activity studied for this reaction [49]. In contrast to monometallic Cu complexes, which catalyze exclusively the hydroboration of CO2 with pinacolborane to

Figure 19.2  Structures of complexes 20–25.

19.4  CO2 Activation

Scheme 19.10  (a) Catalytic pinacolborane oxidation via CO2 reduction using 20–25. (b) Proposed catalyst activation and hypothetical catalytic mechanism.

produce formate, analogous diheterometallic complexes containing Cu−Fe, Cu−W and Cu−Mo give rise to a mixture of pinacolborane-formate and CO. The reaction selectivity on CO vs. formate was controlled by tuning the electronic nature of the Cu/M pair, being the Cu-Mo complex which generates catalytically the largest amount of CO. Heterometallic complexes as catalytic activators of CO2 have been extended to the polymerization reaction. Polyolefins, cheap and easy to produce, are extensively used in industry for a wide variety of applications, being frequently found in packaging, automotive and electrical components, lubricants, and medical devices, among other products. One of the most interesting features of polyolefins is how their properties can be deeply changed by the catalyst used in their synthesis. Modifications of ligands, metal and geometry of the complex, and the employed co-catalyst can deeply modify the molecular weight and branching of the final product. The inclusion of a second catalytic center usually results in important alterations that affect the catalytic activity, polymer mass and structure, and comonomer enchainment selectivity, offering an additional point to tune the polymerization. These effects are as strong as short the M···M distance is, which can be modulated by ligation, counterion and solvent effects [50]. Ring-opening copolymerization (ROCOP) of epoxides and heterocumulenes is an attractive strategy to recycle CO2, transforming it directly into useful materials such as polycarbonates/ester copolymers and also polycarbonate polyols, which can be further processed to synthesize polyurethanes [51–66]. So far, homodinuclear complexes have been used for epoxide and carbon dioxide/anhydride ring-opening copolymerization (ROCOP) but a few examples of heterodinuclear catalysts are known and were reported only recently. The first example of a catalyst for ROCOP was the Zn(II)/Mg(II) complex (30) (Figure 19.3), which is based on a tetraiminodiphenol macrocyclic ligand presented in 2015 [67].

417

418

19  Heterometallic Complexes

Figure 19.3  Structures of complexes 30–32 and cyclohexene oxide/CO2 catalyzed copolymerization.

Complex 30 performs the polymerization reaction significantly better than the equivalent homodimetallic 31 and 32, which contain Mg and Zn, respectively. For example, in the copolymerization of CO2 and cyclohexene oxide at 1 bar, the TOF and turnover number (TON) obtained with 30 were more than double the values obtained with 31, while 32 has not shown catalytic activity for the reaction. Also, 30 showed excellent efficiency for CO2 uptake and high reaction control, giving polymers with narrow molecular weight distribution and an almost unitary ratio of carbonate linkages. The high degree of control of the polymerization mediated by 30 was also supported by a higher molecular weight of the obtained polymer and a marked increase of TOF (624 h−1) at lower catalyst loadings, which is significantly competitive with results obtained with mononuclear catalysts containing Cr and Co [68–70]. In general, complex 30 showed a higher polymerization selectivity, control, and activity in comparison with the homodinuclear analogues and also combinations of them. The advantages offered by the heteronuclearity seemed related to the augmented epoxide coordination offered by the Zn, together with a faster carbonate attack provided by the Mg (Scheme 19.11). Additionally, this synergy between the ­different metal centers also activates the halide co-ligand to start the polymerization, which was found innocent when the homometallic di-zinc complex 32 was used [71–74].

Scheme 19.11  Proposed mechanism for CO2/epoxide copolymerization catalyzed by the dinuclear complexes 30–32.

19.4  CO2 Activation

Another interesting approach was the use of heterobimetallic catalysts containing the {(salen) Al} moiety for the lactide ring-opening polymerization (ROP) [75–77], which underlined the real importance of choosing the adequate metal centers. The monometallic complexes containing Al (34), Mg (35a, 35b) and Zn (36) based on the salen-type ligand (33) were prepared and compared with their heterometallic derivatives containing Al/Mg (37), Al/Zn (38), Al/Li (39) and Al/Ca (40) (Scheme 19.12). Whereas monometallic Mg and Zn complexes were found to be inactive in ­rac-lactide ROP (rac-LA ROP), Al/Mg and Al/Zn combinations speeded up the catalytic reaction to 11 times compared with the mono-Al (34). On the other hand, Al/Mg and Al/Zn combinations inhibited the polymerization rates. Based on ab initio molecular dynamics (AIMD) calculations, structural studies and reaction kinetics, the authors propose that the improvement of the catalytic activity obtained with the Al/Mg and Al/Zn heterometallic complexes can arise from various factors: one of the reasons is the capacity of the second metal to provide additional coordination

Scheme 19.12  Synthesis of complexes 34–40.

419

420

19  Heterometallic Complexes

sites for monomer binding, exhibiting additional Lewis acid sites and also weakening the Al–Cl bond, speeding up the reaction initiation. Also, the chloride bridge causes the Al center to adopt a square pyramidal geometry, that augments the vacant sites at the aluminium, thus shortening the induction period. Finally, the proximity of the metal center seems to compensate for the structural rigidity of the Al/Mg and Al/Zn complexes. It is important to point out that the use of AIMD calculations (applied for the first time to ROP by the authors) revealed fundamental details about the reactivity, offering a deeper and detailed understanding of the features that drive the activity of the studied catalysts.

19.5  O2 and H2 Generation from Water The world’s increasing fuel demand is directing many scientific efforts toward the development of practical processes that allow energy production from the greener and cheapest source at our disposal: water. In this sense, water can be split to cleanly generate oxygen and hydrogen, which can be subsequently used as fuel. So far, many electrode systems and heterogeneous catalysts have been proposed for electrochemical and photoelectrochemical water splitting. With respect to homogeneous catalysis, many transition metal complexes were found to be practical for the generation of oxygen and hydrogen from water, but only in the last decade the scientific efforts, taking example from the natural photosystem II (PS-II), started to consider multimetallic systems as an attractive concept to solve some of the problems inherent to water splitting [78–81]. An ideal artificial system inspired by PS-II should be composed of an anode consisting of a photosensitizer that allows the oxidation of two water molecules by a water oxidation catalyst (WOC) (Eq. 19.1), while, at the cathode side, four electrons are employed by a hydrogen evolving catalyst (HEC) to reduce four protons and generate two molecules of hydrogen (Eq. 19.2). Due to its complexity, the water oxidation step (Eq. 19.1) is the most problematic process.

2H2O → O2 +4H+ + 4e− +

−

4 H + 4 e → 2H2

(19.1) (19.2)

Among the proposed methods to obtain an efficient WOC, the radical coupling between the oxygen of a hydroxocerium(IV) ion and a metal oxo complex is an attractive path. The ruthenium(III) complex [Ru(L)(pic)3] (41) (H3L = 2,2′-iminodibenzoic acid, pic = 4-methylpyridine) [82], is an interesting example of a catalyst designed by taking into account and putting together various beneficial factors targeted to oxidation of liquid water: (i) the presence of a negatively charged ligand to stabilize high valent metal intermediate species; (ii) the choice of a redox non-innocent ligand that can cooperate with the metal in the electron transfers; (iii) water coordination to the metal to support proton-coupled electron transfer. In the presence of sacrificial ceric ammonium nitrate at pH = 1, complex 41 can generate O2 with TON = 200 and very high TOF (0.168 s−1). The mechanism of the water oxidation mediated by 41 was investigated by kinetic studies and DFT calculations. What emerged is that the oxidation of water occurs at the ruthenium center, where the metal-ligand cooperation allows for the storage of many oxidative equivalents at a single site, which is mandatory for the required four electrons transfer to generate one O2 molecule. The formation of the O-O bond proceeds via formal high-valent Ru(VII) species involving a multimetallic Ru-Ce intermediate (Scheme 19.13) with the participation of the anionic tridentate ligand in the electron-transfer process.

19.5  O2 and H2 Generation from Water

Scheme 19.13  The proposed catalytic mechanism for the 41/CeIV catalyzed water oxidation at pH = 1.

On the other hand, experimental results support that heterometallic complexes are valuable multielectron transfer platforms and therefore they can be a beneficial catalyst for water reduction reactions. Production of H2, particularly from water, is one of the most current exciting and demanding research lines. The palpable climatic change, extensively produced by the atmospheric accumulation of human-generated CO2, obliges us to pay deep attention to seek new energy vectors different to fossil fuels, being hydrogen one of the most promising. Due to the well-known reaction of H2 with O2 gives rise to H2O and a large amount of energy. Nevertheless, one of the big limits to the general use of H2 as a substitute for non-renewable energy sources is developing ways to synthesize it by a practical and economical procedure. Water should be the natural favorite source for generating H2 but the process requires a large amount of energy, which could be obtained from a renewable source such as the sun. Therefore, very intense research efforts are

421

422

19  Heterometallic Complexes

Figure 19.4  Top panel: structures of complexes 42 and 43. Bottom panel: representation of the molecular structure of 43 as found by single-crystal X-ray diffraction; hydrogen atoms and anion were omitted for clarity.

ongoing to obtain catalysts whose reactivity is driven by light. Although several mononuclear species have been reported up to date, the exploration of multimetallic molecular photocatalysts started only recently. In 2021, Verani et al. presented a very interesting approach to water photoreduction [83]. The idea, supported by DFT, was to covalently bind two Ru(II) photosensitizer to a [NiII(oxime)] complex. The Ru(II) domains, serve as photoactivators, providing to the Ni center the electrons needed for proton reduction. To prove and support this hypothesis, the dimetallic [(bpy)2RuIINiII(L1)](ClO4)2 (42) and a trimetallic [(bpy)2RuIINiII(L2)2RuII(bpy)2](ClO4)2 (43) (Figure 19.4) were synthesized. Both complexes are able to generate the low-valent precursor involved in hydride formation before generating dihydrogen. Nevertheless, complex 42 was found ineffective as a catalyst for the process due to the high energy barrier required for the protonation of the Ni(I) center after photoactivation. Further calculations and experimental clues obtained using 42 in presence of [Ru(bpy)3]2+ revealed that a second electron was necessary to trigger the reaction. Thus, complex 43 was synthesized and under blue light, as expected, the predicted protonation of the Ni(I) with the formation of a [RuII(NiII-H–) RuII] species was achieved. Finally, this species was able to react with one proton to generate hydrogen. The entire catalytic process required 24 h, being the reaction TON of 49 in H2O/CH3CN at pH = 11, using triethylamine as a sacrificial electron donor (Scheme 19.14). The experimental and DFT studies suggested that in the catalytic process monometallic Ni(0) species were not formed and the second electron coming from the photoexcitation presumably remains on a bpy ring of one of the Ru(II) domains. One of the most recent works published concerning catalytic H2 addresses photo-generation using a bimetallic complex (44–46), showing that complexes containing noble metals are still ­components of reliable strategies to develop useful catalysts for this process, even if exploiting abundant metals should be the target for designing new catalysts [84]. Among the published

19.5  O2 and H2 Generation from Water

Scheme 19.14  Proposed mechanism for the photocatalyzed proton reduction, being catalyzed by 43.

complexes, the most active for photocatalytic H2 production was the Ru(II)-Rh(III) complex 46, which was found to be able also to photocatalyze the generation of H2 more efficiently than its parts and other similar complexes (Figure 19.5). Only a small quantity, almost a trace amount, of H2 was photo-generated by irradiation at λ = 480 nm from an Ar-saturated mixture of dimethylacetamide−triethanolamine (4:1, v/v) together with the reductant 1,3-dimethyl-2-phenyl-2,3-dihydro-1H-benzo-[d]imidazole. Nevertheless, when a proton source was introduced in the reaction (3,5-difluorophenol ­(3,5-F2−PhOH) or 2-(1,3-dimethyl-2,3-dihydro-1H-benzo[d]imidazol-2-yl)phenol) the generation of H2 increased significantly. Interestingly, when the reaction was carried out under CO2 without acid, the main product generated was H2 (30.9 μmol; TONH2 =157.5), producing small amounts of HCOOH and CO. It is important to stress that the high quantum yield of the reaction (ΦH2 = 16.4%) was supported by the fact that the reaction of triethanolamine with CO2 generates protons [85] and gives rise to a zwitterionic alkyl carbonate that acts as an acid, being the H+ source. Under an Ar atmosphere and without acid, there are not enough protons for the photoproduction of H2.

Figure 19.5  Structure of complexes 44–46.

423

424

19  Heterometallic Complexes

The high photochemical H2 generation mediated by 46 was justified as a consequence of the faster electron transfer from the Ru photosensitizer unit to the Rh catalyst center [86]. The photocatalytic ability of the Ru(II)−Rh(III) heterometallic complexes 44–46 is dependent on the diimine ligand coordinated to the Rh, which determines the final product under the same reaction conditions. A strong electron-donating methoxy substituent at the 6,6′-­positions of the bipyridine bridging ligand induces a high and selective formation of H2, while an electroattracting substituent mostly leads to the formation of HCOOH, which was previously observed for Ru-mononuclear complexes containing bpy. The complex 44, in which the ­bridging ligand bipyridine does not support substitutions, is not active for producing photocatalytically H2. This was explained by the authors on the basis that RuII−RhIII(H) and/or RuII−Rh(Cp*H) species do not react with a proton, even under the CO2 atmosphere, producing an excited Ru unit that reacts with CO2, giving rise finally to HCOOH. In contrast, an electron-donating substituted at the 6,6′-positions of the diimine ligand, which is also close to the Rh metal center, could help a mutual interaction between the proton source and “hydride reduced” species in the intermediates RuII−RhIII(H)−OMe and/or RuII−Rh(Cp*H)−OMe and the faster generation of H2 [87].

19.6  Conclusions Recent results in processes catalyzed by heterometallic complexes are well-defined evidences of the huge advantages that the synergic collaboration between metals can provide to homogeneous catalysis when they are forming part of a single complex. Nevertheless, the introduction of additional metallic centers in the structure of a catalyst complicates enormously its synthesis and the interpretation of the mechanisms, increasing the effort needed to improve its performance. Some indications of the structure-activity-relationship of multimetallic catalysts can be extracted from the obtained results, which give some general indications about their design and possible properties. Firstly, the chosen metals need to display an affinity with the substrate in the starting complex and/or reaction intermediates, modulating the balance between activation and stabilization. Then, the distance between the metals is extremely important and should be weighed based on the substrate size. Also, for reactions involving electron transfers, groups connecting metals electronically are necessary. Finally, for photocatalytic reactions, an accurate evaluation of the combined excited state properties is essential to drive the reactivity far from unproductive pathways. In addition to all of these factors, the designed catalyst must be stable and robust to air and light and preferably built with non-noble metals.

Note The authors declare no conflict of interest.

Acknowledgments The authors thank Junta de Andalucía for funding the group PAI FQM-317 and the project PY20_00791, and the University of Almería for the project UAL2020-RNM-B2084 (both projects co-funded by the European Commission FEDER program).

References

References 1 Wender, P.A. and Miller, B.L. (2009). Nature 460 (7252): 197–201. 2 Trost, B. (1991). Science 254 (5037): 1471–1477. 3 Sawamura, M. and Ito, Y. (1992). Chem. Rev. 92 (5): 857–871. 4 Ma, J.-A. and Cahard, D. (2004). Angew. Chem. Int. Ed. 43 (35): 4566–4583. 5 Yamamoto, H. and Futatsugi, K. (2005). Angew. Chem. Int. Ed. 44 (13): 1924–1942. 6 Kanai, M., Kato, N., Ichikawa, E., and Shibasaki, M. (2005). Synlett 2005 (10): 1491–1508. 7 Taylor, M.S., Jacobsen, E.N., Jacobsen, E.N., and Taylor, M.S. (2006). Angew. Chem. Int. Ed. 45 (10): 1520–1543. 8 Mukherjee, S., Yang, J.W., Hoffmann, S., and List, B. (2007). Chem. Rev. 107 (12): 5471–5569. 9 Shibasaki, M., Kanai, M., Matsunaca, S., and Kumagai, N. (2009). Acc. Chem. Res. 42 (8): 1117–1127. 10 Park, J. and Hong, S. (2012). Chem. Soc. Rev. 41 (21): 6931–6943. 11 Ward, M.D., Hunter, C.A., and Williams, N.H. (2018). Acc. Chem. Res. 51 (9): 2073–2082. 12 Jongkind, L.J., Caumes, X., Hartendorp, A.P.T., and Reek, J.N.H. (2018). Acc. Chem. Res. 51 (9): 2115–2128. 13 Brown, C.J., Toste, F.D., Bergman, R.G., and Raymond, K.N. (2015). Chem. Rev. 115 (9): 3012–3035. 14 Gianneschi, N.C., Masar, M.S., and Mirkin, C.A. (2005). Acc. Chem. Res. 38 (11): 825–837. 15 Yoshizawa, M., Klosterman, J.K., and Fujita, M. (2009). Angew. Chem. Int. Ed. 48 (19): 3418–3438. 16 Otte, M. (2016). ACS Catal. 6 (10): 6491–6510. 17 Pluth, M.D., Bergman, R.G., and Raymond, K.N. (2009). Acc. Chem. Res. 42 (10): 1650–1659. 18 Raynal, M., Ballester, P., Vidal-Ferran, A., and van Leeuwen, P.W.N.M. (2014). Chem. Soc. Rev. 43 (5): 1660–1733. 19 Hasell, T. and Cooper, A.I. (2016). Nat. Rev. Mater. 1 (9): 1–14. 20 Gramage-Doria, R., Armspach, D., and Matt, D. (2013). Coord. Chem. Rev. 257 (3–4): 776–816. 21 Gale, P., Gunnlaugsson, T., and Ballester, P. (2010). Chem. Soc. Rev. 39 (10): 3810–3830. 22 Jansze, S.M. and Severin, K. (2018). Acc. Chem. Res. 51 (9): 2139–2147. 23 Wise, M.D., Ruggi, A., Pascu, M. et al. (2013). Chem. Sci. 4 (4): 1658–1662. 24 Kumar, G. and Gupta, R. (2013). Chem. Soc. Rev. 42 (24): 9403–9453. 25 Safont-Sempere, M.M., Fernández, G., and Würthner, F. (2011). Chem. Rev. 111 (9): 5784–5814. 26 Lal Saha, P.M., Schmittel, M., Saha, M.L., and Schmittel, M. (2012). Org. Biomol. Chem. 10 (24): 4651–4684. 27 Ghosh, S. and Isaac, L. (2009). Complex self-sorting systems. In: Dynamic Combinatorial Chemistry: In Drug Discovery, Bioorganic Chemistry, and Materials Science (ed. B.L. Miller), 118–154. John Wiley & Sons, Ltd. 28 Batuecas, M., Gorgas, N., and Crimmin, M.R. (2021). Chem. Sci. 12 (6): 1993–2000. 29 Campos, J. (2020). Nat. Rev. Chem. 4 (12): 696–702. 30 Deacy, A.C., Kilpatrick, A.F.R., Regoutz, A., and Williams, C.K. (2020). Nat. Chem. 12 (4): 372–380. 31 Zippel, C., Israil, R., Schüssler, L. et al. (2021). Chem. Eur. J. 27 (61): 15188–15201. 32 Brand, S., Elsen, H., Langer, J. et al. (2019). Angew. Chem. Int. Ed. 58 (43): 15496–15503. 33 Kurumada, S., Takamori, S., and Yamashita, M. (2019). Nat. Chem. 12 (1): 36–39. 34 Grams, S., Eyselein, J., Langer, J. et al. (2020). Angew. Chem. Int. Ed. 59 (37): 15982–15986. 35 Hicks, J., Vasko, P., Heilmann, A. et al. (2020). Angew. Chem. Int. Ed. 59 (46): 20376–20380. 36 Kurumada, S., Sugita, K., Nakano, R. et al. (2020). Angew. Chem. Int. Ed. 59 (46): 20381–20384. 37 Rekhroukh, F., Chen, W., Brown, R.K. et al. (2020). Chem. Sci. 11 (30): 7842–7849. 38 Hooper, T.N., Brown, R.K., Rekhroukh, F. et al. (2020). Chem. Sci. 11 (30): 7850–7857. 39 Kysliak, O., Görls, H., and Kretschmer, R. (2020). Chem. Commun. 56 (57): 7865–7868.

425

426

19  Heterometallic Complexes

40 Cui, F.H., Hua, Y., Lin, Y.M. et al. (2022). J. Am. Chem. Soc. 144 (5): 2301–2310. 41 Das, A., Goswami, S., Sen, R., and Ghosh, A. (2019). Inorg. Chem. 58 (9): 5787–5798. 42 Bagherzadeh, S. and Mankad, N.P. (2015). J. Am. Chem. Soc. 137 (34): 10898–10901. 43 Banerjee, S., Karunananda, M.K., Bagherzadeh, S. et al. (2014). Inorg. Chem. 53 (20): 11307–11315. 44 King, R.B. (1970). Acc. Chem. Res. 3 (12): 417–427. 45 Laitar, D.S., Müller, P., and Sadighi, J.P. (2005). J. Am. Chem. Soc. 127 (49): 17196–17197. 46 Lee, G.R. and Cooper, N.J. (1985). Organometallics 4 (4): 794–796. 47 Zieliński, M., Zielińska, A., and Papiernik-Zielińska, H. (1994). J. Radioanal. Nucl. Chem. 183 (2): 301–311. 48 Halpern, J. and Kemp, A.L.W. (1966). J. Am. Chem. Soc. 88 (22): 5147–5150. 49 Zhang, L. and Hou, Z. (2013). Chem. Sci. 4 (9): 3395–3403. 50 McInnis, J.P., Delferro, M., and Marks, T.J. (2014). Acc. Chem. Res. 47 (8): 2545–2557. 51 Takeda, N. and Inoue, S. (1978). Makromol. Chem. 179 (5): 1377–1381. 52 Moore, D.R., Cheng, M., Lobkovsky, E.B., and Coates, G.W. (2003). J. Am. Chem. Soc. 125 (39): 11911–11924. 53 Luinstra, G.A. (2008). Polym. Rev. 48 (1): 192–219. 54 Kember, M.R., Knight, P.D., Reung, P.T.R., and Williams, C.K. (2009). Angew. Chem. Int. Ed. 48 (5): 931–933. 55 Nakano, K., Hashimoto, S., and Nozaki, K. (2010). Chem. Sci. 1 (3): 369. 56 Vagin, S.I., Reichardt, R., Klaus, S., and Rieger, B. (2010). J. Am. Chem. Soc. 132 (41): 14367–14369. 57 Klaus, S., Lehenmeier, M.W., Herdtweck, E. et al. (2011). J. Am. Chem. Soc. 133 (33): 13151–13161. 58 Jutz, F., Buchard, A., Kember, M.R. et al. (2011). J. Am. Chem. Soc. 133 (43): 17395–17405. 59 Kember, M.R. and Williams, C.K. (2012). J. Am. Chem. Soc. 134 (38): 15676–15679. 60 Wu, G.-P., Darensbourg, D.J., and Lu, X.-B. (2012). J. Am. Chem. Soc. 134 (42): 17739–17745. 61 Lehenmeier, M.W., Kissling, S., Altenbuchner, P.T. et al. (2013). Angew. Chem. Int. Ed. 52 (37): 9821–9826. 62 Liu, Y., Ren, W.-M., Liu, J., and Lu, X.-B. (2013). Angew. Chem. Int. Ed. 52 (44): 11594–11598. 63 Childers, M.I., Longo, J.M., van Zee, N.J. et al. (2014). Chem. Rev. 114 (16): 8129–8152. 64 Ellis, W.C., Jung, Y., Mulzer, M. et al. (2014). Chem. Sci. 5 (10): 4004. 65 Liu, Y., Ren, W.-M., He, -K.-K., and Lu, X.-B. (2014). Nat. Commun. 5 (1): 5687. 66 Kissling, S., Altenbuchner, P.T., Lehenmeier, M.W. et al. (2015). Chem. Eur. J. 21 (22): 8148–8157. 67 Garden, J.A., Saini, P.K., and Williams, C.K. (2015). J. Am. Chem. Soc. 137 (48): 15078–15081. 68 Darensbourg, D.J. and Fitch, S.B. (2007). Inorg. Chem. 46 (14): 5474–5476. 69 Sugimoto, H. and Kuroda, K. (2008). Macromolecules 41 (2): 312–317. 70 Nakano, K., Nakamura, M., and Nozaki, K. (2009). Macromolecules 42 (18): 6972–6980. 71 Darensbourg, D.J., Lewis, S.J., Rodgers, J.L., and Yarbrough, J.C. (2003). Inorg. Chem. 42 (2): 581–589. 72 Darensbourg, D.J. and Billodeaux, D.B. (2005). Inorg. Chem. 44 (5): 1433–1442. 73 Darensbourg, D.J., Ulusoy, M., Karroonnirum, O. et al. (2009). Macromolecules 42 (18): 6992–6998. 74 Buchard, A., Kember, M.R., Sandeman, K.G., and Williams, C.K. (2011). Chem. Commun. 47 (1): 212–214. 75 Isnard, F., Lamberti, M., Lettieri, L. et al. (2016). Dalton Trans. 45 (40): 16001–16010. 76 Chen, L., Li, W., Yuan, D. et al. (2015). Inorg. Chem. 54 (10): 4699–4708. 77 Gaston, A.J., Greindl, Z., Morrison, C.A., and Garden, J.A. (2021). Inorg. Chem. 60 (4): 2294–2303. 78 Yoshida, M., Masaoka, S., Abe, J., and Sakai, K. (2010). Chem. Asian J. 5 (11): 2369–2378. 79 Kimoto, A., Yamauchi, K., Yoshida, M. et al. (2012). Chem. Commun. 48 (2): 239–241. 80 Yoshida, M., Kondo, M., Torii, S. et al. (2015). Angew. Chem. 127 (27): 8092–8095. 81 Yoshida, M., Kondo, M., Torii, S. et al. (2015). Angew. Chem. Int. Ed. 54 (27): 7981–7984.

References

82 Kundu, A., Dey, S.K., Dey, S. et al. (2020). Inorg. Chem. 59 (2): 1461–1470. 83 El Harakeh, N., de Morais, A.C.P., Rani, N. et al. (2021). Angew. Chem. Int. Ed. 60 (11): 5723–5728. 84 Ghosh, D., Fabry, D.C., Saito, D., and Ishitani, O. (2021). Energy Fuels 35 (23): 19069–19080. 85 Sampaio, R.N., Grills, D.C., Polyansky, D.E. et al. (2020). J. Am. Chem. Soc. 142 (5): 2413–2428. 86 Yamazaki, Y., Ohkubo, K., Saito, D. et al. (2019). Inorg. Chem. 58 (17): 11480–11492. 87 Wang, W.H., Hull, J.F., Muckerman, J.T. et al. (2012). Energy Environ. Sci. 5 (7): 7923–7926.

427

429

20 Metal-Organic Frameworks in Tandem Catalysis Anirban Karmakar* and Armando J.L. Pombeiro Centro de Química Estrutural, Institute of Molecular Sciences, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, Lisboa, Portugal * Corresponding author

20.1  Introduction Catalysis plays a fundamental role in chemical industry. In traditional catalysis, the substrate is activated by a specific catalyst by minimising the reaction energetic barrier [1]. Over the last few decades the concept of multicatalysis, in which multiple reactions can be performed in a single pot using multifunctional catalysts, has evolved rapidly [2, 3]. In this context, tandem or cascade reactions that comprise two or more catalytic reactions performed in a single pot without any purification or separation of the intermediates are relevant [4]. The tandem reactions are often catalysed by two different catalysts or a bifunctional catalyst having multiple active sites. The reactant(s) is (are) activated by an active site of the catalyst to yield an intermediate, which is further activated by the same or another active site present in the catalyst to produce the final product (Figure 20.1) [5, 6]. Thus, waste, time, energy, solvents, and reagents are saved. In the context of sustainable chemistry, the tandem reactions can provide effective and ecologically friendly chemical synthesis methods [7]. The design of multifunctional MOFs to catalyse tandem reactions remains challenging. In fact, to obtain an effective catalytic process, a pathway should be attained in which the catalyst can promote the different reactions in a one-pot process. Moreover, in multifunctional catalysts the active sites, e.g. acidic and basic sites, tend to neutralize each other. Several heterogenous catalysts, such as Pd0/CaCO3, Pd0/BaCO3, and Au NPs supported on ZrO2, Pd-Co NPs on silica, Pd-Os/MgO catalysts, zeolites, covalent organic frameworks (COFs), MOFs, and others have been reported [5, 8– 11]. Among them, MOFs are attracting significant consideration, consistent with their favourable structural properties. The possibility of introducing various acid and basic sites in a single MOF affords an excellent prospect to explore MOFs as heterogenous catalysts in the area of tandem reactions [12]. Thus, recently several functionalized MOFs have been reported that can effectively catalyse various tandem reactions [13].

Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 2, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

430

20  Metal-Organic Frameworks in Tandem Catalysis Metal node

Reactants Active site A Intermediate Active site A

Functional linker

Active site B

Product

Product

(a)

(b)

Guest species

(c)

Figure 20.1  Tandem reactions catalysed by one active site (a) or two different active sites (b). (c) Various active sites on a metal-organic framework (MOF), namely metal nodes, catalytically active guest species, and functional organic linkers. Reproduced with permission from Ref 14 / Royal Society of Chemistry.

In this book chapter, we illustrate functionalized MOFs that have been used as catalysts for various liquid phase tandem reactions, thus contributing to understanding their significance as useful materials in tandem type catalysis. In the first part, favourable structural features of MOFs in this context are discussed. The main content of this chapter is systematized according to the various reported examples of particular tandem reactions catalysed by MOFs. The chapter does not intend to cover all the tandem reactions promoted by functionalized MOFs, but rather to demonstrate some important examples. Lastly, conclusions for future MOFs catalytic applications are specified.

20.2  MOFs as Catalysts for Tandem Reactions MOFs, also known as porous coordination polymers, form a particular class of porous crystalline materials composed of metal ions or clusters and organic linkers [14–17]. Their relevance has been recognized in chemical and materials sciences due to their adaptable architectures, high surface areas, tunable pore sizes, as well as their applications in catalysis, gas storage and separation, sensing, drug delivery, proton conduction, and other techniques [18–20]. Among the applications, MOFs are gaining much consideration as heterogenous catalysts particularly in areas of tandem [21], asymmetric [22, 23], and photocatalysis [24, 25].

20.2.1  Active Sites in MOFs The structures of MOFs can be tuned by changing the metal nodes and ligands, which provides a huge variety of possible compositions and structures [26], with a diversity of properties including concomitant acidity and basicity. Other important features of MOFs include their high surface areas and pore volumes, which permit active guest molecules to be introduced into the pores and allow substrates admittance to the internal active sites [27]. Functional MOFs can catalyse organic reactions via three different types of active sites: (i) open metal centres or clusters; (ii) organic ligands holding different functional organic groups (e.g. amine, amide, urea, thiol, and sulfonate); (iii) catalytically active guest molecules present in the MOF pores [13]. This blend of types of MOF active sites allows MOFs to behave as relevant multifunctional materials for tandem reactions (Figure 20.1b).

20.3  Examples of MOFs Used as Catalysts for Tandem Reactions

20.2.2  Advantages and Limitations of MOFs MOFs own an important place in heterogeneous catalysis offering several advantages, including: (i) easy separation, recovery, and reuse of catalyst with preservation of catalytic activity; (ii) homogenously dispersed active sites which can improve the catalytic efficiency; (iii) tunable pore sizes that can promote regio-selective catalysis; (iv) low formation of by products; and (v) various types of stabilized catalytically guest species, such as NPs, metal complexes, or polyoxometallates, into their pores [28, 29]. The structural stability of MOFs during the catalytic process is an important concern in view of the limited thermal and chemical stability of the metal–ligand bonds. The solvents, high reaction temperature, or certain reagents can cause partial or total destruction of the crystal structures [30, 31]. Moreover, leaching of metal ions can result from a strong binding affinity of reagents or substrates. Another relevant point to be considered is the deactivation of MOF as a catalyst. It can undergo partial or total deactivation after a few recycling steps due to pore blocking, metal leaching, or structure collapse [32]. Thus, a careful survey of MOF stability under the reaction conditions is important. MOF stability after the catalytic process must be monitored by a combination of powder X-ray diffraction analysis and porosity measurements of the reused material. Moreover, the hot filtration test and chemical analysis of the liquid phase can provide significant information about leaching during the catalytic process. It is also recommended to find the reasons for deactivation and suggest a regeneration protocol.

20.3  Examples of MOFs Used as Catalysts for Tandem Reactions MOFs have been used as heterogenous catalysts for various liquid phase organic transformations owing to characteristics such as their insolubility in common organic solvents, thermal and photostability, and good distribution of catalytically active sites [33]. MOFs having suitable metal and ligand systems can also catalyse several types of tandem reactions and examples are discussed in this reaction, addressing their activity, selectivity, and mechanistic roles.

20.3.1  Deacetalization-Knoevenagel Condensation An interesting example of tandem reactions concerns the deacetalization-Knoevenagel condensation, in which the reaction of benzaldehyde dimethyl acetal and malononitrile directly produces benzylidene malononitrile. Different MOFs decorated with acidic and basic groups have ­effectively catalysed such reactions (Scheme 20.1) [34–36].

CN H3C

O

O

O

CH3 NC

H

NC

CN

MOF catalyst Benzaldehyde dimethyl acetal

Benzaldehyde

Benzylidene malononitrile

Scheme 20.1  Tandem deacetalization-Knoevenagel condensation reactions of benzaldehyde dimethyl acetal and malononitrile. Ref [35] / Royal Society of Chemistry.

431

432

20  Metal-Organic Frameworks in Tandem Catalysis

For example, Zhou et al. reported a Cu(II) MOF formulated as [Cu2(L1)(H2O)]n (1) using a ­pyridine based amido carboxylate pro-ligand, 5,5ʹ-[(pyridine-3,5-dicarbonyl)bis-(azanediyl)] diisophthalate (H4L1) (Figure 20.2a), and this framework shows an efficient catalytic activity for the tandem one-pot deacetalization-Knoevenagel condensation reaction of benzaldehyde dimethyl acetal and malononitrile [34]. By using 0.5 mol% of 1 as catalyst for a such a reaction, 100% yield of benzylidene malononitrile was reached within 12 hours at 50 oC. Moreover, they have recycled the catalyst for three cycles and only a slight decrease in reaction yield (from 100% to 92%) was observed. Powder x-ray diffraction analysis before and after the catalysis did not show any alteration in the peaks pattern, confirming catalyst stability under the reaction conditions. The authors have also proved that the presence of both the Cu(II) sites and the amide groups from the ligands is crucial for this reaction. Addition of p-toluenesulfonic acid facilitates the deacetalization reaction but hampers the Knoevenagel condensation, whereas the addition of ethylenediamine (blocking the Cu(II) sites) only promotes the Knoevenagel condensation and hampers the deacetalization reaction, decreasing the overall yield of benzylidene malononitrile. Recently, we have synthesized a two dimensional MOF, [Zn5(L2)4(OH)2(H2O)4]n.8n(DMF).4n(H2O), in which DMF is dimethyl formamide, (2), by the solvothermal reactions between the amide functionalized dicarboxylic acid 4,4ʹ-{(pyridine-2,6-dicarbonyl)bis(azanediyl)}dibenzoic acid (H2L2) and zinc(II) nitrate (Figure 20.2b) [35]. This MOF heterogeneously catalyses the tandem deacetalization-Knoevenagel condensation reactions carried out under conventional heating, microwave, or ultrasonic irradiation. Comparative studies show that ultrasonic irradiation (yield of 98% after two hours) provides the most favourable method compared to microwave or normal heating (yield of 91–93% after three hours). Moreover, the catalysts can be reused at least for five consecutive cycles without losing activity significantly.

Figure 20.2  Synthesis and crystal structures of metal-organic frameworks (MOFs) 1 (a) and 2 (b). Adapted from Refs [34] and [35].

20.3  Examples of MOFs Used as Catalysts for Tandem Reactions

Figure 20.3  Synthesis and crystal structures of metal-organic frameworks (MOFs) 3 (A) and 4 (B). Ref [36] / Frontiers Media S.A.

In another work, we have reported the pyridine based amide functionalized tetracarboxylic acid 5,5ʹ-{(pyridine-2,6-dicarbonyl)bis(azanediyl)}diisophthalic acid (H4L3) and its coordination chemistry with Zn(II) and Cd(II) ions. The reactions of H4L3 with Zn(II) and Cd(II) nitrates led to the formation of the two 2D MOFs [Zn2(L3)(H2O)4]n.4n(H2O) (3) and [Cd3(HL3)2(DMF)4]n.4n(DMF) (4), respectively (Figure 20.3) [36]. On account of the presence of Lewis acid (Zn or Cd centers) and basic (uncoordinated pyridine and amide groups) sites, 3 effectively (yield of 99% after 3 h) catalyses the one-pot cascade deacetalization-Knoevenagel condensation reactions, under considerably mild conditions. These MOFs act as heterogeneous catalysts and can be recycled a few times without losing their activity. The stability of these MOFs after the catalytic reactions were proved by powder X-ray diffraction and FT-IR analysis.

20.3.2  Deacetalization-Henry Reaction Deacetalization-Henry reactions are another important type of tandem reactions which involve the conversion of an acetal to aldehyde (first step) followed by the Henry or nitroaldol reaction between the intermediate aldehyde and nitromethane or nitroethane, leading to the formation of a nitroalkene as the final product (Scheme 20.2).

Scheme 20.2  Example of deacetalization-Henry tandem reactions of acetal and nitroalkane. Adapted from Ref [37].

In 2012, Li and co-workers prepared a Cr(III) MOF, formulated as [Cr3(F)(O)(H2O)2(BDC)3]n (MIL-101-Cr) (BDC = 1,4-benzene dicarboxylate), by solvothermal reaction between Cr(NO3)3·9H2O and 1,4-benzene dicarboxylic acid. Later, they prepared a bifunctionalized MOF, namely MIL-101SO3H-NH2 (5), by introducing sulfonic acid (SO3H–) and ethylenediamine (NH2CH2CH2NH2) groups into the MIL-101-Cr MOF via post-synthetic modification methods (Scheme 20.3) [37]. They tested the catalytic activity of MIL-101-SO3H-NH2 (5) towards the one-pot deacetalizationHenry tandem reactions of benzaldehyde dimethyl acetal and nitromethane, achieving 97% yield

433

434

20  Metal-Organic Frameworks in Tandem Catalysis

Cr O

Cr O

Cr O H2N

Cr O

NHBOC

O Cr

NHBOC

Cr O

ClSO3H

O Cr

O Cr

Cr O

N H2

SO3H

CHCl3

Benzene O Cr

N H2

O Cr

O Cr

MIL-101-Cr

NHBOC

Cr O

165 oC, benzene

Cr O

N H2

NH2

SO3H

MW O Cr

O Cr

MIL-101-SO3H-NH2 (5)

Scheme 20.3  Synthesis of MIL-101-SO3H-NH2 (5) via post-synthetic modification of MOF-101-Cr (BOC = tert-butyloxycarbonyl protecting group). Adapted from Ref [37].

of 2-nitrovinyl benzene as the final product after 24 h of reaction time at 90 oC. Upon performing the reaction with monofunctionalized MIL-101-SO3H and MIL-101-NH2 (only one of the two essential functional groups of the ligand is present), no such final product was detected. This experiment demonstrates that the presence of both sulfonic acid and amine groups is necessary to promote the abovementioned tandem catalytic process. Moreover, the use of catalyst 5 with an excess amount of acid (p-toluene sulfonic acid) or base (ethylamine) would inhibit the activity probably due to the formation of ion pairs. They have also reused the catalyst for three times and no changes in the framework crystallinity or structure were observed by means of powder X-ray diffraction analysis.

20.3.3  Meinwald rearrangement-Knoevenagel Condensation Meinwald rearrangement concerns the conversion of epoxides into aldehydes or ketones by ringopening and 1,2-shift of an hydride or alkyl group, whereas Knoevenagel condensation is the condensation reaction between aldehydes or ketones with nitriles to form benzylidene malononitriles (Scheme 20.4).

Scheme 20.4  Example of Meinwald rearrangement-Knoevenagel condensation tandem reactions. Adapted from Ref [38].

In 2012, Kim et al. reported a 3D Al(III) MOF, [NH2-MIL-101(Al), 6] formulated as [Al(OH)(BDCNH2)]n (BDC-NH2 = 2-aminoterephthalate) (Figure 20.4), which effectively catalyses the tandem Meinwald rearrangement-Knoevenagel condensation reactions. The tandem reaction between 2-methyl-2-phenyloxirane and malononitrile in the presence of catalyst 6 at 60 oC resulted in 70% yield of the final product 2-(2-phenylpropylidene)malononitrile within 48 h of reaction time [38]. The Al(III) metal centres can act as Lewis acid catalytic sites and catalyse the first step (Meinwald rearrangement) of the tandem reactions, which is ring opening of 2-methyl-2-phenyloxirane to generate the intermediate 2-phenylpropanal [38]. The second step (Knoevenagel condensation) is

20.3  Examples of MOFs Used as Catalysts for Tandem Reactions

Figure 20.4  Crystal structure and trinuclear Al(III) secondary building unit (SBU) in NH2-MIL-101(Al) (6). Adapted from Ref [38].

mostly promoted by free amino (-NH2) moieties of the 2-amino terephthalate which act as Brønsted base. Upon performing the tandem reaction of 2-methyl-2-phenyloxirane and malononitrile in the presence of AlCl3 or a mixture of AlCl3 and 2-aminoterephthalate as a catalyst, the Meinwald rearrangement (first step) occurred with 85% of conversion, but the Knoevenagel condensation failed. This failure may be to the lack of free amine groups of 2-aminoterephthalate available for catalysis. However, the reaction of 2-phenylpropanal and malononitrile using 2-aminoterephthalate as catalyst produces 65% of 2-(2-phenylpropylidene)malononitrile. The results confirm that the presence of both Al(III) sites and free amine (-NH2) group is essential to promote such tandem reactions. Moreover, hot filtration experiments established that the tandem epoxide ring opening and Knoevenagel condensation reaction happened in heterogeneous medium and powder X-ray diffraction analysis indicated that the structural integrity of the catalyst 6 remained unaltered after two reaction cycles.

20.3.4  Reductive Amination of Aldehydes with Nitroarenes An interesting example of one-pot three step tandem reactions is the reductive amination of aldehydes with nitroarenes or hydrogenation-reductive amination. This overall reaction involves three steps, firstly the chemoselective reduction of the nitro compound to the corresponding amine in presence of H2 and catalyst, followed by the condensation between the aromatic amine and the carbonyl group of aldehyde to produce an imine compound, and lastly the hydrogenation of the resulting imine to produce the corresponding secondary amine (Scheme 20.5).

Scheme 20.5  Example of tandem hydrogenation-reductive amination reactions. Ref [39] / Elsevier.

435

436

20  Metal-Organic Frameworks in Tandem Catalysis

Scheme 20.6  Synthesis of the Ir(II) complex bearing metal-organic framework (MOF) 7a via post-synthetic modification of the Zr(IV) MOF (7). Adapted from Ref [39].

Corma et al. prepared a Zr(IV) MOF, formulated as [Zr6O4(OH)4(BDC-NH2)6]n (7) (BDC-NH2 = 2-aminoterephthalate), by the solvothermal reaction of a Zr(IV) salt and 2-aminoterephthalic acid. Later, they incorporated an iridium complex into the MOF 7 via post-synthetic modification as shown in Scheme 20.6 to produce an Ir(II) complex bearing MOF (7a) able to catalyse hydrogenation-reductive amination tandem reaction [39]. Performing the reaction of benzaldehyde and nitrobenzene at 100 oC, in the presence of isopropanol, H2 (6 bar) and MOF catalyst 7a, resulted in 99% conversion of nitrobenzene to the secondary amine with a higher selectivity (99%) towards the amine compared to the imine product after 24 h of reaction time [39]. In this reaction, the nitrobenzene was reduced to the aniline in the presence of the iridium site and hydrogen. Then, the aniline and the benzaldehyde formed an imine via a condensation reaction which was mostly catalysed by the Zr(IV) acid site of 7a. The final step, conversion of imine to the secondary amine by H2 was also catalysed by Ir sites. Moreover, the authors have also used various substituted aromatic and aliphatic aldehydes, as well as substituted nitrobenzene, and achieved 90–100% conversion to the corresponding products. They have recycled the catalyst 7a for four times and it maintained catalytic activity. Hot filtration test, powder x-ray diffraction and SEM analyses suggest that no leaching occurred during the catalytic process and the overall structure of the catalyst remained intact.

20.3.5  Epoxidation–Ring Opening of Epoxide Farha et al. reported an Hf(IV) MOF, formulated as [Hf6(O)4(OH)4(FeTCP-Cl)3]n (8), prepared by the solvothermal reaction of HfOCl2, meso-tetra(4-carboxyphenyl)-porphyrin-Fe(III) chloride (H4FeTCP-Cl), and benzoic acid as a modulator (Figure 20.5). Later, to ensure every porphyrin was metalated the synthesized MOF 8 was suspended in a DMF solution containing anhydrous FeCl3 and surprisingly a new material 8a was formed [40]. In MOF 8a, the Fe(III) ions from anhydrous FeCl3 were coordinated to the hydroxy (–OH) and water ligands, which was confirmed by single-crystal X-ray diffraction analysis and diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS). They have tested the catalytic activity of MOF 8a towards the tandem styrene epoxidation–ring opening of epoxide reaction. The reaction of styrene with trimethylsilyl azide (TMSN3) in presence of tert-butyraldehyde (to regenerate the catalyst), O2 (5 atm) and MOF 8a (as catalyst) in acetylnitrile (MeCN) medium at 60 oC led to formation of a protected 1,2-hydroxylamine within 10 h of reaction time (Scheme 20.7) [40]. The first step of this tandem reaction which is epoxidation of styrene to produce styrene oxide intermediate was mainly catalysed by the Fe(III) sites of 8a, and the reaction of styrene epoxide with TMSN3 was catalysed by the Lewis acid Hf(IV) nodes of 8a. By attempting to perform the

20.3  Examples of MOFs Used as Catalysts for Tandem Reactions

Figure 20.5  Structure of metal-organic framework (MOF) 8 and its [Hf6(O)4(OH)4] secondary building unit (SBU) and FeTCP-Cl ligand. Ref [40] / American Chemical Society.

Scheme 20.7  Example of tandem epoxidation of styrene and ring opening of 2-phenyloxirane. Adapted from Ref [40].

reaction with MOF 8, no final product was obtained. Moreover, no leaching of the heterogenous 8a catalyst was observed during the catalytic reaction.

20.3.6  Oxidation-Esterification Incorporation of metal NPs into MOFs pores can result in a catalytically more active system [41]. To study the catalytic activity of NPs incorporated MOFs, Fischer and co-workers prepared two zeolite imidazolate frameworks, namely [Zn(MeIM)2]n (9) (MeIM = imidazolate-2-methyl) and [Zn(ICA)2]n (10) (ICA = imidazolate-2-carboxyaldehyde), and embedded them with Au NPs to produce MOFs 9Au and 10Au, respectively [41]. They then studied the catalytic activities of these Au NPs encapsulated MOFs towards the tandem oxidation-esterification reactions of benzyl alcohol to methyl benzoate in the presence of methanol and O2 (Scheme 20.8) [41]. The loading of Au NPs was varied between 5 and 30 wt.% and they were distributed throughout the MOFs with particle sizes ranging from 1 to 5 nm. Upon testing the catalytic activity of MOFs 9 and 10 (without incorporated Au NPs) towards the tandem

437

438

20  Metal-Organic Frameworks in Tandem Catalysis

Scheme 20.8  Tandem oxidation-esterification reactions of benzyl alcohol to methyl benzoate using Au nanoparticles (NPs) supported metal-organic frameworks (MOFs) (9 and 10). Adapted from Ref [41].

oxidation-esterification reaction of benzyl alcohol to methyl benzoate, no activity was observed. However, by using the Au NPs incorporated MOFs 9Au and 10Au as catalysts, a conversion of 81 and 13% with a selectivity towards methyl benzoate of 98% and 50%, respectively, was obtained. Moreover, small amounts of benzaldehyde were formed during this tandem reaction, but no benzaldehyde dimethyl acetal was detected. The same research group also reported metal oxide and nanoparticle encapsulated MOFs for the tandem oxidation-esterification reaction of benzyl alcohol to methyl benzoate [42]. Firstly they prepared a Zn(II) MOF, formulated as [Zn4O(BDC)3]n (11), and loaded it with ZnO and TiO2, and prepared ZnO@11 and TiO2@11, respectively. The reaction of 11, ZnO@11 and TiO2@11 with [ClAuCO] produced intermediate materials denoted as [ClAuCO]@11, [ClAuCO]/ZnO@11 and [ClAuCO]/TiO2@11. These composites were decomposed at 100 °C under hydrogen and produced Au@11, Au/ZnO@11 and Au/TiO2@11. The characterization of these materials revealed an homogeneous distribution of Au NPs (size range within 1–3 nm) over the MOF. The catalytic activity of these materials was then tested towards liquid-phase oxidation-esterification of benzyl alcohol to methyl benzoate in the presence of MeOH and O2. Unlike the previous case, these materials did not show any activity without the presence of a base. However, upon performing the reaction in the presence of K2CO3 as base, Au@11 displayed 50% conversion of benzyl alcohol with 48% yield of methyl benzoate in 30 min. The presence of metal oxide improved the benzyl alcohol conversion, as well as the selectivity to the methyl ester. For example, in the presence of K2CO3 base, Au/ZnO@11 and Au/TiO2@11 resulted in 66 and 72% selectivity, respectively, towards methyl benzoate with trace amounts of benzaldehyde and no formation of benzoic acid.

20.3.7  Oxidation-Hemiacetal Reaction Mixed NPs encapsulated MOFs have already emerged in heterogenous tandem catalysis [43]. For example, Luque et al. prepared a Au–Pd NPs supported MOF (12) by immersing a Cr(III) MOF, formulated as [Cr3(O)(F)(H2O)2(BDC)3]n.25nH2O (BDC = 1,4-benzenedicarboxylate), into a solution containing Au(III) and Pd(II) NPs. They have tested its activity in the aerobic oxidation-hemiacetal tandem reactions of toluene and its derivatives to conversion into the corresponding esters [43]. The solvent-free oxidation-hemiacetal reaction of toluene was carried out in the presence of 1 wt% of catalyst 12 (Au:Pd ratio of 1:1.5) at 120  oC and under 1.0 MPa of O2 and after 48 h of

20.3  Examples of MOFs Used as Catalysts for Tandem Reactions

Scheme 20.9  Preparation of benzyl benzoate from toluene via tandem oxidation-aldol condensation catalysed by Au–Pd nanoparticles (NPs) supported metal-organic framework (MOF) (12). Adapted from Ref [43].

reaction time 98.6% of toluene conversion with 93% selectivity for benzyl benzoate was achieved (Scheme 20.9). Moreover, only 2 and 3% of benzaldehyde and phenyl benzoate were detected, respectively. They have also tested this reaction without any metal catalyst and with the parent Cr(III)-MOF as catalyst and almost no conversion was achieved in both cases. The oxidation-hemiacetal tandem reactions of other substrates, such as o-, m-, and p-xylenes; 2-, 3-, 4-methoxytoluene; and 4-fluorotoluene, were also tested using 12 as a catalyst, and the system was also efficient [43]. For p-xylene, 40.2% conversion (TON value of 1206) with 65% selectivity for benzyl benzoate was obtained by performing the reaction at 120 oC for 48 h. The oxidation of substituted toluene resulted in the development of the corresponding esters as the major products with an overall selectivity within the range of 66–71% without formation of acid or CO2. The catalyst 12 was stable under the reaction conditions and preserved its structure after three catalytic cycles as demonstrated by powder X-ray diffraction. This tandem reaction is mostly catalysed by the Au-Pd NPs and proceeds through the oxidative conversion of toluene to benzyl alcohol and benzaldehyde, followed by the formation of the hemiacetal [PhCH(OH)OCH2Ph] via reaction of these intermediates, which finally is further oxidized to benzyl benzoate.

20.3.8  Asymmetric Tandem Reactions In this context, the design and synthesis of MOF based catalysts for enantioselective tandem catalysis is a further challenging task. Homochiral MOFs containing both acid and basic sites have been engaged in multi-component tandem asymmetric synthesis [44]. For example, Duan et al. synthesized two homochiral enantiomorphs MOFs (13a and 13b), formulated as [ZnW12O40].[Zn2(NH2-BPY)2(HPYI)2(H2O)(CH3CN)], by the combination of the Keggin polyoxometalate [ZnW12O40]6− anion, NH2-BPY (3-amino-4,4ʹ-bipyridine), Zn(II) ions and L- or D-pyrrolidine-2-yl-imidazole (PYI) (Figure 20.6). By using these homochiral MOFs they have performed olefin epoxidation followed by the conversion of epoxide into cyclic carbonate using CO2, and obtained good yield and selectivity [44]. The reaction of styrene, tert-butyl hydroperoxide (TBHP), and CO2 in the presence catalyst 13a at 50  oC led to the formation of (R)-phenyl(ethylene carbonate) with 92% yield and 80% enantiomeric excess (ee) within 96 h of reaction time (Scheme 20.10) [44]. By using 13b as catalyst,

439

440

20  Metal-Organic Frameworks in Tandem Catalysis

Figure 20.6  Synthesis (a) and crystal structure (b) of homochiral enantiomorphs metal-organic frameworks (MOFs) (13a and 13b). Ref [44] / Springer Nature.

Scheme 20.10  Example of asymmetric tandem reactions catalysed by MOFs 13a or 13b. Adapted from Ref [44].

(S)-phenyl(ethylene carbonate) was obtained as the final product with similar yield and enantiomeric excess (ee). The asymmetric epoxidation of styrene (first step) is mostly catalysed by the [ZnW12O40]6– Keggin anion and the PYI ligand. But the second step of this tandem reaction (asymmetric cyclic carbonate formation) is primarily catalysed via Lewis acidic Zn(II) ions present in the MOF 13a or 13b, which interact synergistically and activate CO2 insertion in view of the spatial location. Moreover, the amine (-NH2) group not only acts as a basic site but also increases the CO2 adsorption and promotes the formation of the cyclic carbonate product. The authors also tested the reactivity of their catalysts towards various substituted styrene derivatives and obtained good product yields (72–83%) and selectivities (55–70%). The removal of catalyst from the reaction mixture after 48 h stopped the reaction and no additional conversion was observed. The catalysts were reused at least three cycles with a moderate loss of activity (from 92 to 88% yield) and a small decrease in selectivity (from 80 to 77% ee). The powder X-ray diffraction analyses before and after catalysis indicate that the MOF catalysts remain stable during the catalytic process.

20.3.9  Photocatalytic Tandem Reactions Photocatalysis is an important type of catalysis in which light can be used as an energy source to accomplish organic transformations under mild conditions. In this context the photocatalytic tandem reactions, in particular with MOF-based catalysts, deserve a particular attention [45, 46]. Amine-functionalized ligand-based MOFs coupled with photoactive metal nodes have been widely used for photocatalytic tandem reactions. For example, Matsuoka et al. have used the amine functionalized Zr(IV) MOF [Zr6O4(OH)4(BDC-NH2)6]n (7) (BDC-NH2 = 2-aminoterephthalate) for one-pot tandem reactions between benzyl alcohol and malononitrile to produce benzylidene malononitrile

20.4  Conclusions

Scheme 20.11  Example of photocatalytic tandem reactions catalysed by metal-organic frameworks (MOFs) 7 and 14. Adapted from Ref [45].

(Scheme 20.11) [45]. This overall conversion comprises the photocatalytic oxidation of benzyl alcohol to benzaldehyde, followed by the Knoevenagel condensation of the latter with malononitrile to produce an α,β-unsaturated benzylidene malononitrile. The reaction of benzyl alcohol with an excess amount of malononitrile in the presence of MOF 7 as catalyst under UV-light irradiation at 363 K yielded 91% benzylidene malononitrile after 48 h. The reaction did not procced without UV-irradiation and the use of a lower temperature (298 K) only promotes the first step of the tandem reaction (photocatalytic oxidation of benzyl alcohol) leading to the formation of benzaldehyde as major product (yield 89%). Moreover, the use of a Zr(IV)-benzenedicarboxylate MOF without the amine group or the test of ZrO2 as catalyst for the above mentioned reaction only produces 28% or 2% of final product, respectively, suggesting a significant role of the amine (-NH2) moiety in the reaction. The authors have also tested the activities of other amine containing MOFs e.g. the Zn-MOF-NH2 [Zn4O(BDC-NH2)3]n and the Ti-MOF-NH2 [Ti8O8(OH)4(BDC-NH2)6]n and their catalytic activities are lower than that of MOF 7, suggesting a higher catalytic effect of Zr(IV) compared to the other metals. The recyclability of MOF 7 was tested and a slight decrease in catalytic activity was observed. Li and co-workers employed an Fe-based MOF, [Fe3(O)(Cl)(BDC-NH2)3]n (14) (BDC-NH2 = 2-aminoterephthalate), as the catalyst for the tandem photooxidation-Knoevenagel condensation in the presence of O2 and visible light at room temperature [46]. On account of extensive iron oxoclusters as well as the amine (-NH2) group in MOF 14, it shows intensive light absorption in the visible light region. In accord, MOF 14 efficiently catalyses the one-pot tandem reaction between benzyl alcohol and malononitrile via tandem photooxidation and base-catalysed Knoevenagel condensation, achieving 72% yield of benzylidene malononitrile after 40 h. As in the above earlier case, the reaction did not take place without catalyst or visible light irradiation. The authors also tested the catalytic activities of MOF 14 towards a variety of substituted aromatic alcohols and active methylene compounds and obtained product yields in the 20–76% range. The heterogeneous nature of the catalyst 14 was confirmed by a hot filtration experiment. Moreover, the catalyst can be recycled at least for three times without obvious reduction of the activity. Powder X-ray diffraction analyses, FT-IR and N2 adsorption analyses after the catalytic reaction did not show significant changes, suggesting the stability of MOF 14 during the tandem reaction.

20.4  Conclusions In recent decades, the use of MOFs as heterogeneous catalysts for different liquid-phase organic catalytic reactions, including tandem ones, has improved markedly, as illustrated in this chapter. MOFs can catalyse an organic reaction via different active sites. In most of the cases, the unsaturated metal nodes (Lewis acidic centres) or the functional active sites on the organic linkers (acid/ base sites) are involved in the catalytic transformations. But in some cases the catalytically active NPs or metal complexes incorporated in the MOFs pores can also catalyse the organic reactions. Two main different methods have been used to incorporate the active sites into MOFs: (i) the direct

441

442

20  Metal-Organic Frameworks in Tandem Catalysis

method and (ii) the post-synthetic method. The direct method is simple and easy to incorporate different types of functional ligands into the MOF, but provides only a low control on the final structure and porosity. In contrast, the post-synthetic method can afford a good control on the structure and porosity of the MOFs, but its methodology is not easy. Due to the straightforwardness of the direct method, it has been extensively used for the synthesis of various catalytically active MOFs, such as Cu-MOF (1), Zn-MOFs (2, 3 and 13), Cd-MOFs (4), Al-MOF (6), and Hf-MOF (8), systems. However, in some cases, the incorporation of catalytically active NPs and metal complexes into MOFs sites were done via post-synthetic modifications. For example, an Ir(II) complex was incorporated in the Zr-MOF (7) by the post-synthetic modification method, being hard to obtain using the direct method. Moreover, in the Cr-MOF (5), the basic sites (ethylene diamine) and acidic sites (sulfonic acid) were also introduced via a post-synthetic modification process. These functionalized MOFs can act as effective catalysts for different types of tandem reactions including deacetalization-Knoevenagel condensation, deacetalization-Henry reaction, Meinwald rearrangement-Knoevenagel condensation, nitroarene reductive N-alkylation with aldehydes, epoxidation–epoxide ring-opening, and oxidation-esterification. High yields and selectivities can be achieved owing to the synergistic effect of metal sites, functional organic groups or encapsulated NPs or complex molecules within the structure. For example, MOF 1 catalyses the tandem deacetalization-Knoevenagel condensation where the deacetalization reaction is mostly catalysed by the Lewis acidic Cu(II) metal sites and the Knoevenagel condensation is promoted by a basic amide group present in the ligand. But in some tandem reactions the catalytic mechanisms are not clear and thus further characterization methods accompanied by theoretical calculations are needed for such a purpose. Although important developments have been achieved in different tandem reactions catalysed by MOFs, their role in asymmetric and photocatalysis is still limited. A few of such MOFs (7, 13a, 13b and 14) have been used for such a purpose, but, the product yield and selectivity (enantiomeric excess) are usually not remarkable. Thus, the use of MOF based catalysts in asymmetric and in photocatalysis deserves to be further explored. Although hundreds of MOFs have already been reported as catalysts for a significant number of tandem reactions, the practical industrial applications of MOFs materials are still very limited. Some of the MOFs display high porosity and stability, but significant attention is required to produce high yielded and low-cost MOFs in an industrial scale. It is anticipated that this area will expand and MOFs will be used as robust catalysts for various tandem organic reactions of industrial significance.

Acknowledgements This work was supported by the Fundação para a Ciência e Tecnologia (FCT), Portugal, projects UIDB/00100/2020 and UIDP/00100/2020 of Centro de Química Estrutural and LA/P/0056/2020. A. Karmakar expresses his gratitude to Instituto Superior Técnico and FCT for Scientific Employment contract (Contrato No: IST-ID/107/2018) under Decree-Law no. 57/2016, of August 29.

References 1 Climent, M.J., Corma, A., and Iborra, S. (2011). Chem. Rev. 111: 1072–1133. 2 Walsh, C.T. and Moore, B.S. (2019). Angew. Chem. Int. Ed. 58: 6846–6879. 3 Volla, C.M.R., Atodiresei, I., and Rueping, M. (2014). Chem. Rev. 114: 2390–2431.

References

4 Felpin, F.-X. and Fouquet, E. (2008). ChemSusChem 1: 718–724. 5 Wasilke, J.C., Obrey, S.J., Baker, R.T., and Bazan, G.C. (2005). Chem. Rev. 105: 1001–1020. 6 Chng, L.L., Erathodiyil, N., and Ying, J.Y. (2013). Acc. Chem. Res. 46: 1825–1837. 7 Liu, J., Chen, L., Cui, H. et al. (2014). Chem. Soc. Rev. 43: 6011–6061. 8 Islamoglu, T., Goswami, S., Li, Z. et al. (2017). Acc. Chem. Res. 50: 805–813. 9 Climent, M.J., Corma, A., Iborra, S., and Sabater, M.J. (2014). ACS Catal 4: 870–891. 10 Paul, A., Karmakar, A., Guedes da Silva, M.F.C., and Pombeiro, A.J.L. (2021). Catalysts 11: 90. 11 Dhakshinamoorthy, A. and Garcia, H. (2014). ChemSusChem 7: 2392–2410. 12 Zhang, Y., Huang, C., and Mi, L. (2020). Dalton Trans. 49: 14723–14730. 13 Huang, Y.-B., Liang, J., Wang, X.-S., and Cao, R. (2017). Chem. Soc. Rev. 46: 126–157. 14 Karmakar, A. and Pombeiro, A.J.L. (2019). Coord. Chem. Rev. 395: 86–129. 15 Karmakar, A., Guedes da Silva, M.F.C., and Pombeiro, A.J.L. (2014). Dalton Trans. 43: 7795–7810. 16 Tăbăcaru, A., Pettinari, C., and Galli, S. (2018). Coord. Chem. Rev. 372: 1–30. 17 Karmakar, A., Hazra, S., and Pombeiro, A.J.L. (2022). Coord. Chem. Rev. 453: 214314. 18 Pettinari, C., Pettinari, R., Nicola, C.D. et al. (2021). Coord. Chem. Rev. 446: 214121. 19 Karmakar, A., Paul, A., and Pombeiro, A.J.L. (2017). CrystEngComm 19: 4666–4695. 20 Kou, J., Lu, C., Wang, J. et al. (2017). Chem. Rev. 117: 1445–1514. 21 Karmakar, A., Soliman, M.M.A., Alegria, E.C.B.A. et al. (2022). Catalysts 12: 294. 22 Dybtsev, D.N. and Bryliakov, K.P. (2021). Coord. Chem. Rev. 437: 213845. 23 Mo, K., Yang, Y., and Cui, Y. (2014). J. Am. Chem. Soc. 136: 1746–1749. 24 Li, Y., Xu, H., Ouyang, S., and Ye, J. (2016). Phys. Chem. Chem. Phys. 18: 7563–7572. 25 Xiao, J.-D. and Jiang, H.-L. (2019). Acc. Chem. Res. 52: 356–366. 26 Karmakar, A., Titi, H.M., and Goldberg, I. (2011). Cryst. Growth Des. 11: 2621–2636. 27 Dhakshinamoorthy, A., Asiri, A.M., and Garcia, H. (2016). Chem. Eur. J. 22: 8012–8024. 28 Kang, Y.-S., Lu, Y., Chen, K. et al. (2019). Coord. Chem. Rev. 378: 262–280. 29 Gascon, J., Corma, A., Kapteijn, F., and Llabrés i Xamena, F.X. (2014). ACS Catal. 4: 361–378. 30 Karmakar, A., Martins, L.M.D.R.S., Hazra, S. et al. (2016). Cryst. Growth Des. 16: 1837–1849. 31 Karmakar, A., Rúbio, G.M.D.M., Guedes da Silva, M.F.C. et al. (2016). RSC Adv. 6: 89007–89018. 32 Conley, E.T. and Gates, B.C. (2022). Chem. Mater. 34: 3395–3408. 33 Wang, Q. and Astruc, D. (2020). Chem. Rev. 120: 1438–1511. 34 Park, J., Li, J.-R., Chen, Y.-P. et al. (2012). Chem. Commun. 48: 9995–9997. 35 Karmakar, A., Soliman, M.M.A., Rúbio, G.M.D.M. et al. (2020). Dalton Trans. 49: 8075–8085. 36 Karmakar, A., Paul, A., Rúbio, G.M.D.M. et al. (2019). Front. Chem. 7: 699. 37 Li, B., Zhang, Y., Ma, D. et al. (2012). Chem. Commun. 48: 6151–6153. 38 Srirambalaji, R., Hong, S., Natarajan, R. et al. (2012). Chem. Commun. 48: 11650–11652. 39 Pintado-Sierra, M., Rasero-Almansa, A.M., Corma, A. et al. (2013). J. Catal. 299: 137–145. 40 Beyzavi, H., Vermeulen, N.A., Howarth, A.J. et al. (2015). J. Am. Chem. Soc. 137: 13624–13631. 41 Esken, D., Turner, S., Lebedev, O.I. et al. (2010). Chem. Mater. 22: 6393–6401. 42 Müller, M., Turner, S., Lebedev, O.I. et al. (2011). Eur. J. Inorg. Chem. 1876–1887. 43 Liu, H., Li, Y., Jiang, H. et al. (2012). Chem. Commun. 48: 8431–8433. 44 Han, Q., Qi, B., Ren, W. et al. (2015). Nat Commun 6: 10007. 45 Toyao, T., Saito, M., Horiuchi, Y., and Matsuoka, M. (2014). Catal. Sci. Technol. 4: 625–628. 46 Wang, D. and Li, Z. (2015). Catal. Sci. Technol. 5: 1623–1628.

443

445

21 (Tetracarboxylate)bridged-di-transition Metal Complexes and Factors Impacting Their Carbene Transfer Reactivity LiPing Xu1,2, Adrian Varela-Alvarez1, and Djamaladdin G. Musaev1 1 Cherry L. Emerson Center for Scientific Computation and Department of Chemistry, Emory University, 1515 Dickey Drive, Atlanta, Georgia, USA 2 School of Chemistry and Chemical Engineering, Shandong University of Technology, Zibo, China

21.1  Introduction Multi-transition metal complexes have emerged as important thermal and photocatalysts in numerous vital synthetic and biological processes, as effective photosensitizers, and as vital multifunctional materials. For example, bourgeoning research has identified tetra-bridged paddlewheeldi-Rh, and di-Ru complexes as effective catalysts for C–H and C–C π-bond alkylation (by utilizing diazocarbenes N2CR1R2 as a carbene source) [1–12] and amination (which operates best with sulfamate-derived nitrene species (NSO3R)) (Figure 21.1) [13]. It has been demonstrated that metal centers stabilize reactive carbenes (and nitrenes) for consequent insertion into activated bonds (Figure 21.2) in a controllable and highly predictable manner during carbene (and nitrene) transfer reactions [14–18]. Some of the reported tetra-bridged ­paddlewheel-di-metallic complexes, such as the (tetracarboxylate)-di-Rh complexes, also have emerged as photosensitizers that adsorb visible light to initiate charge separation and charge transfer processes [19–21]. In many cases, these complexes act as the building blocks of multifunctional porous metal-organic frameworks (MOFs) [22]. These and other important features and applications of the tetra-bridged paddlewheel-di-metallic complexes have made them a target of numerous fundamental investigations. The existing extensive experimental and computational studies have emphasized the stability of these species, nature of the bridging ligands, and tunability of the redox potentials of metal centers as the success-limiting aspects [23–35]. To establish the factors impacting the stability and reactivity of these species under the catalytic conditions, the atomistic-level understanding of mechanisms of the targeted processes, roles of the transition metals (M) and their electronic structures, and roles of bridging and auxiliary ligands (L) have been shown to be important. Knowledge of these aspects of the tetra-bridged paddlewheel-dimetallic complexes is critical to improve their efficacy and broader utilization. Currently, the literature includes multiple high impact investigations of the tetra-bridged paddlewheel-[Rh2] and –[Ru2] complexes [7, 8, 12–18, 23–35]. However, the development of inexpensive analogs with the earth abundant first-row transition metals (such as Co, Ni, and Cu) requires additional research.

Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 2, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

446

21  (Tetracarboxylate)bridged-di-transition Metal Complexes and Factors Impacting Their Carbene Transfer Reactivity

Figure 21.1  Schematic representation of the broadly used tetra-bridged paddlewheel-di-Rh catalysts in this study.

Figure 21.2  Schematic representation of the diazocarbene (N2CR1R2) decomposition by the (tetracarboxylate)-bridged-di-Rh catalysts, and following carbene transfer to C–H bond.

Therefore, this chapter aims to discuss electronic structures of the (RCOO)4-[M2] complexes, and their carbene derivatives {(RCOO)4-[M2]}=CR1R2, where M = Co, Ni, Cu, Rh, Pd, and Ag; R = Me, Ph and CF3; R1 = p-BrPh; and R2 = COOMe. We anticipate that the comparison of the obtained data for different Ms and Rs will enable us to identify impact of the nature of the metal-centers and R-substituents of the bridging carboxylates on the stability and electronic structure of these complexes, as well as thermodynamic properties of diazocarbene decomposition (i.e. the reaction of N2=CR2 + {(RCOO)4-[M2]} → {(RCOO)4-[M2]}=CR1R2 + N2) by these complexes.

21.3  Results and Discussion

21.2  Computational Procedure In this paper, we used the B3LYP density functional approximation [36–38] with Grimme’s empirical dispersion-correction (D3) [39, 40] and the Becke-Johnson (BJ) damping-correction [41] in conjunction with the 6-31G(d,p) basis sets for C, H, N, and O atoms [42] and LANL2DZ basis sets (with their corresponding electrochemical potentials) for transition metal atoms and Br [43, 44]. Bulk solvent effects were incorporated for all calculations (including geometry optimizations and frequency calculations) using the self-consistent reaction field polarizable continuum model (IEF-PCM) [45–47]. We chose dichloromethane as the solvent. The reported thermodynamic data were computed at a temperature of 298.15 K and at 1 atm of pressure. Unless otherwise stated, energies are given as ΔH/ΔG in kcal/mol, although only the Gibbs free energies will be discussed. We have calculated several lower-lying electronic states (with different spin multiplicities) for all reported species, including antiferromagnetically coupled (so called open-shell singlet) states. Regardless, below we will discuss only the energetically lowest representatives of the low-spin and high-spin states. All reported structures were fully optimized without any geometry constraints. Previously, it was reported that the computational methods used in this paper accurately describe the energies and geometries of organometallic compounds [48–52]. Frequency calculations were carried out to verify the nature of the located stationary points. Graphical analysis of the imaginary vibrational normal modes was used to confirm the nature of the located equilibrium structures. All calculations were performed by the Gaussian-09 suite of programs [53]. NBO 3.1 program, which is included in the Gaussian-09 suite of programs, was used to obtain natural bond orbitals (NBOs), atomic net charges, densities of spin, and Wiberg bond indexes (WBI) at the optimized geometries.

21.3  Results and Discussion 21.3.1  Geometrical and Electronic Structures of the (RCOO)4-[M2] Complexes The geometry and electronic structure of the (tetracarboxylate)-[M2] complexes have been the subject of numerous studies [8, 30, 54–64]. These studies have identified the most critical two types of bonds that directly impact to structural stability of these complexes. They are: (i) the donor-­ acceptor M–O(carboxylate), and (ii) the M(dzz) + M(dzz) sigma bonds. The former requires empty or partially occupied dxy-AOs of metals, whereas the latter is expected to be more effective if the M-centers will have a partially occupied dzz-AO (see Figure 21.3). M=Rh. The Rh(II)-centers with d7-electron configuration is expected to be ideal for the formation of strong Rh-O(carboxylate) donor-acceptor bonds, as well as the M(dzz) + M(dzz) covalent bonds in the (tetracarboxylate)-[Rh2] complex (see Figure 21.4). Figure 21.3  Schematic representation of the important M–O(carboxylate) and metal-metal σ-bonds in the (RCOO)4-di-metallic complexes.

447

448

21  (Tetracarboxylate)bridged-di-transition Metal Complexes and Factors Impacting Their Carbene Transfer Reactivity

Figure 21.4  3D presentation of the lowest singlet and triple state (RCOO)4-[Rh2] and (RCOO)4-[Co2] complexes. For important geometry parameters of these species, see Table 21.1.

Consistently, previous extensive computational analyses have shown that, in its ground singlet [σ(dzz+dzz)]2[π(dxz+dxz)]2[π(dyz+dyz)]2[δ(dxy+dxy)]2[δ*(dxy-dxy)]2[π*(dxz-dxz)]2[π*(dyz-dyz)]2[σ*(dzz-dzz)]0 electronic state, this complex has a single σ[Rh(s)–Rh(s)] bond, along the Rh-O(carboxylate) donor-acceptor bonds (see Figure 21.5). Our NBO analyses of the (MeCOO)4-[Rh2] complex show that the Rh-O(carboxylate) donor-acceptor bonds are polarized toward the O-centers with 11.5% (Rh) + 88.5% (O) content. In (MeCOO)4-[Rh2], Rh-centers are reduced, and have only +0.72 |e|. In other words, (i) almost 2.56 |e| electron densities were transferred from four carboxylates to the Rh2-core; and (ii) rhodium centers are in their +1 oxidation states. The calculated value of the Rh– Rh distance, 2.394–2.421 Å (Table 21.1), is consistent with the single bond character of the Rh–Rh bond. The calculated Wiberg bond index (WI) of 0.83–0.82 supports this assignment. The natural bond analyses shown that, the bonding σ-orbital has only 1.66 electrons, and its anti-bonding component, σ* orbital, holds ca 0.27 electrons, in the (MeCOO)4-[Rh2] complex. Interestingly, the energetically lowest excited triplet states of the (RCOO)4-[Rh2] complexes are the (σ)2(π)4(δ)2(δ*)2(π*)3(σ*)1 triplet states, which can be formed from the singlet ground state by the π* → σ* one electron transition (which is symmetry-allowed only under C1-symmetry [65]). These triplet states are calculated to be ~21–23 kcal/mol higher in free energy than the corresponding singlet states (Table 21.1). Spin density analyses show that in the excited triplet state, each Rh

Figure 21.5  Electronic configuration of the ground singlet electronic state of the (MeCOO)4-[Rh2] complex.

21.3  Results and Discussion

449

Table 21.1  The calculated geometry parameters (in Å) of the (RCOO)-[M2] complexes, at their energetically lowest low-spin (L) and high-spin (H) electronic states, as well as energy difference, Δ(L–H), in kcal/mol, between these electronic states (see Figures 21.4, 21.6, and 21.8 for the labeling used). Complex R

State

M1-M2

M1-O1

M1-O3

M1-O5

M1-O7

M2-O2

M2-O4

M2-O6

M2-O8

M=Rh;

1

2.394

2.065

2.064

2.065

2.064

2.064

2.065

2.064

2.065

3

2.452

2.073

2.072

2.073

2.072

2.072

2.073

2.072

2.073

1

2.395

2.060

2.060

2.060

2.060

2.060

2.060

2.060

2.060

3

2.454

2.068

2.068

2.068

2.068

2.068

2.068

2.068

2.068

1

2.421

2.066

2.067

2.067

2.066

2.067

2.066

2.067

2.066

3

2.486

2.079

2.080

2.077

2.075

2.080

2.079

2.075

2.077

1

2.357

1.919

1.919

1.920

1.920

1.920

1.920

1.919

1.919

0.00

3

2.452

1.912

1.910

1.911

1.908

1.941

1.923

1.941

1.924

3.5

1

2.360

1.916

1.915

1.916

1.916

1.916

1.916

1.916

1.915

0.00

3

2.438

1.929

1.923

1.923

1.930

1.925

1.920

1.919

1.925

5.4

1

2.439

1.924

1.923

1.926

1.925

1.926

1.926

1.924

1.925

0.00

3

2.558

1.928

1.925

1.928

1.923

1.945

1.925

1.947

1.925

2.6

1

2.623

2.066

2.067

2.066

2.067

2.066

2.067

2.066

2.067

0.00

3

2.532

2.232

2.233

2.234

2.235

2.042

2.041

2.041

2.041

2.4

1

2.622

2.061

2.061

2.061

2.061

2.061

2.061

2.061

2.061

0.00

3

2.534

2.230

2.228

2.230

2.230

2.036

2.037

2.036

2.036

2.3

1

2.675

2.071

2.072

2.072

2.072

2.070

2.070

2.070

2.070

1.7

3

2.586

2.245

2.246

2.252

2.255

2.042

2.040

2.040

2.042

0.00

1

2.427

1.895

1.895

1.895

1.895

1.895

1.895

1.895

1.895

0.00

3

2.340

1.958

1.957

1.957

1.957

1.957

1.957

1.957

1.957

2.6

1

2.428

1.891

1.891

1.891

1.891

1.891

1.891

1.891

1.891

0.00

3

2.338

1.952

1.952

1.952

1.952

1.952

1.951

1.951

1.951

3.5

1

2.498

1.900

1.899

1.899

1.900

1.900

1.900

1.900

1.900

0.00

3

2.409

1.964

1.963

1.963

1.964

1.963

1.964

1.963

1.963

0.4

1

2.886

2.431

2.432

2.432

2.431

2.064

2.064

2.064

2.064

1.2

3

2.727

2.221

2.204

2.217

2.205

2.203

2.219

2.207

2.219

0.00

1

2.914

2.440

2.441

2.444

2.437

2.052

2.052

2.052

2.053

1.4

3

2.730

2.208

2.205

2.201

2.204

2.201

2.204

2.207

2.205

0.00

1

2.847

2.226

2.226

2.223

2.223

2.233

2.234

2.236

2.237

1.6

3

2.806

2.224

2.218

2.223

2.211

2.215

2.222

2.215

2.226

0.00

Me

A A

Ph

A A

CF3

A A

M=Co;

Me

Aos A

Ph

Aos A

CF3

Aos A

M=Pd;

Me

A A

Ph

A A

CF3

A A

M=Ni;

Me

A A

Ph

A A

CF3

A A

M=Ag;

Me

A A

Ph

A A

CF3

A A

Δ(L–H)

0.00 21.9 0.00 23.2 0.00 21.2

(Continued)

450

21  (Tetracarboxylate)bridged-di-transition Metal Complexes and Factors Impacting Their Carbene Transfer Reactivity

Table 21.1  (Continued) M1-M2

M1-O1

M1-O3

M1-O5

M1-O7

M2-O2

M2-O4

M2-O6

M2-O8

Complex R

State

M=Cu;

1

2.521

1.988

1.988

1.988

1.988

1.988

1.988

1.988

1.988

0.00

3

2.507

1.973

1.972

1.973

1.973

1.973

1.973

1.973

1.973

0.3

Me

Aos A

Ph

1

2.517

1.967

1.967

1.967

1.967

1.967

1.967

1.967

1.967

0.00

3

2.512

1.967

1.967

1.967

1.967

1.967

1.967

1.967

1.967

0.1

Aos A

CF3

Δ(L–H)

1

Aos

3

A

2.633

1.990

1.992

1.992

1.990

1.992

1.990

1.990

1.992

0.00

2.618

1.974

1.975

1.974

1.975

1.974

1.975

1.974

1.975

0.7

center has almost one (0.95 |e|) unpaired spin, and the Rh–Rh bond has slightly larger than halfbond character with 0.55 Wiberg bond indexes. NBO analysis also show the existence of two beta Rh–Rh bonding orbitals in the triplet state complex. The first one is a Rh–Rh σ-bond, whereas the second one is a Rh–Rh π-bond. Thus, the Rh–Rh bond with half σ- and half π-bonds has more than half-bond character. Bearing in mind that a one-electron π-bond is intrinsically weaker than oneelectron σ-bond, one can explain the ~0.06Å elongation of the Rh–Rh bond distance in the triplet state complexes compared with the singlet state complexes. The data presented in Tables 21.1 and 21.2 for the (RCOO)4-[Rh2] complexes (where R = Me, Ph and CF3) show only a slight impact of the R-groups on the calculated geometry of and singlettriplet energy splitting in these complexes. M= Co(II). As seen in Table 21.1, replacement of Rh(II) in [RCOO]4–[Rh2] by its 3d-analog, Co(II), leads to formation of the high-spin [RCOO]4–[Co2] complex with almost one unpaired Table 21.2  The calculated important spin densities (in |e|) of the (RCOO)4-[M2] complexes, at their energetically lowest low-spin and high-spin electronic states (see Figures 21.4, 21.6, and 21.8 for the labeling used). Complex R

State

M1

M2

O1

O2

O3

O4

O5

O6

O7

O8

M=Rh;

Me

3

0.95

0.95

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.01

Ph

3

0.95

0.95

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.01

CF3

3

0.96

0.96

0.00

0.03

0.00

0.03

0.00

0.03

0.00

0.03

Me

1

-0.95 0.94

0.02

0.02

0.02

0.02

-0.02

-0.02

-0.02

-0.02

M=Co;

A A A Aos

3

1.00

1.01

0.00

-0.03

0.00

-0.03

0.00

0.00

0.01

0.00

1

0.89

-0.89

-0.02

-0.02

-0.02

-0.02

0.02

0.02

0.02

0.02

3

1.01

1.10

-0.02

0.00

0.00

-0.02

0.00

0.00

0.01

0.01

1

-0.94 0.94

0.02

0.02

0.02

0.02

-0.02

-0.02

-0.02

-0.02

3

1.00

1.07

0.02

0.04

0.00

0.02

0.00

-0.01

-0.01

-0.01

Me

3

1.27

0.31

0.02

0.02

0.02

0.02

0.09

0.09

0.09

0.09

Ph

3

1.29

0.31

0.09

0.09

0.09

0.09

0.02

0.02

0.02

0.02

CF3

3

1.32

0.31

0.02

0.02

0.02

0.02

0.08

0.08

0.08

0.08

A

Ph

Aos A

CF3

Aos A

M=Pd;

A A A

21.3  Results and Discussion

Table 21.2  (Continued) Complex R

State

M1

M2

O1

O2

O3

O4

O5

O6

O7

O8

M=Ni;

Me

3

1.05

1.05

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.01

Ph

3

1.04

1.04

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.01

CF3

3

1.07

1.07

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.01

Me

3

0.35

0.35

0.15

0.15

0.15

0.15

0.15

0.15

0.15

0.15

Ph

3

0.36

0.36

0.15

0.15

0.15

0.15

0.15

0.15

0.15

0.15

CF3

3

0.38

0.38

0.14

0.14

0.14

0.14

0.14

0.14

0.14

0.14

Me

1

0.57

−0.57

0.10

0.10

0.10

0.10

3

−0.10

−0.10

−0.10

−0.10

0.57

0.57

0.10

0.10

0.10

0.10

0.10

0.10

0.10

0.10

1

0.58

−0.58

0.10

0.10

0.10

0.10

3

−0.10

−0.10

−0.10

−0.10

0.58

0.58

0.10

0.10

0.10

0.10

0.10

0.10

0.10

0.10

1

0.59

−0.59

0.10

0.10

0.10

0.10

3

−0.10

−0.10

−0.10

−0.10

0.59

0.59

0.10

0.10

0.10

0.10

0.10

0.10

0.10

0.10

M=Ag;

M=Cu;

A A A A A A Aos A

Ph

Aos A

CF3

Aos A

α-spin on each Co-centers. For the [RCOO]4–[Co2] complex, the antiferromagnetically coupled singlet ground electronic states of these complexes are lower in energy with the (s)0.30(p)0.34(dxy)0.61 (dxz)1.99(dyx)1.99(dx2-y2)1.97(dz2)1.05 valence electronic configurations of the Co-centers. The partial population of the Co(dxy) orbital allows the formation of the strong Co-carboxylate interactions. The generated Co–O bonds have almost 11–12% contributions from the Co-centers and 89–88% contributions from the O-centers of carboxylate. Thus, these Co–O bonds are highly polarized toward oxygen centers. As a result of these interactions, some electron densities have transferred from carboxylates to the Co-centers: in [MeCOO]4–[Co2], Co-centers possess only +0.74 |e| positive charges, whereas each acetate ligand has a total of -0.42 |e| negative charge. As we could expect from the electronic configurations of the Co-centers in [RCOO]4–[Co2] reported here, the lowest singlet electronic states of these complexes may have a Co(sdzz)–Co(sdzz) σ-bond. Indeed, calculations show the presence of the Co–Co σ-bond with 1.77 |e| electron population. Because of the presence of such Co–Co bonding, the Co–Co distances are 2.357, 2.360, and 2.439 Å for R = Me, Ph, and CF3, respectively. Because the calculated singlet-triplet energy differences are small and changes via 2.6 kcal/mol (R = CF3) 6, AuCl4− is transformed to [Au(OH)nCl4−n]− (n = 1–3), and the particle sizes of gold nanoparticles are below 4 nm [36]. Bond and colleagues [21] showed that pH 9 is the best for Au/TiO2, as most species in solution were anionic gold complexes, with residual chloride. Nevertheless, the best pH depends on the isoelectric point of the support. Urea can also be used as an alternative to adjust the pH [3, 54], as it slowly decomposes in the solution, forming hydroxyl ions that are consumed as they are formed. Louis and colleagues realised that urea allowed them to obtain the same size of Au nanoparticles as NaOH (2–3 nm) [55]. Other authors also used this method, obtaining very active catalysts [56, 57]. However, DP is not adequate for some supports, such as activated carbon [11, 14, 16] or zeolites [58], as they have high isoelectric points. DP has advantages over CP as all of the active phase is on the support surface and not buried inside [3, 14]. It also produces narrow particle size distributions; however, it is preferrable that the support has a surface area above 50 m2/g [3, 14]. Figure 23.2 shows some examples of gold nanoparticles prepared by DP.

23.2.5  Liquid-phase Reductive Deposition (LPRD) The liquid-phase reductive deposition (LRPD) procedure was described by Sunagawa et al. and includes mixing aqueous solutions of HAuCl4 and NaOH, aging for 24 hours at room temperature in the dark to complete hydroxylation, adding the support, dispersing ultrasonically for 30 minutes, and aging at 100 °C overnight [59]. The resulting solid is washed many times to remove chloride and dried. Adsorption of the gold ions takes place on the support surface, where reduction also happens. This method was effectively used to obtain Au nanoparticles on several supports [19, 28, 35, 46–48, 60, 61].

23.2  Preparation Methods

Figure 23.2  Transmission electron microscopy (TEM) images of Au/Al2O3 (a) and Au/CeO2 (c) prepared by deposition precipitation (DP), with respective size gold nanoparticle size distributions (b, d). Reproduced with permission from Ref [22] / Elsevier.

23.2.6  Ion-Exchange This method consists of replacing the ions of the support surface by gold ions. It is very efficient for zeolites, but adding active species inside the cavities of these materials, instead of having them on the surface, poses some obstacles, like the shortage of adequate cations or cationic complexes [3]. However, this method proved to be successful for several Au/zeolites [62, 63]. Pitchon and colleagues used direct anionic exchange (DAE) of the Au species with OH− groups of the support [64, 65]. HAuCl4 (aqueous solution) is poured into the support, the mixture is heated to 70 °C and maintained there for 1 hour, then it is filtered, washed with warm water, dried during the night, and calcined in air at 300 °C. To assure a complete removal of chloride ions, a fraction of the dried catalyst is washed with a concentrated solution of ammonia. Nevertheless, caution must be taken, as gold and ammonia might produce fulminating gold that is explosive [1, 14, 16].

23.2.7  Photochemical Deposition (PD) Photochemical deposition (PD) procedure ensures metal deposition on semiconductors, with metal ions being simultaneously reduced by conduction band electrons [66]. The method can be improved by the use of sacrificial electron donors (e.g. methanol, 2-propanol, or formaldehyde) that can

485

486

23  Environmental Catalysis by Gold Nanoparticles

provide a large quantity of electrons. The Au in aqueous solution is mixed with the sacrificial electron donor and the support, sonicated for 30 minutes, and photodeposited using an ultraviolet lamp. This procedure was used to prepare Au on TiO2 [66–68], ZnO [60, 66, 69] and other supports.

23.2.8  Ultrasonication (US) Ultrasonication (US) is analogous to PD but with no photodeposition; the sample is sonicated for 8 hours. It was a serendipitous discovery made by Carabineiro et al. while attempting to prepare Au/ZnO by PD [60]. The sample was going to be sonicated over 30 minutes (then photodeposited), but it was forgoten in the sonicator for 8 hours. As the mixture showed a deep purple colour, like those prepared by PD, it was filtered, washed, dried, tested in CO oxidation, and found to be a very active catalyst [60]. A TEM image and the size distribution are depicted in Figure 23.3. US was also used to prepare Au on Fe2O3 [25], MgO [44], CuO, La2O3, NiO, and Y2O3 materials [26], but the results were not so good as with Au/ZnO [60].

23.2.9  Vapor-Phase and Grafting Vapor-phase and grafting procedures are similar; the difference is whether a solvent is used. In the vapor-phase (chemical vapor deposition), a volatile gold compound is carried to a support with a high area, with the help of an inert gas, and then reacts chemically with the support surface to form a precursor of the active species [14, 16, 36]. The most commonly used precursors are AuCl3 or HAuCl4, but other compounds without chloride have been used. In physical vapor deposition, Au is vaporised and deposited on the support in high vacuum conditions [16]. According to 3M (Minnesota, USA) very active Au catalysts can be prepared like this for several supports, including water soluble materials, or those not suitable for DP, such as SiO2 [70]. This procedure is cheap, reproducible, needs no washing or thermal treatments, and is not toxic. Very stable Au nanoparticles (up to 600 °C) were prepared on Al2O3 by this technique [71]. In grafting, a solution of a gold complex reacts with the support surface, and the formed species are converted to catalytically active forms. Gold phosphine complexes were grafted onto several

Figure 23.3  Au/ZnO prepared by ultrasonication (US): transmission electron microscopy (TEM) image (a) and Au nanoparticle size distribution (b). Reproduced with permission from Ref [60] / Elsevier.

23.3  Properties of Gold Nanoparticles

hydroxides [3, 14, 72], as they have –OH groups that are able to react with the Au compounds. The procedure includes room temperature vacuum drying and air calcination, which leads to a simultaneous transformation of the precursors to oxides and Au particles. Au–phosphine complexes are great choices as they decompose by heat, forming metallic Au metal at a temperature similar to the value needed to converted metal hydroxides into oxides. These ligands are also expected to delay the growth of Au nanoparticles into larger ones. Gold can be supported on MCM–41, SiO2, SiO2–Al2O3, or activated carbon, as highly dispersed nanoparticles using a Au acetylacetonate complex, which is much more effective for these supports than liquid-phase procedures [73].

23.2.10  Bi- and Tri-metallic Au Catalysts It is already well known that Au is an effective catalyst alone, but it is also effective when combined with other metals [14]. Gold-based bimetallic materials have a large potential for several reactions, such as selective oxidations and hydrogenations, C–C coupling, and photocatalysis, due to their high activities and selectivities in mild conditions [74]. Bimetallic Au catalysts can be prepared by many procedures, as described in many reviews [75].

23.2.11  Post-treatment and Storage Several post-treatments can be used, such as calcination and reduction [1, 14, 16, 21, 46, 64, 72, 76–78]. However, several catalysts can be used without any further treatment. In fact, reduction or calcination can have negative effects in some cases [21, 79]. Also, the size of gold nanoparticles can increase with thermal treatments [1, 14, 16]. However, materials prepared by COL need a heat treatment for decomposition of the organic scaffold [28–34]. In terms of storage, freshly prepared samples should be kept at 0 °C and calcined materials should be kept in a cold place. Once dried, samples should be stored in a desiccator, under vacuum, in the dark, and be reduced immediately before use [16, 80].

23.3  Properties of Gold Nanoparticles Gold nanoparticles have very different properties than bulk gold. The decrease in size of a gold particle leads to changes in the metal structure [81] and a large increase in the quantity of low coordinated surface atoms, such as edge and corner atoms [81]. It also produces changes in the electronic properties, as the metallic character is lost and particles become non-metallic or semiconductive (caused by quantum size effects) [81]. These changes in structure and electronic properties produce alterations in the Au–Au bond distance [81]. For example, for ~1 nm sizes, the bond length is shorter (2.72 Å) than that of the bulk gold (2.88 Å) [81]. Moreover, if nanoparticles are supported, their behaviour will depend on the type of support, method of preparation, pre-treatment, calcination, and other factors [1, 3, 14, 16, 79]. Some of the more important properties of Au nanoparticles are succinctly discussed in the following.

23.3.1  Activity It is widely known that the presence of small (usually 99% was reported for Au/TiO2 based catalysts, using H2 and O2 [106]. For Au nanoparticles above 2 nm, high selectivity to propene oxide is obtained [14, 107], but propane is obtained instead for sizes smaller than 2 nm [11, 14, 107]. Haruta used Au/TS-1-K-1 and found out that Au nanoparticles above 2 nm gave mostly acrolein, while sizes below 2 nm gave propene oxide [102]. ●

The selective oxidation of CO in the presence of H2, or PROX, discussed hereafter, is also a good example. Often, H2 is obtained from a source containing carbon, like a natural gas, which leads to CO production. This gas can be detrimental, even in small quantities, as it can poison Pt-based proton exchange membrane (PEM) fuel cells (as CO tightly binds to Pt) [14, 108, 109]. Unlike PGMs, gold-based materials are more active for CO oxidation than for H2 oxidation. Moreover, Au is nearly insensitive to CO2, and its activity is enhanced by moisture [12, 110–113]. Those are products of CO oxidation (CO + O2 ⇄ CO2) and water-gas shift (CO + H2O ⇄ CO2 + H2) reactions (also discussed later), common in the production and purification of H2. This makes gold a good candidate for PROX, as shown in several articles [51, 65, 67, 69, 77, 82, 88, 101, 114].

23.3.3  Durability If considering Tammann temperature, Au should sinter around 532 °C (half of the melting point of metallic Au: 1064 °C) [14]. However, small gold nanoparticles might sinter at a much lower temperature, as the melting values should significantly diminish with decreasing particle size (this means that a Au nanoparticle of ~2 nm should melt around 300 °C [14, 20], a much lower value than bulk gold). However, experience has been showing that gold catalysts have much higher durability than expected [1, 14]. Some examples can be found in literature for this, including: Au on cobalt oxide on zirconia-stabilised ceria, titania, and zirconia, that was able to resist at 500 °C for 157 hours, although with some deactivation [92]. ● Adding transition metal oxides to obtain Au/MOx/Al2O3 prevented Au sintering during methane oxidation up to 700 °C [115]. ● Au/Al2O3 showed good activity in NO conversion, surviving pre-treatments of 600 °C in air for 24 hours, several cycles at 150–500 °C, and being kept overnight at 500 °C [52]. Another Au/ Al2O3 sample was hydrothermally stable at 600 °C for 96 hours [116]. Au/Al2O3 can also be very efficient for the conversion of glucose to gluconic acid (>99% selectivity) and durability (no loss of activity or selectivity up to 110 days) [117], with 3.8 ton of product being produced per gram of Au in 70 days. ● Au/ZnO, used for PROX, resisted for 350 hours at 80 °C [118]. ● A Au2Sr5O8 material patented by Toyota Motor Company had improved durability (tested up to 800 °C for 5 hours) [119]. ● Nanoaligned rutile rods limited the growth of Au up to 800 °C, and CO oxidation was very high even after exposure of the material to high temperature [120]. ● Au/alumina with complete removal of chloride by washing with ammonia through DAE, were resistant to oxidation at 600 °C and to water [64]. Similar catalysts were also stable up to 600 °C for 100 hours [116]. ●

489

490

23  Environmental Catalysis by Gold Nanoparticles

1%Au/AC with YCl3 showed 87.8% conversion for acetylene hydrochlorination to vinyl chloride monomer (VCM) and almost 100% selectivity [121]. The catalyst kept high activity for >2300 hours at 30 h-1 GHSV (C2H2) at 180 °C. ● A review refers several strategies to stabilise gold nanoparticles on solid supports, like developing gold core-shell or yolk-shell structures [122], which also increased catalyst durability. One example is gold nanoparticles within mesoporous frameworks, which enhanced assured thermal stability of Au nanoparticles up to 800 °C and a good durability (>130 h at 375 °C) for aerobic oxidation of benzyl alcohol and CO oxidation [123]. ● Au/TiO2 catalysts with 3D nanorod structures, inside a stable polymorph were able to maintain activity for CO oxidation below 115 °C, even after several cycles at 800 °C, being promising for industrial applications [120]. ● A Keggin-type POM (polyoxometalate), Cs4[α-SiW12O40]·nH2O, was used as a support for Au/Al2O3. The Au/La2O3/Al2O3 catalyst showed a good performance, being stable in the presence of CO2 [101, 185]. Au/MgO materials, modified with Mn and Fe, had higher stability, activity, and selectivity than unmodified Au/MgO [220]. The addition of these modifiers changed the type of metal ion–metal nanocluster active sites linked to CO activation, and the new sites were more active for temperatures higher than room temperature. MgO-supported Au catalysts modified by La, Pb, Sm, V, and

H2, CO, O2 Conversion (%)

498

100 (a)

(b)

75 50 25 0 100 (c)

(d)

75 50 25 0

0

100

200 300 Temperature (ºC)

400

100

200 300 Temperature (ºC)

400

Figure 23.9  Conversion of CO (⚪), H2 (∆), and O2 (♦) on (a) Au/Al2O3, (b) Au/MnOx/Al2O3, (c) Au/MgO/Al2O3, and (d) Au/MnOx/MgO/Al2O3 in the presence of H2. Reproduced from Ref (148), with permission from Elsevier

23.4  Reactions Efficiently Catalyzed by Gold Nanoparticles

Y were also tested [221]. The formed interfaces of the multi-component catalysts allowed a new type of active site ensembles able to adsorb and activate CO and O2, but prevented activation of H2. Au/Fe2O3-TiO2 (1:4) showed higher CO conversion than single Au/TiO2 and Au/Fe2O3, especially below 100 °C, due to the presence of reducible Au species [181]. Au on mixed Ni-Fe supports were also studied [200]. Laguna et al. reported that adding 10% Fe to Au/ceria resulted in an improvement of catalytic activity, with the composite having a higher CO conversion than Au/ceria below 150 °C, and being scarcely affected by CO2 and H2O from the gas stream [222]. The strong interaction between Fe2O3 and CeO2 contributed to larger oxygen mobility, leading to a higher catalytic activity of Au/Ce-Fe-O [65, 208]. Nice results were obtained for Au/La-Ce-O, which was stable, even with 10% water in the feed stream [223]. Also improved results were obtained when doping Au/ceria with Cu [182, 193], Mn [208], Ni [224], Sm [225], Y [51], Zn [225, 226], or Zr [207, 226, 227]. However, no improvement was found when Co3O4 was added to CeO2 [208]. H2 oxidation (accompanied by Co3O4 reduction to inactive CoO) occurred, together with methane formation (with a contribution from Co being reduced to the metallic state). However, addition of Co3O4 to TiO2 had better results, as it increased the oxygen defects on TiO2, forming more strongly bound Au atoms [213]. Co3O4 amorphous together with TiO2 enhanced electronic interaction between Co3O4-TiO2 and Au, stabilising the Au particles and improving the catalytic activity. Au/CeO2-MOx/Al2O3 catalysts (M = Cr, Cu, Fe, La, Ni, Y) were also used [202]. Cu and Fe enhanced the catalytic performance of Au/CeO2/Al2O3. The Cu-containing material was water resistant and CO oxidation efficient. Better catalysts were obtained adding Pt to Au. For example, Au/ZnO had good stability at 80 °C, but its activity was slightly decreased above 350 hours [190]. But when small quantities of Pt were added, the stability improved, up to 500 hours [196]. Also, better results were found for Au–Pt/ ceria than for Au/ceria [228]. Zhang et al. [196] tested Au–Pt/ZnO catalysts and found that 1% Pt improved the stability, but a larger amount of Pt had a detrimental effect on the CO2 selectivity. Hutchings et al. [204, 205] prepared a Au/Fe2O3 catalyst by CP of Au3+ and Fe2O3 and calcination in two stages. They obtained a commercially acceptable material to competitively oxidate diluted CO in the presence of moisture excess of H2 and CO2. Calcining the material at 400 and then at 550 °C yielded a catalyst able to remove >99.5% of the CO at 80 °C, the operating temperature of fuel cells [204, 205]. Concerning the reaction mechanism, the lattice oxygen of the oxide supports can also participate in the reaction (without Au or being enhanced by Au), and the Mars–van Krevelen mechanism is accepted, together with Au perimeter effects [114, 147, 148, 198, 209, 215, 216, 227].

23.4.3  Water-gas Shift Water-gas, syngas or synthesis gas is a mixture of CO and H2, first synthesised from the reaction of steam with coke [14]. Hydrogen can be obtained from biomass and fossil fuels (coke, coal, naphtha, oil, and natural gas) [194, 229, 230]. Steam reforming of methane from natural gas is a common and cheap process used to produce around half of the H2 used globally [231]. However, the resulting gas contains a small amount of CO, which needs to be removed [187], as mentioned previously. The water-gas shift (WGS) reaction (CO+H2O→CO2+H2) is often used to reduce CO and produce H2 [231], and can be combined with steam reforming of methane (and other hydrocarbons) [232], allowing tuning of the H2/CO ratio in the water gas [233] to produce methanol [234], ammonia [235], or liquid hydrocarbons (like kerosene, gasoline and lubricants; that is, the

499

500

23  Environmental Catalysis by Gold Nanoparticles

Fischer-Tropsch synthesis) [231], and purification of H2 for fuel cells [231]. Highly purified hydrogen is mandatory for fuel cells (as referenced previously), and also for the synthesis of ammonia, as CO acts as a poison, not just for the fuel cell Pt electrode anode (as already mentioned), but also for the Fe catalyst used in the production of ammonia [236, 237]. WGS is also a key step in automobile exhaust reactions, and H2 is a very effective reductant to remove NOx [14]. WGS is a relatively mild, equilibrium-limited, exothermic reaction, that is thermodynamically favored at low temperatures, although the kinetics is decreased under such conditions, reducing the yield of H2 and increasing the quantity of catalyst needed to attain sufficient CO conversion [238–242]. The way to obtain a good reaction rate and a high CO conversion is to carry the WGS reaction in two steps: high-temperature (HT-WGS) and low-temperature (LT-WGS) [242–244]. HT-WGS operates around 350–500 °C and has a rapid reaction rate yet low CO conversion. LT-WGS, performed around 200–350 °C, is favored by thermodynamics to convert CO, tuning the CO:H2 ratio of the gas coming out of the HT-WGS reactor, but has slower reaction rate, as previously mentioned [238–242]. A good part of the feed CO is converted in the first reactor due to the fast kinetics. The amount of CO is further reduced in the second reactor, at lower temperature, given the exothermal nature of the reaction. Two different types of materials are industrially used for the WGS reaction: HT-WGS uses Fe–Cr oxides, whereas LT-WGS needs Co–Mo or Cu–Zn oxides [14, 229, 231, 242, 243]. Catalysis containing Cu has disadvantages, such as extended pre-reduction time, sensitivity to air and water, susceptibility to being poisoned by sulfur, slow kinetics at low temperature [231, 245]. Thus, they need precise reduction conditions, which makes them be very reactive to air after activation (pyrophoric), which is unsafe during operations. This makes them unsatisfactory for automotive applications [246]. Cu catalysts also operate from 200–300 °C, whereas gold is active from 100–200 °C. This is important because lowering the temperature causes the CO equilibrium concentration to be improved and the CO amount in the reformate gas to be lowered, avoiding the use of further CO removing systems that are still used in fuel processing commercial systems (as CO is poisonous, as already referred) [14]. Gold has an extra advantage, as it selectively catalyzes this oxidation (PROX reaction), as its selectivity for hydrogen/oxygen competing reaction is much lower at room temperature, as already mentioned. Noble metals (like Pt and Au) on oxide supports efficiently catalyze LT-WGS [1, 14, 16, 47, 48, 84, 152, 229–231, 244, 246–262]. They are promising for fuel cells, even though they are expensive [229, 236]. There is the need for catalysts that are: (i) non-pyrophoric; (ii) usable without requiring pre-reduction treatments; and (iii) durable in quick heating and cooling cycles. Au catalysts are good candidates and have been successfully used in WGS [1, 14, 16, 47, 48, 152, 229–231, 244–246, 251, 252, 254, 256–260, 262–268]. When Au is added to metal oxides, there is an improvement in the activity of LT-WGS, explained by a synergistic effect between Au and the oxide [231]. Moreover, as described previously, a Chevron patent from 2005 already claimed that gold on sulphated ZrO2 could be used for HT-WGS and LT-WGS [129], which is very advantageous. In the 1990s, Andreeva et al. found that Au/Fe2O3 was unexpectedly highly active for WGS [247, 269]. CO conversion had started already at 120 °C, being more efficient than the WGS industrial catalysts of that time. Those authors kept studying gold and reported that Au/TiO2 and Au/ZrO2 were also active, even more than Au/Co3O4 [249, 251]. Shortly afterwards, Haruta also reported a high activity for Au/TiO2, compared to Au/ZnO, Au/ Al2O3, and Au/Fe2O3 [84, 248]. Andreeva et al. also tested mixed oxide supports (Fe2O3-ZnO and Fe2O3-ZrO2) and amorphous TiO2 and ZrO2 [250]. Well-crystallised metal oxides were more efficient than amorphous oxides. Similar results were found for mixed oxides [250, 251].

23.4  Reactions Efficiently Catalyzed by Gold Nanoparticles

Flytzani-Stephanopoulos reported the high activity of Au/ceria for WGS [260, 263, 266, 270, 271]. This is very important for automotive catalysts, as CeO2 is widely used in automotive threeway catalysts of emission-control, given its ability to quickly change oxidation state if the redox potential of the exhaust gases change. Several studies reveal that CeO2 is a very active support [1, 14, 16, 229–231, 246, 251, 252, 254, 256, 258–260, 262, 263, 265, 266, 270, 271]. It is known that ceria and ceria-containing materials show high capacities to transport oxygen and can easily shift between reduced and oxidised states (Ce3+ to Ce4+), and this can be enhanced by adding transition metal ions into the ceria lattice [231]. Thus, adding rare earth metals (REMs) to ceria has been tested. REMs are highly active when added to ceria, leading to enhanced thermal stability and improved catalytic activity [267]. Au/ceria doped with La and Gd also displayed higher catalytic activity [271]. Au/CeO2–ZrO2 catalysts are also very active for WGS and allow H2 to be obtained from H2O at 100 °C [272]. AuCe50Zr50 showed high activity and stability due to high Au dispersion [272]. Au/CeO2/Al2O3 was also investigated and transition metals like Cu, Fe and Zn were used as ceria dopants, improving the redox activity and, consequently, the activity [273]. Adding ZrO2 to Au/Fe2O3 highly improved activity and stability, but the addition of CaO, Cr2O3, CuO, La2O3, MgO, or NiO had a smaller effect [274]. Some Au/MxOy/TiO2 materials (MxOy = CaO, Al2O3, ZrO2 Y2O3, and REM oxides) had higher activity than Au/TiO2 if calcined at high temperature [275]. Au/carbide were also used [276]. Au/MoCx also showed exceptional LT-WGS activity. Au was well dispersed on the carbide and interacted with the support by strong charge transfer. Several reviews on Au catalysts for WGS showed that not just the nature of the support, but also the gold size, oxidation state, and preparation method are important factors [1, 14, 16, 229–231, 251, 252, 256, 258–260, 263, 265, 277]. DP and CP are the most common procedures used to prepare catalysts for LT-WGS reaction [14, 229–231, 246, 247, 251, 252, 256, 258–260, 265, 269, 278]. Carabineiro et al. used Au/Fe2O3 [47] and Au/TiO2 [48] obtained by LPRD and DIM, but DP produced more active catalysts [47, 48]. Bimetallic Au-Ru catalysts on Fe2O3 were also used [279]. When compared to Au/Fe2O3 and Ru/ Fe2O3 analogues, the bimetallic system had better activity at 100–300 °C. Au-M/Fe2O3 (M = Ag, Bi, Co, Cu, Mn, Ni, Pb, Ru, Sn, or Tl) were also used, with Ru and Ni having the best results [280]. Concerning Au–Cu/TiO2 [281], the presence of Cu+ enhanced the dispersion of reduced gold and also helped keep it highly dispersed during the reaction. Other bimetallic catalysts include Au–Pt/ CeO2 [282], Au–Pd/CeO2 [282], and Au–Re/CeO2 [245]. WGS, catalyzed by Au or other catalysts, is usually believed to progress by two pathways: a redox regenerative mechanism and an associate mechanism [259, 265, 277]. In the former, CO is reacted with a reducible oxide, forming CO2 and partially reducing the support, which is then oxidised by water, replenishing oxygen and releasing hydrogen. This takes place with a reducible support and at high temperature. The second mechanism includes the development of surface CHxOy intermediates (formates and carboxyls), and their decomposition to H2 and CO2, with possible reduction and reoxidation of the metal oxide support (Figure 23.10). The mechanism depends on the type of catalyst, preparation method, pre-treatment, reaction temperature, and gas composition [259, 265]. Au catalysts might suffer WGS activity loss due to agglomeration of gold nanoparticles (this happens for Au/ceria-zirconia, which are among the most active materials for LT-WGS [261], and Au/ ceria [264]). Au/ceria catalysts can deactivate due to poisoning or blockage of the active sites by hydrocarbons, carbonates, or formates [264], formed from CO and H2, on oxygen-deficient sites of ceria [254]. Nevertheless, regeneration is possible by calcination and 95% of the initial activity can be restored by heating at 400 °C in air [254]. Au/Fe2O3 deactivated due to large reduction of the support surface area with time of use [283]. However, rapid deactivation of Au/Co3O4 was due to reduction of the support [251]. Gold can

501

502

23  Environmental Catalysis by Gold Nanoparticles H C

O

O

H

O

O

H

C

H O

Au support

H

H

H

support

O

C

O

H

a)

C b)

O H

Au

Au support

O

C H

H

C

O

O

O

C

O

Au support

H

O

Au support

Au support

Figure 23.10  Carboxyl (a) and formate (b) mediated mechanism in the low-temerature water-gas shift (LT-WGS) reaction. The gold nanoparticle and support only show the position of surface sites. Adapted from Ref [277].

stimulate reduction of the supports, by activation of H2 on Au and spill-over of the atomic hydrogen to the support [251]. Au/CeZrO4 deactivated due to water that progressively removed the gold atoms from the support, diminishing the metal-support interaction [284]. The rate of exchange in Au0–CO infrared bands was well related with deactivation, while Aud+–CO species vanished more quickly on stream [278]. These considerations show that deactivation of Au catalysts in WGS is dependent on the type of catalyst and reaction conditions, such as temperature and gas composition [259]. Regeneration by calcination at high temperature is able to remove intermediates and strong adsorbates, but also re-disperse gold nanoparticles in the support and recover the surface oxygen [259].

23.4.4  Total Oxidation of Volatile Organic Compounds (VOCs) Volatile organic compounds (VOCs) are present in printing, painting, petroleum refinery, motor vehicles and fuel storage, being harmful pollutants to humans and environment [14, 285]. VOCs are divided into several types: Saturated or unsaturated aliphatic (including methane, propane and propene, with alkanes being the less reactive). ● Oxygenated (alcohols, such as methanol, ethanol, and propanol; aldehydes, like formaldehyde; ketones, such as acetone; esters, like ethyl acetate). The order of reactivity is: alkanes99% regioselectivity for the terminal epoxide with the former system, whereas 97% regioselectivity toward the internal epoxide was observed with the latter [22]. The same catalyst modified with chiraphos allowed the enantioselective terminal epoxidation of a variety of dienes with ee in the range 63–98% [23] (Figure 24.5). Previously, using the chiraphos, –CF3 derivative, we had been

24.3  Enantioselective Oxidations

Figure 24.5  Enantioselective epoxidation of terminal alkenes and BV oxidation of cyclic ketones with hydrogen peroxide using chiral Pt(II) complexes as catalysts.

able to asymmetrically epoxidize also a very simple molecule like propene obtaining propene epoxide with 41% ee [24]. Although these enantioselectivities may not be exceptional in some cases, it must be borne in mind that simple alkenes, possessing no extra functional groups in the molecule, lack secondary binding to the metal, a factor that is recognized to play a fundamental role in increasing enantioselectivity. Similarly, the desymmetrization of a series of racemic or meso ketones to yield the corresponding chiral lactones was successfully achieved using the hydroxo-bridged [(P–P)Pt(μ-OH)]22+ complexes again modified with chiral diphosphines [25]. Some examples are reported in Figure 24.5. The use of hydrogen peroxide in these reactions, the possibility of exploiting the wide library of phosphine ligands (chiral and non-chiral) available to tune the steric and electronic properties of the catalysts, the high product yields observed in most cases, the high product selectivities (even enantio-), the absence of oxidant waste in undesired side reactions, and the mild conditions applied, combine to make these oxidation reactions some good examples of ante litteram Green

521

522

24  Platinum Complexes for Selective Oxidations in Water

Chemistry, since most of them were discovered long before the twelve principles of Green Chemistry were established [26]. However, an important step forward in this respect would be the possibility to easily separate the catalyst from the reaction mixture and recycle it, and to remove the organic solvent (usually dichloromethane or dichloroethane) used to dissolve all reaction ingredients (except hydrogen peroxide). In fact, the solvent constitutes about 80% of the total mass of a reaction and in the industrial practice this sets two major problems: (i) the environmental acceptability of the solvent; and (ii) its recovery and recycling. With respect to the former point, methanol, acetone, and toluene are by far the most widely accepted solvents in industry for obvious reasons, while recovery and recycling have an efficiency that rarely exceeds 70%. In the pharmaceutical industry, where the standards of purity required are more stringent, the amount of solvents recycled having a sufficient purity is no more than 30%, the remaining part being simply incinerated to produce heat [27].

24.4  Water as the Reaction Medium In a chemical reaction the solvent is required to play several roles at the same time: (i) to ensure contact between substrates of different polarity; (ii) to control heat transfer; and (iii) to favor the interaction that leads to the final transformation. In this framework, among all possible liquids, water is certainly the one with the smallest impact on the environment. Water as a possible reaction medium has many advantages over organic solvents [28]: It is environmentally safe, non flammable, non toxic, cheap, readily available anywhere and requires no synthetic procedure to obtain. ● It does not contribute to greenhouse emissions and its E factor is assumed equal to zero [29]. ● It has a tunable acidity, and a large heat capacity and heat of evaporation that make it an ideal choice for exothermic reactions. ● It has a high polarity and H-bond donor and acceptor functionalities coexist. ● It is immiscible with most organic solvents making catalyst separation possible (at least in principle) in water/organic biphasic reactions. ●

On the other hand, it also has a few but crucial disadvantages that have precluded its widespread use in organic synthesis, first of all its poor solubility properties with respect to most organic substrates and catalysts (this is the major problem); furthermore, some chemical functionalities (especially organometallics) are not stable in the presence of water. The hydrophobic effect that makes most organics insoluble in water is essentially an entropic effect that is due to the existence in liquid water of an extensive hydrogen bonding network that causes a large cohesive energy. When a hydrophobic organic molecule is added to water, dissolution (i.e. molecular dispersion into the aqueous medium) would require extensive hydrogen bond breaking. Conversely, aggregation (i.e. two-phase formation) will minimize hydrogen bond disruption and will restrict the organic molecule motion, thus resulting in an increase in order and a decrease in entropy. In recent years, the interest in catalysis in water was stimulated by the Ruhrchemie-Rhone Poulenc process for the hydroformylation of propene and butene catalyzed by rhodium complexes modified with sulfonate-containing phosphine ligands in order to make them soluble in water [30]. Historically, it has been known since the 1980s that the hydrophobic effect can greatly accelerate reactions between poorly water-soluble substrates. Rideout and Breslow [31] found that the Diels–Alder cycloaddition reaction in water as the solvent was hundreds of times faster than in

24.4  Water as the Reaction Medium

organic protic and aprotic solvents, an observation that excluded simple polarity or H-bonding effects on catalysis. It is commonly accepted [32] that the hydrophobic effect forces molecules or apolar groups to aggregate in order to minimize their contact with water. This leads to a compacting of the reagents in the transition state thus increasing reaction rate. Concepts like the hydrophobic effect and the donor-acceptor hydrogen bonding ability of water have allowed rationalization of the enhanced productivity as well as regio-, diastereo-, and enantioselectivity in several catalytic reactions [33]. Mother Nature has always handled reactions between hydrophobic substrates using water as the sole solvent. How does this happen? Because of the hydrophobic effect, enzymes, the natural catalysts, fold in water and, thanks to a complex network of hydrogen bonding and dipole-dipole interactions, they assume their characteristic structure where the hydrophilic groups of the macromolecule are exposed on the surface in contact with water, while the hydrophobic portions lie inside the structure where the active site is generally located. This complex strategy, based on an amphiphilic nature, allows a hydrophobic phase to be well dispersed in water as well as to be a suitable reaction medium for organic transformations. These processes are pivoted by weak supramolecular interactions (again dipole-dipole, hydrogen-bonding, and others) that drive not only the reaction rate and the extent of the catalytic phenomenon, but also product as well as substrate selectivity. The simplest way to mimic, at least in part, enzymes behavior is to use a surfactant. Soaps have been used by humans for about 2000 years to remove fats. News on their preparation is reported by Pliny the Elder in his Naturalis Historia [34], in which he describes a strange practice by Gauls, who used to melt together tallow and wood ashes obtaining by cooling a solid material called sapo. Surfactants, the active ingredients of soaps, are amphiphilic molecules with a polar hydrophilic head and a hydrophobic tail that above a certain concentration (the so-called critical micellar concentration or cmc) self-assemble in solution to give nanometric aggregates (micelles) where hydrophilic heads are concentrated on the surface in contact with water, while hydrophobic tails point inside. Clearly, they are much smaller than enzymes, and, unlike enzymes, do not possess that large variety of internal functional groups and chemical complexity capable of driving exceptional activity and selectivity. However, similarly to enzymes, they constitute an organic nanophase dispersed in water capable of dissolving hydrophobic molecules. Surfactants are a large class of cheap chemicals, commercially available as cationic, anionic, and neutral molecules. Micellar media are a good choice to perform homogeneous catalysis in water for several reasons: They enhance catalyst and reagents solubility in water because their inside is an organic phase. They behave like an organic solvent and therefore transition metal complexes already optimized to be used as catalysts in solution need no ligand modification, unlike with catalysts made soluble in water for which polar groups must be introduced into the ligands, possibly altering their electronic properties. ● Catalyst local concentration inside the micelle is about two orders of magnitude higher than in a regular solvent, thus a beneficial effect on reactivity is possible. ● Product isolation is feasible via extraction with apolar solvents immiscible with water, followed by phase separation. ● With appropriate extraction solvents the catalyst remains confined in the micellar medium and in principle can be recycled. ● ●

From a certain point of view, the behavior of micelles resembles that of enzymes, isolating species from the bulk solvent (water) and playing several roles at a time (such as improving

523

524

24  Platinum Complexes for Selective Oxidations in Water

solubilization of organic reagents in water, favoring compartmentalization of reagents, and imparting unique chemo-, regio- and stereoselectivities). A limitation of micellar catalysis is related to the amounts of substrate that can be loaded into the micelles, usually lower than in common organic solvents, although the higher activities and selectivities often observed with micelles can partially compensate the disadvantage of working in diluted media. The distribution of organic species between bulk water and micelles depends on their polarity, charge, and dimension. Apolar substrates are almost exclusively hosted inside micelles. On the other hand, charged micelles tend to concentrate species of opposite charge on their surface, hence in cationic micelles the surface local pH is slightly more basic than in the bulk solution, and the opposite is observed for anionic micelles [35]. Depending on the nature of the micelle, upon dissolution a transition metal complex can experience a second coordination sphere due to non-covalent interactions with the micelle, and as in metallo-enzymes [36] this can play a significant role in driving the activity and selectivity. The interactions between catalyst and surfactants can also influence catalyst separation. Under the best conditions it could be possible to simply efficiently extract the product, leaving the catalyst and the surfactant in the micellar medium. This occurs when the metal catalyst and surfactant are oppositely charged and the organic reagents and reaction products are rather apolar and easily removable from the micellar medium. In other cases, product isolation is more difficult because extraction with a solvent removes partially the surfactant and the catalyst from the aqueous phase. This often occurs with neutral surfactants, neutral metal complexes, or organocatalysts.

24.5  Catalytic Oxidation Reactions in Water On the basis of these principles, the catalytic enantioselective oxidation reactions reported previously seem to be the ideal choice to test homogeneous catalysis in micellar media. In fact, (i) water is already present both as the solvent for H2O2 and as the reaction product in the original system; (ii) with the solvent dichloromethane (DCM) or 1,2-dichloroethane (DCE), the reaction medium is already biphasic; and (iii) the Pt catalysts are quite stable in water, a condition that does not apply to many transition metal and organometallic complexes. Our first attempt was performed in the oxidation of a series of thioanisoles. Selectivity issues were twofold: one concerned the sulfoxide/sulfone balance and the other the extent of enantioselectivity of the sulfoxide. Reactions were performed using the R-binap derivative of the hydroxobridged [(P–P)Pt(μ-OH)]22+ complexes as catalysts and the results obtained in the micellar medium were compared to those observed with a regular solvent. Some representative data are reported in Figure 24.6 and show the superior performance of the micellar medium both with respect to the sulfoxide/sulfone selectivity and in the enantioselectivity of the sulfoxide [37]. The catalyst being cationic, the best option as surfactant turned out to be an anionic one like sodium dodecyl sulfate (SDS). A cationic surfactant like cetrimonium bromide (CTABr) was the worst, whereas a neutral one was only slightly better than the latter. The micellar medium could not only improve the chemoselectivity of the process (sulfoxide/sulfone 80:1 in CH2Cl2 and >200:1 in water/SDS), but also significantly increase the enantioselectivity (up to 88%). Figure 24.6 reports also some data on the scope of the reactions in water/SDS as the reaction medium. As shown an electron-releasing substituent in the aromatic ring increases the activity but decreases the enantioselectivity, while the opposite occurs with electron-withdrawing substituents. Optimization of the reaction conditions requires an investigation on the effect of the surfactant concentration. Figure 24.6 shows that both the catalytic activity and the enantioselectivity of

24.5  Catalytic Oxidation Reactions in Water

Figure 24.6  Sulfoxidation of thioanisoles in CH2Cl2 vs. micellar media and effect of the surfactant concentration on the activity and selectivity.

the reaction are strongly influenced by the latter parameter, showing a maximum. This test must be done for each reaction investigated because the outcome can change dramatically. A study of the enantioselective epoxidation of terminal alkenes was also performed in micellar media using the same catalyst and the same conditions reported previously in Figure 24.7a summarizes some data using [(chiraphos)Pt(C6F5)(H2O)]+ as catalyst in H2O/Triton-X100 as the reaction medium that turned out to be the best choice [38]. The use of neutral surfactants was crucial to ensure sufficient catalyst solubilization as demonstrated by 2D-nuclear magnetic resonance (NMR) experiments. In the case of 4-methyl-2-pentene the epoxide was produced with 84% ee, while in dichloroethane under the same conditions, the enantioselectivity observed was 58% ee with very similar product yields. In this epoxidation system catalyst separation and recycling was possible by extraction of the reaction products with n-hexane. The micellar medium containing the catalyst was recycled at least three times with negligible differences in terms of activity and enantioselectivity [38].

525

526

24  Platinum Complexes for Selective Oxidations in Water

The asymmetric Baeyer–Villiger oxidation of cyclic ketones with hydrogen peroxide catalyzed by chiral diphosphine-Pt(II) catalysts was extensively investigated in water in the presence of different surfactants (Figure 24.7b) [39]. The reaction resulted extremely sensitive to the combination of substrate, catalyst and surfactant employed and only careful optimization of all three parameters enabled good results in terms of asymmetric induction and yield of lactone. It was also found that the oxidation of meso-cyclobutanones in micellar media allowed the reaction to proceed in high yields and better enantiomeric excess (ee up to 56%) with respect to organic media. Extension to meso-cyclohexanones resulted in a general decrease in yields but an enhancement of enantioselectivity (ee up to 92%) among the best reported in the literature and second only to enzymatic catalysis. Some striking effects on the extent of stereocontrol exerted by micelles in enantioselective reactions was observed in the Baeyer-Villiger oxidation of some meso-cyclobutanones catalyzed by the Co(salen) complex shown in Figure 24.7c [40]. This was completely inactive when employed in several organic solvents. The same complex turned out to induce significant asymmetric induction when employed in water, with best results using SDS as surfactant. Particularly interesting was the oxidation of chiral racemic cyclobutanones as shown in Figure 24.7c. Here the catalyst is called to control both the regioselectivity (lactone 1 vs. lactone 2) and the enantioselectivity. In the specific case reported in Figure 24.7c, in DCE as the solvent the Co catalyst is poorly active and non enantioselective (ee = 0%), while in the micellar medium the yield in stereoisomer 1 jumps to 80% and its ee to 88%. This striking effect is only due to the use of Triton-X114 that was found to be the best surfactant. In the latter examples, careful surfactant optimization was necessary since, as observed in other oxidation reactions, every combination of substrate and catalyst behave in a different way when using different surfactants and this needs optimization. Over the years we have investigated also other (non oxidation) reactions in micellar media using the same class of cationic Pt(II) complexes, like the hydration of terminal alkynes to yield methylketones [41] and the hydroformylation of alkenes to give exclusively linear aldehydes [42]. In both cases the use of H2O/SDS allowed both the increase of product yields and the recycle of the catalyst several times. Other reactions studied were the hydration of nitriles using Ru catalysts in H2O/ Triton-X114, again with improved yields and catalyst recycling [43] and the Diels-Alder reaction between cyclopentadiene and methacrolein catalyzed by a Cr(salen) complex in H2O/SDS with better yields with respect to organic solvents [44]. Over the past 15 years the general subject of micellar catalysis using soluble metal complexes as catalysts in hydrogenation, C–C and C–heteroatom bond formation, olefin metathesis, cross-coupling, cycloaddition reactions, has been extensively investigated and reviewed [45] leading in some cases to potential large-scale applications [46]. Moreover, in order to extend the scope of these reactions, a decade ago Lipshutz introduced the concept of designer surfactants [47] (i.e. some new amphiphilic molecules specifically designed and developed for catalytic applications in water) [48] based for example on natural products like α-tocopherol or β-sitosterol as the hydrophobic part, often coupled with hydrophilic poly(ethylene-glycol) methyl ether chains of different length. In the recent years the field of designer surfactants has flourished with many more surfactants that further widen the applicability of micellar catalysis to an even larger range of reactions [49]. Alternatively, metallo-surfactants [50] have been proposed in which the hydrophilic head-group of the surfactant is a ligand for metal centers connected to a long alkyl chain forming metallo-micelles adorned on the surface with many catalytically active metal centers. This all in one approach has the advantage to provide often very good catalytic properties at the expense of a much more complicated synthesis.

24.6  The Catalyst/micelle Interaction

Figure 24.7  Asymmetric epoxidation (a) and Baeyer-Villiger oxidation of ketones (b, c) catalyzed by chiral Pt (a, b) and Co (c) complexes in different micellar media.

24.6  The Catalyst/micelle Interaction The improvement in conversion and selectivity very often observed when moving from a traditional organic solvent to a micellar medium stems from the interactions between the micelle and the catalyst “dissolved” in it. Even if surfactants such as SDS or CTABr have a very simple internal

527

528

24  Platinum Complexes for Selective Oxidations in Water

structure (alkyl chains piled essentially side by side) there will always be a gradient of polarity on going from the micelle surface, where the polar groups are located, toward the core where the hydrophobic part is concentrated. Catalyst positioning in the micelle will depend on a reciprocal affinity basis with portions of different polarity. To examine this issue, we chose three similar platinum catalysts, all cationic and containing diphosphines, already successfully tested in oxidation reactions (see previously), and investigated their positioning with respect to SDS micelles using 2D nuclear Overhauser effect (NOESY) NMR experiments [37–39], by checking the interactions between the diphosphine aryl groups and different portions of SDS alkyl chain (Figure 24.8). As shown in the figure, the bridging hydroxo complex used for sulfoxidation (Figure 24.8 left) was found to interact with the –CH2– group adjacent to the sulfate groups, indicating that the catalyst is essentially sitting on the micelle surface in contact with the anionic groups. Less expected was the catalyst/micelle interaction observed between the Pt pentafluoro derivative used in epoxidation (Figure 24.8 center). Here the NOESY cross peaks indicated an interaction between the phosphine phenyl groups and the terminal –CH3 of the SDS alkyl chain, indicating that the complex is deeply buried in the apolar core of the micelle, somehow surprisingly given its cationic nature. Finally, the bis-aquo complex used for the Baeyer-Villiger oxidation shows clear interactions between the phosphine phenyl groups and the –CH2– groups present along the surfactant chain, thus indicating that it is roughly halfway between the micelle’s surface and core. These experiments demonstrate that even simple supramolecular aggregates like micelles can strongly influence catalyst positioning by mutual interactions and even apparently small differences in the catalyst nature can lead to quite different situations. Hence the constrain of reaction partners in a confined anisotropic space (at variance with solutions) will exert a tighter supramolecular control on reactions and will influence the steric and electronic demand of the transition state in catalyzed reactions, thus leading to improvement in activity and selectivity. For example, in the case of the experiment reported Figure 24.7c, from the extent of the enantioselectivites

Figure 24.8  A graphical representation of the interactions between the complexes indicated on top with SDS micelles. Ellipses on phosphorus represent the phenyl rings.

24.6  The Catalyst/micelle Interaction

observed it can be calculated from the Eyring equation that the ΔΔG≠ values of the diastereomeric transition states when the reaction is carried out in DCE and the same reaction performed in H2O/ Triton-X114 differ by about –6.8 kJ/mol. To exploit sucessfully these effects in micellar catalysis it is clear that the nature and the concentration of the surfactant must be critically selected and carefully optimized. The feature just discussed is typical of enzyme behavior where activity and selectivity issues are amplified by the high variety and complexity of supramolecular interactions occurring inside the enzyme structure. These interactions are capable of driving certain molecules to the active site, lowering the energy of the transition state, and producing specifically only one product enantiomer. They are also the basis for the ability of enzymes to pick only one specific molecule among many others present in biological fluids and lead to its transformation, a property known as substrate selectivity. Can micelles behave similarly? We tested the Diels-Alder reaction between cyclopentadiene and a series of α,β-unsaturated aldehydes differing by the length of the alkyl chain (C1 to C7) adjacent to the double bond, that were fed all together into the reaction mixture. The reaction was catalyzed by a Cr(salen) complex and was tested both in chloroform and in water/SDS (Figure 24.9a). By analyzing the yields of the

Figure 24.9  Substrate selectivity in (a) a Diels-Alder reaction between cyclopentadiene and unsaturated aldehydes of different length and (b) hydrogenation of the same aldehydes with Pd metal. In both cases, the graphs report a comparison between the use of an ordinary solvent and a micellar medium.

529

530

24  Platinum Complexes for Selective Oxidations in Water

various cycloaddition products it was observed that in chloroform differences related to the alkyl chain length of the aldehyde were negligible. On the contrary in water/SDS an increase in yields was observed. Here the cycloaddition product bearing the longest alkyl chain was found to be favored over the one having the shortest alkyl chain by a factor of about 3.5. More striking evidence for substrate selectivity effect was observed in the hydrogenation of the α,β-unsaturated aldehydes with Pd metal particles generated in situ either in tetrahydrofuran (THF) or in water/sodium dodecylbenzenesulfonate (SDBS) micellar medium (Figure 24.9b) [51]. Pd particles produced in either medium had the same size and morphology as was demonstrated by TEM measurements. When the mixture of aldehydes was fed into THF the initial hydrogenation rates for the individual aldehydes decreased on going from the shortest to the longest by a factor of 3.6. The contrary occurred in water/SDBS, with the longest aldehyde being preferred by a factor of 330. As in the Diels-Alder reaction the substrate selectivity rule seems to rely on the higher lipophilicity of longer aldehydes and their higher affinity with the micelle. It should be emphasized that in both cases we performed really competitive experiments because substrates were fed all together into the reaction mixture.

24.7  Environmental Acceptability Evaluation and Possible Industrial Applications The green character of micellar catalysis as well as the advantages with respect to traditional catalysis in solution were evaluated in several reactions by comparing the E-factor of the regular reaction carried out in organic solution and the same reaction performed in a micellar medium. In these evaluations also the final work-up to isolate the products was taken into account. As an example, we consider three C–C bond forming reactions reported by Lipshutz et al. some years ago [52], all catalyzed by Pd complexes (Figure 24.10): (i) a Heck coupling between ethyl acrylate and 3-bromoquinoline; (ii) a Suzuki-Miyaura reaction between phenylboronic acid and 2-iodo-5-bromotoluene; and (iii) a Sonogashira coupling between 3-butynol and a 4-bromobenzoate ester. In all three cases the use of the micellar medium allowed to increase the yields, simplify the reaction procedure and lower the transition metal content. Notably in all three cases the E-factors including work-up were drastically reduced using micellar media based on neutral surfactants instead of the traditional procedure. From what we have reported, it is clear that the industrial application potential of micellar catalysis is quite high. Indeed, the pharmaceutical industry has already adopted this technology either in the total synthesis of some active pharmaceutical ingredients or in the preparation of important intermediates. In particular, research groups at Novartis [53] and Abbvie [54] tested the use of micellar media in a series of reactions of industrial interest, addressing crucial issues such as increasing yields, making product isolation easier, and reducing the amount of catalyst. In 2016 [46b], they reported the total synthesis of an active pharmaceutical ingredient, showing only the functional groups involved in the chemical transformations and hiding the sensitive information of the molecules utilized. The final product synthesis consisted of five different steps and for each of them the product yield, the time necessary to accomplish it and the cost difference were evaluated (Table 24.1) comparing the traditional methodology in organic solvents with the one based on micellar media, both applied on industrial scale. The surfactant technology, based on the nonionic designer surfactant TPGS-750-M in water proved more straightforward and highly advantageous all across the entire synthetic route with environmental as well as economic and productivity benefits. As can be seen from Table 24.1, this resulted in reduction of organic solvent use, reduction of water use, improved yields, milder

24.7  Environmental Acceptability Evaluation and Possible Industrial Applications

Figure 24.10  Three C–C bond forming reactions: comparison of the E-factors between the traditional procedure based on organic solvents and the alternative use of aqueous micellar media. Table 24.1  Synthesis of an active pharmaceutical ingredient: comparison between the traditional organic methodology and the use of micellar media (highlighted in yellow). Step

Conditions

Yield (%)

Cycle time (hr)

1 SNAr

reactant 1.2 eq, i-PrOH/toluene, 80 °C, 5 hours

87

104

reactant 1.0 eq, TPGS-750-M/water, RT, 12–16 hours

75

61

reactant 1.4 eq, Pd, t-AmOH, 85 °C, six hours

=

2 Suzuki Miyaura

reactant 1.0 eq, Pd, TPGS-750-M/water, RT, four to six hours

61

NaOH, MeOH/water, RT, eight hours, MeTHF NaOH, RT, four to six hours

87

53

4 amide bond formation

reactant 1.2 eq, acetonitrile, 50 °C, 10 hours

76

105

reactant 1.0 eq, TPGS-750-M/water, RT, 10–12 h

80

76

HCl, MeOH/water, 0 °C to RT, 16 hours

92

62

HCl, MeOH/water, 0 °C to RT, 16 hours

92

62

Overall

+4%

–38%

24

3 hydrolysis

5 final deprotection

Cost diff

70

137

organic traditional

42.5

469

micellar

48

276

–7%

0% −17%

531

532

24  Platinum Complexes for Selective Oxidations in Water

reaction conditions, reduction of cost, and reduction of cycle time, with the latter being an important issue in the pharmaceutical industry that typically operates in discontinuous, making use of multi-purpose facilities already present in the infrastructure. To quote the authors, “Quantitatively, the differences for some of these virtues approached 50% in favor of surfactant technology”.

24.8  Conclusions Despite the moderate applications of platinum complexes in homogeneous catalysis, the class of very similar electron-poor cationic complexes containing diphosphines that we have investigated over the past forty years has proven useful to elucidate and give experimental support to a series of concepts that can be used as guidelines to identify potential catalysts for oxidation reactions using hydrogen peroxide as oxidants. These concepts are: (i) soft Lewis acidity that makes possible the (ii) electrophilic oxidation, thus opening the way to the oxidation of electronpoor olefins and ketones with unprecedented activity, selectivity (including enantio-), product yields, without any waste of oxidant, and under very mild conditions (room temperature, in air). Additional advantages of these Pt complexes are: (iii) the possibility to tune the steric and electronic properties of the metal center through an appropriate choice of commercially available diphosphine ligands, including chiral ones to perform enantioselective transformations; and (iv) full compatibility with water necessary to operate with hydrogen peroxide. The latter property allowed these Pt complexes (v) to be used in aqueous media where they proved to be ideal to test the different features of micellar catalysis, thus removing organic solvents at least as reaction media. Commercially available surfactants can be used as simple auxiliaries to create a separate phase in water, made of self-assembled devices (micelles) that can be seen as dispersed nanoreactors, in which to dissolve catalyst, reactants, products. Micellar media can improve activity and selectivity and, in many cases, make it possible catalyst separation and recycling. As we have shown, the gradient of polarity present in micelles can result in different catalyst and substrate positioning. This internal anisotropy is typical of enzymes, involving not only areas of different polarity but also hydrogen bonding and other weak interactions. The consequence of that is an amplification of the free energy differences of the possible transition states of a reaction, thus influencing its outcome and driving selectivity at any level. The different chemical affinity of different substrates for micelles, mainly driven by differences in hydrophobic properties, enable traditional metal catalysts to display unusual substrate selectivity [55], an important property in common with enzymes and also with zeolites, although on a different selection basis. Practical examples have shown that the use of micellar media can significantly reduce the E-factor in organic transformations. However this is not a simple green chemistry advance, but rather a new technological approach that can lead to significant benefits across the entire synthetic route, thus paving the way to industrial applications.

References 1 BASF. (2009). New BASF and Dow HPPO plant in Antwerp completes start-up phase. https:// www.basf.com/ru/ru/media/news-releases/2009/03/p-09-154.html (accessed 19 April 2023). 2 Chemical Online. (2000). Sumitomo to commericalize new caprolactam process. https://www. chemicalonline.com/doc/sumitomo-to-commercialize-new-caprolactam-pro-0001 (accessed 19 April 2023).

References

3 (a) Strukul, G., Ros, R., and Michelin, R.A. (1982), Inorg. Chem. 21: 495–500; (b) Strukul, G., Michelin, R.A., Orbell, J.D., and Randaccio, L. (1983). Inorg. Chem. 22: 3706–3713; (c) Strukul, G. and Michelin, R.A. (1984). JCS Chem. Commun. 1538–1539; (d) Strukul, G. and Michelin, R.A. (1985). J. Am. Chem. Soc. 107: 7563–7569. 4 Zanardo, A., Pinna, F., Michelin, R.A., and Strukul, G. (1988). Inorg. Chem. 27: 1966–1973. 5 Roussel, M. and Mimoun, H. (1980). J. Org. Chem. 45: 5381–5383. 6 (a) Suresh, C.H. and Koga, N. (2001). J. Phys. Chem. A 105: 5940–5944; (b) Bond, G.C. (2000). J. Mol. Catal. A-Chem. 156: 1–20. 7 For some example of homologous diphosphine complexes structural data, see: (a) Maddox, A.F., Rheingold, A.L., Golen, J.A. et al. (2008). Inorg. Chim. Acta 361: 3283–3293; (b) Ghent, B.L., Martinak, S.L., Sites, L.A. et al. (2007). J. Organom. Chem. 692: 2365–2374. 8 (a) Langford, C.H. and Gray, H.B. (1965). Chapter 2. Square-Planar Substitutions, pag. 18, In: Ligand Substitution Processes (1966) W.A. Benjamin, Inc. Ed., New York (USA). In: Ligand Substitution Processes. (ed. C.H. Langford and H.B. Gray), 18. New York: W.A. Benjamin, Inc; (b) Anderson, G.K. and Cross, R.J. (1980). Chem. Soc. Rev. 9: 185–215; (c) Appleton, T.G., Clark, H.C., and Manzer, L.E. (1973). Coord. Chem. Rev. 10: 335–422; (d) Chval, Z., Sip, M., and Burda, J.V. (2008). J. Comput. Chem. 29: 2370–2381. 9 (a) Aye, K.T., Vittal, J.J., and Puddephatt, R.J. (1993). J. Chem. Soc., Dalton Trans. 1835–1839; (b) Deacon, G.B., Lawarenz, E.T., Hambley, T.W. et al. (1995). J. Organomet. Chem. 493: 205–213; (c) Rashidi, M., Nabavizadeh, M., Hakimelahi, R., and Jamali, S. (2001). J. Chem. Soc., Dalton Trans. 3430–3434. 10 (a) Uson, R., Fornies, J., Espinet, P., and Alfranca, G. (1980). Synth. React. Inorg. Met.-Org. Chem. 10: 579–50-. (b) Uson, R., Fornies, J., Martinez, F., and Tomas, M. (1980). J. Chem. Soc., Dalton Trans. 888–894. 11 Criegee, R. (1948). Justus Liebigs Ann. Chem. 560: 127–135. 12 Del Todesco Frisone, M., Pinna, F., and Strukul, G. (1983). Organometallics 12: 148–156. 13 (a) Scarcia, V., Furlani, A., Longato, B. et al. (1988). Inorg. Chim. Acta 153: 67–70; (b) Gavagnin, R., Cataldo, M., Pinna, F., and Strukul, G. (1998). Organometallics 17: 661–667; (c) Sgarbossa, P., Scarso, A., Michelin, R.A., and Strukul, G. (2006). Organometallics 26: 2714–2719. 14 (a) Michelin, R.A., da Silva, M.F.C.G., and Pombeiro, A.J.L., (2001). Coord. Chem. Rev. 218: 75–112; (b) Belluco, U., Michelin, R.A., Uguagliati, P., and Crociani, B. (1983). J. Organomet. Chem. 250: 565–587. 15 Pizzo, E., Sgarbossa, P., Scarso, A. et al. (2006). Organometallics 25: 3056–3962. 16 Bond, G.C. (1987). Heterogeneous Catalysis: Principles and Applications, 2e. Oxford: Clarendon Press. 17 Michelin, R.A., Pizzo, E., Sgarbossa, P. et al. (2005). Organometallics 24: 1012–1017. 18 Wursche, R., Debaerdemaeker, T., Klinga, M., and Rieger, B. (2000). Eur. J. Inorg. Chem. 2063–2070. 19 (a) Cataldo, M., Nieddu, E., Gavagnin, R. et al. (1999). J. Mol. Catal. A Chemical 142: 305–316; (b) Nieddu, E., Cataldo, M., Pinna, F., and Strukul, G. (1999). Tetrahedron Lett. 40: 6987–6990; (c) Battaglia, L., Pinna, F., and Strukul, G. (2001). Can. J. Chem. 79: 621–625. 20 Pignat, K., Vallotto, J., Pinna, F., and Strukul, G. (2000). Organometallics 19: 5160–5167. 21 Scarso, A., Colladon, M., Sgarbossa, P. et al. (2010). Organometallics 29: 1487–1497. 22 Colladon, M., Scarso, A., Sgarbossa, P. et al. (2007). J. Am Chem. Soc. 129: 7680–7689. 23 Colladon, M., Scarso, A., Sgarbossa, P. et al. (2006). J. Am Chem. Soc. 128: 14006–14007. 24 Sinigalia, R., Michelin, R.A., Pinna, F., and Strukul, G. (1987). Organometallics 6: 728–734. 25 (a) Gusso, A., Baccin, C., Pinna, F., and Strukul, G. (1994). Organometallics 13: 3442–3451; (b) Paneghetti, C., Gavagnin, R., Pinna, F., and Strukul, G. (1999). Organometallics 18: 5057–5065.

533

534

24  Platinum Complexes for Selective Oxidations in Water

26 (a) Anastas, P.T. and Warner, J.C. (1998). Green Chemistry: Theory and Practice. New York: Oxford University Press; (b) Anastas, P.T. and Kirchhoff, M. M. (2002). Acc. Chem. Res. 35: 686–695; (c) Bolm, C., Beckmann, O., and Dabard, O.G.A. (1999). Angew. Chem. Int. Ed. 37: 1198–1209. 27 (a) Seyler, C., Capello, C., Hellweg, S. et al. (2006). Ind. Eng. Chem. Res. 45: 7700–7709; (b) Capello, C., Fischer, U., and Hungerbühler, K. (2007). Green Chem. 9: 927–934. 28 (a) Cortes-Clerget, M., Yu, J., Kincaid, J.R.A. et al. (2021). Chem. Sci. 12: 4237–4266; (b) Lipshutz, B.H., Gallou, F., and Handa, S. (2016). ACS Sust. Chem. Eng. 4: 5838–5849. 29 Sheldon, R.A. (2005). Green Chem. 7: 267–278. 30 Joó, F. (2001). Aqueous Organometallic Catalysis. Dordrecht: Kluwer. 31 Rideout, D.C. and Breslow, R. (1980). J. Am. Chem. Soc. 102: 7816–7817. 32 (a) Gawande, M.B., Bonifácio, V.D.B., Luque, R. et al. (2013). Chem. Soc. Rev. 42: 5522–5551; (b) Simon, M.-O. and Li, C.-J. (2012). Chem. Soc. Rev. 41: 1415–1427. 33 (a) Lindström, U.M., (2006). Angew. Chem., Int. Ed. 45: 548–551; (b) Adams, D., Dyson, P., and Tavener, S. (2004). Chemistry in Alternative Reaction Media. Chichester: Wiley; (c) Li, C.J. (1993). Chem. Rev. 93: 2023–2035; (d) Li, C.J. (2005). Chem. Rev. 105: 3095–3166; (e) Pirrung, M.C. (2006). Chem. Eur. J. 12: 1312–1317; (f) Narayan, S., Muldoon, J., Finn, M.G. et al. (2005). Angew. Chem., Int. Ed. 44: 3275–3279. 34 Pliny the Elder. Naturalis Historia, 28, ch. 47. 35 (a) Breslow, R., Maitra, U., and Rideout, D., (1983). Tetrahedron Lett. 24: 1901–1904; (b) Grieco, P.A., Garner, P., and He, Z. (1983). Tetrahedron Lett. 24: 1897–1900; (c) Dey, A. and Patwari, G.N. (2011). J. Chem. Sci. 123: 909–918. 36 Zhao, M., Wang, H.-B., Ji, L.-N., and Mao, Z.-W. (2013). Chem. Soc. Rev. 42: 8360–8375. 37 Scarso, A. and Strukul, G. (2005). Adv. Synth. Catal. 347: 1227–1234. 38 Colladon, M., Scarso, A., and Strukul, G. (2007). Adv. Synth. Catal. 349: 797–801. 39 Cavarzan, A., Bianchini, G., Sgarbossa, P. et al. (2009). Chem. Eur. J. 15: 7930–7939. 40 Bianchini, G., Cavarzan, A., Scarso, A., and Strukul, G. (2009). Green Chem. 11: 1517–1520. 41 Trentin, F., Chapman, A.M., Scarso, A. et al. (2012). Adv. Synth. Catal. 354: 1095–1104. 42 Gottardo, M., Scarso, A., Paganelli, S., and Strukul, G. (2010). Adv. Synth. Catal. 352: 2251–2262. 43 Cavarzan, A., Scarso, A., and Strukul, G. (2010). Green Chem. 12: 790–794. 44 Trentin, F., Scarso, A., and Strukul, G. (2011). Tetrahedron Lett. 52: 6978–6981. 45 (a) Sorella, G.L., Strukul, G., and Scarso, A. (2015). Green Chem. 17: 644–683; (b) Gallou, F. (2020). Chimia 74: 538–548; (c) Sar, P., Ghosh, A., Scarso, A., and Saha, B. (2019). Res. Chem. Int. 45: 6021–6041; (d) Scarso, A. and Strukul, G. (2019). In: Green Synthetic Processes and Procedures (ed. R. Ballini), 268–288. London: The Royal Society of Chemistry. 46 (a) Kar, S., Sanderson, H., Roy, K. et al. (2022). Chem. Rev. 122: 3637–3710; (b) Gallou, F., Isley, N.A., Ganic, A. et al. (2016). Green Chem. 18: 14–19. 47 Lipshutz, B.H. and Ghorai, S. (2012). Aldrichimica Acta 45: 3–16. 48 (a) Lipshutz, B.H. (2018). Curr. Opin. Green Sust. Chem. 11: 1–8; (b) Lipshutz, B.H., Ghorai, S., and Cortes-Clerget, M. (2018). Chem. Eur. J. 24: 6672–6695; (c) Lipshutz, B.H. (2017). J. Org. Chem. 82: 2806–2816. 49 Lorenzetto, T., Berton, G., Fabris, F., and Scarso, A. (2020). Catal. Sci. Technol. 10: 4492–4502. 50 Mehta, S. and Kaur, R. (2022). Metallosurfactants: From Fundamentals to Catalytic and Biomedical Applications. Weinheim: Wiley-VCH. 51 La Sorella, G., Canton, P., Strukul, G., and Scarso, A. (2014). ChemCatChem. 6: 1575–1578. 52 Lipshutz, B.H., Isley, N.A., Fennewald, J.C., and Slack, E.D. (2013). Angew. Chem. Int. Ed. 52: 10952–10968.

References

53 (a) Li, X., Thakore, R.R., Takale, B.S. et al. (2021). Organ. Lett. 23: 8114–8118; (c) Yu, T.Y., Pang, H., Cao, Y. et al. (2021). Angew. Chem. Int. Ed. 60: 3708–3613. 54 (a) Petkova, D., Borlinghaus, N., Sharma, S. et al. (2020). ACS Sustainable Chem. Eng. 8: 12612– 1267; (b) Sharma, S., Ansari, T.N., and Handa, S. (2021). ACS Sustainable Chem. Eng. 9: 12719–12628. 55 (a) Lindbäck, E., Dawaigher, S., and Wärnmark, K. (2014). Chem. Eur. J. 20: 13432–13481; (b) Otte, M. (2016). ACS Catal. 6: 6491–6510.

535

537

25 The Role of Water in Reactions Catalyzed by Transition Metals A.W. Augustyniak and A.M. Trzeciak University of Chemistry, Faculty of Chemistry, 14 F. Joliot-Curie, Wrocław, Poland

25.1  Water as a Solvent in Organic Reactions Water is the most environmentally friendly solvent, cheap and easily available. In addition, because water is non-toxic and non-flammable, it is highly recommended as a medium for laboratory and industrial syntheses [1]. However, the high polarity of water complicates the performance of organic reactions because most organic compounds are non-polar and insoluble in water. In this context, the mediating effect of water in the course of organic reactions should be mentioned [2, 3]. In 1980, Rideout and Breslow reported an acceleration of the Diels-Alder reaction between nonpolar substrates in the presence of water [4, 5]. A similar beneficial effect was confirmed next by other authors for Claisen rearrangement [6]. In 2005, Sharpless introduced the name on water for the reaction in which insoluble reactants are stirred in aqueous suspension [7]. He also mentioned the key role of the phase boundary in the acceleration of organic reactions in these conditions. Organic reactions that benefit from performance on water were reviewed by Fokin et. al [8]. Further interpretations were presented by Jung and Marcus, who focused on the hydrophobic interfacial structure of water and proposed models of on-water and aqueous catalysis [9]. In this kinetic approach, the highest values of rate constants were noted for reactions occurring on the catalyst surface. When reaction takes place at an oil-water interface, the formation of H-bonds with the reactant (B) and with the transition state (AB) facilitated the reaction course (Figure 25.1). In contrast, only a moderate shortening of the reaction time was observed for aqueous homogeneous reactions (Figure 25.1). This could be related to the presence of the H-bond network formed by water molecules around the reactants. Kühne et al. supported some conclusions of Jung and Marcus for the Diels-Alder reaction under on water conditions using the molecular dynamics of the Car-Parrinello method [10]. Calculations of the free energy profiles indicated the role of the dangling OH-bonds at a water-oil interface, whereas the stabilization of the transition state by H-bonding only had a small influence on the free energy barrier. Recently Kitanosono and Kobayashi reviewed different aspects of on water organic reactions as well as reactions performed with and without a catalyst [11]. They proposed a new on water model Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 2, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

538

25  The Role of Water in Reactions Catalyzed by Transition Metals

Figure 25.1  The kinetic data for the on water and homogeneous aqueous reactions.

Figure 25.2  Orientation of water molecules on the acidic surface.

for acidic surfaces considering the partial polarization of water molecules. In this model, three water layers were distinguished according to the orientation of water dipoles. Accordingly, the most important for the reaction course was the orientation of molecules at an interface and the formation of hydrogen bonds (Figure 25.2). The authors mentioned that their model can better explain a decrease in the energy of the transition state as a result of the formation of H– bonds between the catalyst and reactants. The promotional role of water in heterogeneous catalysis was discussed by Li et al., who considered experimental and theoretical aspects of such systems and distinguished different situations describing water assistance [12]. Thus, water molecules accelerate the catalytic reaction due to the stabilization of reactants and intermediates on the catalyst surface via hydrogen bonds. Water can also be involved in H-transfer processes, initiated by the formation of hydrogen bonds with donor or acceptor molecules. In this case, the dehydrogenation of the HA reactant is facilitated by water, which mediates the H-atom transfer to the surface.

25.2  The Role of Water in Heterogeneous Catalytic Systems

Davies discussed the importance of water using examples of the heterogeneously catalyzed oxidation of CO and alcohols or Fischer-Tropsch synthesis [13]. Following some ideas of Roberts [14], he pointed out the contribution of water in the creation of hydroxyls, proton transfer and the stabilization of reactive intermediates. Micek-Ilnicka reviewed the physicochemical properties of heteropolyacid (HPA)-water systems [15]. HPAs are strong mineral acids and are often used as acid-base or redox catalysts, also in industry. HPAs are typically used in hydrated form and water molecules present in the structure can interact with the terminal oxygen of the HPA anion or they can form hydrogen bonds to OH groups. The formation of H5O2+ protonated clusters can also be considered. It is worth noting that water influences acid strength and the number of proton centers in HPAs, which is important for their catalytic activity. For example, the rate of acid-base reactions is related to the concentration of non-solvated protons involved in hydrogen bonds. Moreover, water influences the secondary structure of HPAs and competes with other polar molecules. It can be concluded that the understanding of the accelerating effect of water molecules in catalytic reactions is quite well developed for heterogeneous systems [16]. Thus, water molecules can stabilize reactants and intermediates on the catalyst surface via the H-bond (a solvation-like effect). In this case, water can also be involved in H-transfer processes, but it is not consumed. Alternatively, water may be decomposed on the catalyst surface with the formation of OH and O species, which contribute next in the catalytic reaction, for example in the oxidation of CO to COOH. The formation of �OOH and �OH radicals may occur in the reaction of O2 + H2O. Water can contribute to other specific processes such as the removal of carbon from the catalytic surface, the blocking of active sites or surface reconstruction. In homogeneous systems similar processes can occur with the contribution of water coordinated as a ligand to the active catalyst. Moreover, similarly as in the case of HPAs, the second coordination sphere created by hydrogen bonds can have an additional influence on the access of substrates to the active center and reaction efficiency. In the nextsections, selected examples, which present water as an active component of catalytic systems, including both heterogeneous and homogeneous ones, are discussed.

25.2  The Role of Water in Heterogeneous Catalytic Systems 25.2.1  The Transformations of Furfuryl Derivatives The decarbonylation of 5-hydroxymethylfurfural to furfuryl alcohol was performed with a Pd/ Al2O3 catalyst in a mixed solvent system containing water (Figure 25.3) [17]. The presence of 28 wt% of water caused an increase of selectivity to furfural to 97% due to the suppression of side reactions, such as hydrolysis or etherification. However, at a higher amount of water, 40 wt%, conversion decreased while the contribution of dehydrogenation/polymerization increased. The reaction mechanism was studied using the density functional theory (DFT) method and it was found that the formation of hydrogen bonds between furfuryl alcohol and water hinders further reaction. Moreover, the formation of hydrogen bonds between water and OH groups on the Al2O3 surface decreased the activity of Pd/Al2O3 and retarded side reactions of furfuryl alcohol. The effect of water was disclosed in the ring opening of furfuryl alcohol to 1,2-pentanediol catalyzed by immobilized Pd catalysts, Pd/CeO2, Pt/MgO, and Pd/La-Al2O3 (Figure 25.4) [18]. This reaction presents an important step during the transformation of biomass to valuable chemicals.

539

540

25  The Role of Water in Reactions Catalyzed by Transition Metals

Figure 25.3  The decarbonylation of 5-methoxymethylfurfural.

Figure 25.4  The transformations of furfuryl alcohol to 1,2-pentanediol.

By using DFT calculations, several interactions of water molecules with a support and an active metal phase were identified. An experiment performed with H218O directly evidenced the presence of 18O in CH3CH2CH2CH(18OH) ion originated from 1,2-pentanediol. This is direct proof for the involvement of water in this reaction.

25.2.2  Oxidation and Deoxygenation The activity of palladium catalysts in methane combustion is inhibited by water and different explanations of this effect were proposed [19] One of the recent interpretations is based on limited methane adsorption on the active catalyst surface due to the formation of Pd(OH). Moreover, in the presence of water, the activation of C–H bonds in the methyl group was inhibited. The deprotonation of water on oxygen vacancies also has a negative influence on the formation of PdO and, in consequence, on the combustion of methane. Water is formed by the condensation of surface hydroxyls, while their desorption generates oxygen vacancies (Ov). In these conditions the formation of Pd-OH species is preferred.

Pd−O + H2O + Pd−O V  2 Pd−OH

Another example of water contribution was reported for the Fe-catalyzed hydrodeoxygenation of phenol to benzene (Figure 25.5) [20]. It was evidenced that water decreased the activation barrier due to the formation of Bronsted acid sites on the Fe(110) surface during water splitting. The formed acid centers are involved in proton transfer and promote the formation of benzene. The role of water in the catalytic oxidation of hydrocarbons was recently reviewed by Ovchinnikova et al. [21]. Some important aspects related to the presence of water in the catalytic systems were pointed out. One of them is the formation of new Bronsted acid sites on the catalyst surface. These acid sites influenced the reaction course and, in particular, reaction selectivity. This is because they suppressed the formation of some intermediates. Water can also retard the reoxidation process by substituting oxygenate precursors on the catalyst surface. Moreover, water influences the oxidation of hydrocarbons because it competes with reaction products and intermediates for catalytically active sites. It can also facilitate the desorption of the product. As a source of OH groups, water moderates reaction kinetics.

25.2  The Role of Water in Heterogeneous Catalytic Systems

Figure 25.5  The hydrodeoxygenation of phenol.

An interesting example of the enhancement of the reaction rate in the presence of water was reported by Qiu et al. [22]. They studied the oxidation of different aliphatic and aromatic alcohols containing a sulfur or nitrogen atom catalyzed by an Ru catalyst supported on a carbon nanotube (Ru/CNT) in an oil/water emulsion. The beneficial effect of water was explained by the formation of emulsion droplets with an Ru catalyst reacting at the interfaces. An additional advantage of this system is its excellent selectivity to desired aldehydes and the easy recycling of the catalyst. Mullen et al. reviewed the effect of surface water present on the catalytic activity of gold [23]. They found that the interaction of water with a single crystal gold surface is different in the presence and absence of oxygen. When the surface was covered with atomic oxygen, dissociative adsorption occurred and �OOH and �OH radicals were formed. They enhanced the selectivity of propene epoxidation, increased the efficiency of CO oxidation and the water-gas shift reaction. The solvation effect of water can result in the acceleration of the catalytic reaction due to the stabilization of reactants and intermediates on the surface [12]. The formation of hydrogen bonds, without the splitting of O–H, plays the main role here. Theoretical calculations performed for the reaction of O2 with H2O on the surface of Au10 and Au38 clusters show that O2 is activated and formed the HOO� species in reaction with hydrogen from the H2O molecule.

O2 + H2O → ⋅OOH +⋅OH

The activation of water on the Au cluster occurred in several steps involving the elongation of the O–O bond in O2 (from 1.27 Å to 1.38 Å) and an O–H bond in H2O (from 1.0 Å to 1.42 Å) [24]. As a result HOO� and ⋅OH radicals were formed. This result can be helpful in explaining the activating role of water in propene epoxidation with O2 and H2O on the Au7/α-Al2O3 catalyst. The overall equation for propene epoxidation with water being a catalytic promoter was:

2 C3H6 + O2 + H2O → 2PO + H2O (PO = propene oxide)

Interesting DFT calculations were performed for the conversion of H2O to H2O2 in the presence of FeIVO2+, which occurred according to the scheme:

541

542

25  The Role of Water in Reactions Catalyzed by Transition Metals

Figure 25.6  Oxidation of water with high-spin Fe(IV)-oxo species.

The reaction proceeded in two steps, namely the formation of the O–O bond between the OH2O (water oxygen) and Ooxo ligand followed by the generation of H2O2 (Figure 25.6). In the first step contributed [(H2O)5Fe(IV)O]2+ which coordinated the additional water molecule to the Ooxo ligand. Next, the transition state was formed with O–O at a distance of 1.76 Å. Two electrons were transferred from Ooxo to Fe and, consequently, Fe(IV) was reduced to Fe(II) and HOO– was generated. Shortening the O–O distance to 1.45 Å indicated the formation of the peroxide fragment (in H2O2 the O–O distance is 1.47 Å). In the final step of the reaction, the H2O2 molecule was formed due to the H transfer from water to the HOO– group with the O–O bond shortened to 1.35 Å. At this stage, Fe2+ formed the [Fe(H2O)5]2+ complex. These studies explained the main steps of water oxidation and indicated the special role of a second solvation sphere in stabilizing the Fe–OOH intermediate as well as in proton transfer.

25.2.3  Arylcyanation and C–C Cross-Coupling The arylcyanation of aryl bromides is a method used in the synthesis of aromatic cyanides, including biologically active compounds. It is generally assumed that this reaction is catalyzed by soluble palladium compounds in a homogeneous system. However, recently Fairlamb et al. have shown that water originating from K4[Fe(CN)6].3H2O used as a cyanating agent, changed the reaction mechanism to a heterogeneous one [25]. A heterogeneous pathway is based on the activity of Pd(0) in the form of bimetallic species [LnPd]2 or Pdn aggregates (Figure 25.7). An excess of H2O influenced the equilibrium between these forms and the catalytic cycle based on LnPd species leached from Pdn, which became dominating. In addition, the contribution of Pdn aggregates in the reaction course was confirmed by the Hg(0) test. The amount of water present in the system influenced arylcyanation efficiency and higher turnover frequency (TOF) and turnover number (TON) values were obtained for K4[Fe(CN)6] containing less than 220 ppm of water. A higher amount of water, >4000 ppm, introduced with K4[Fe(CN)6].3H2O caused a decrease in the reaction yield. A solvent-free Suzuki-Miyaura reaction was successfully performed under conventional heating using PdNPs/MWCNT (multi-walled carbon nanotubes) [26]. In this original solid-state system, a small amount of water formed during the trimerization of phenyl boronic acid plays an important role (Figure 25.8). In agreement with this, calcinated phenylboronic acid as well as KBPh4 or PhBF3K, which cannot release water, provided a much lower conversion. The presence of water in the catalytic system facilitates the diffusion of substrates to the catalyst and mass transfer. Water can also be the source of basic nucleophiles, such as an OH– anion, which can substitute the X ligand in the coordination sphere of Pd to form an R-Pd-OH intermediate. More data concerning the reaction course were collected from a SEM examination. In the reaction mixture, particles of a diameter of c. 100 nm formed while grinding reagents were identified. Interestingly, the coupling product present inside these particles was easily isolated by sublimation. At this stage, some

25.2  The Role of Water in Heterogeneous Catalytic Systems

Figure 25.7  Mechanism of arylcyanation.

Figure 25.8  Contribution of water in the solid-phase Suzuki-Miyaura reaction.

morphology changes occurred and new pores were formed due to the removal of the product from particles. However, the catalyst, Pd/MWCNT, was reused with the same activity without Pd loss. The presented procedure enables the minimization of Pd leaching due to solvent elimination. A direct H-atom transfer from H2O to alkenes and alkynes was reported for transfer hydrogenation with the B2(OH)4 and Pd/C catalyst [27]. According to the proposed mechanism, supported by experimental data, in the first step, B2(OH)4 coordinates to Pd via an oxidative addition pathway with the splitting of the B–B bond. Next, water bonded to B(OH)2 is split and acts as a source of the H ligand. Further steps of the catalytic cycle were analogous to other hydrogenation cycles and led to the Pd-alkyl intermediate. In the end the water molecule reacted with B(OH)2 producing the H ligand and B(OH)3 (Figure 25.9). In this reaction, water acted as the only source of hydrogen. Kamal et. al., discussed the role of the hydrophobic effect on the catalyst surface for water mediated Heck and Ulmann coupling with Pd nanoparticles supported on amphiphilic carbon spheres

543

544

25  The Role of Water in Reactions Catalyzed by Transition Metals

Figure 25.9  Hydrogenation of trans-stilbene by water and B2(OH)4.

Figure 25.10  The Heck (left) and the Ullmann (right) coupling over Pd@CSP in water.

(CPS) prepared from glucose (Figure 25.10) [28]. Catalytic tests performed for Heck coupling in different solvents revealed the advantages of water, which was the best solvent considering catalyst activity, selectivity and stability. These advantages were related to the Breslow effect based on the amphiphilic structure of carbon support. The presence of carboxylic groups on its surface additionally limits the aggregation of Pd nanoparticles.

25.2.4  Hydrogenation Experimental and theoretical studies on the mechanism of the hydrogenation of cinnamylaldehyde catalyzed by a Pt3Fe nanocatalyst supported on a carbon nanotube (CNT) showed the involvement of water in this process [29]. It was found that the water molecule forms a bridge enabling hydrogen exchange between aldehyde and Pt sites (Figure 25.11). Synergistic interactions between water and the catalyst during the hydrogenation process resulted in an increase of both aldehyde conversion and selectivity to alcohol. In particular, the water-mediated pathway was energetically favorable, over the water-free one, due to a lower energy barrier. Thus, water mediates a hydrogen exchange and acts as a promoter of selective hydrogenation. The hydrogenation of aromatic acids catalyzed by Pd supported on a carbon nanofiber in water occurred selectively in the aromatic ring [30]. The carboxylic acid groups were not affected and cyclohexanecarboxylic acids were formed as the only products as final products. This high selectivity of hydrogenation was explained by the interaction between water, the substrate and the metal surface. It was assumed that polar OH groups of acid form hydrogen bonds with water molecules blocking their interaction with the metal center. In such a case, the interaction of the aromatic ring with the active metal was preferred and hydrogenation occurred in the ring. An interesting effect of water was observed during the hydrogenation of acetophenone catalyzed by Pd/S-DVB (S-DVB = styrene/divinylbenzene copolymer) in MeOH/H2O and 2-PrOH/H2O mixtures [31]. The reaction proceeded efficiently in pure water or in pure alcohol, while in a mixture of solvents the yield was significantly lower. For example, 91% of acetophenone was converted in pure MeOH, while in the mixture 1.5 MeOH/1.5 H2O conversion decreased to 24% only (Figure 25.12). It was assumed that hydrated clusters of alcohol formed in the mixed solutions interacted with the

25.2  The Role of Water in Heterogeneous Catalytic Systems

Figure 25.11  The hydrogenation of cinnamaldehyde over Pt3Fe@CNT catalyst.

Figure 25.12  The hydrogenation of acetophenone in different solvents.

porous catalyst blocking the access of substrates to Pd active centers. As a result, catalytic efficiency decreased. This inhibiting effect was evident in MeOH/H2O and 2-PrOH solutions, while in BuOH/ H2O it was significantly weaker due to a weaker hydration of bigger BuOH clusters.

25.2.5  Hydroformylation The asymmetric hydroformylation of vinyl acetate catalyzed by the Rh/DNA catalyst in the presence of chiral diphosphine in water proceeded with significantly higher enantioselectivity when compared to the same reaction with the Rh(acac)(CO)2 catalyst [32]. For instance, enantioselectivity (ee) achieved the value of 49% with Rh/DNA + (R)-BINAP and 11% with Rh(acac) (CO)2 + (R)-BINAP (Figure 25.13). Importantly, the addition of free DNA to the reaction catalyzed by Rh(acac)(CO)2 caused an increase of ee to 35%. It was concluded that the synergistic interaction of DNA support with chiral phosphine has an improving effect on the reaction course. The contribution of water in the stabilization of the active species should also be considered. This catalytic system also works well for the asymmetric hydroformylation of styrene.

545

546

25  The Role of Water in Reactions Catalyzed by Transition Metals

Figure 25.13  The asymmetric hydrogenation of vinyl acetate catalyzed by Rh/DNA in water (regioselectivity to 2-acetoxypropanal).

25.2.6  Catalytic Reactions with MOF-based Catalysts in an Aqueous Medium For more than two decades, metal-organic frameworks (MOFs) and covalent organic frameworks (COFs) have been considered exciting platforms for applications in heterogeneous catalysis [33, 34]. Their well-defined structures, high surface areas, and tunable pores create a unique environment around catalytic centers. In the context of catalysis performed in an aqueous medium, it is important to emphasize that the hydrophobic/hydrophilic character of MOFs can accumulate selected substrates around catalytic centers and accordingly influence reaction selectivity. A number of catalytic reactions have been developed using MOFs and COFs in heterogeneous catalysis, including reactions in an aqueous phase [35, 36]. However, still less work has been devoted to specific interactions with water, which can influence the reaction course. In this chapter, special attention was paid to the function of water as a reactant or co-catalyst. 25.2.6.1  Cross-Coupling Reactions

Table 25.1 compiles selected reports concerning the use of Pd-MOFs as catalysts for C–C coupling reactions such as Suzuki–Miyaura, Heck, and Sonogashira, in which water was used as a solvent. The use of water in these reactions is strongly recommended, not only from an environmental point of view but also because of the application of their products in the pharmaceutical industry. However, in most cases, the influence of water on the reaction has not clearly been stated. One of the first reports on a Suzuki reaction in a water medium was presented for Pd@MIL101-Cr as a catalyst (Figure 25.14). Water in this system was primarily a green solvent; however, there was also some information about its other role. It was noticed that the combination of water and a NaOMe base led to the generation of a small amount of methanol and NaOH. This system was able to achieve up to 78% conversion of 4-chloroanisole, while much less, 63% of the product was obtained in a reaction performed in a methanol/water solution in the presence of NaOH. These results confirmed that the formation of MeOH and NaOH in situ increased the efficiency of this reaction [37]. Another MOF material, Pd@UiO-66-NH2, used in the Suzuki reaction in water formed 53% of 4-methoxybiphenyl when bromobenzene was used as a substrate [39]. The proposed reaction mechanism indicates that water participated in the reaction of aryl boronic acid with K2CO3. An anionic species, R-B(OH)3–, formed in this reaction, is considered the most reactive species toward the transmetalation step according to previous works [48]. Martín-Matute et al. used the catalyst Pd@MIL-101-NH2-Cr for Suzuki–Miyaura cross-coupling in water at room temperature (Figure 25.15) [38]. In this case water affected the reactivity, and electron-rich substrates reached full conversion while the electron-poor compound was converted

25.2  The Role of Water in Heterogeneous Catalytic Systems

Table 25.1  C–C coupling reactions catalyzed by MOFs-based Pd catalysts in water. Entry

1

Catalysts

Reaction type

Ref.

Pd/MIL-101-Cr

Suzuki-Miyaura

[37]

2

Pd@MIL-101(Cr)-NH2

Suzuki-Miyaura

[38]

3

Pd/UiO-66-NH2

Suzuki-Miyaura

[39]

4

Pd(0)/MCoS-1

Suzuki-Miyaura, Sonogashira

[40]

5

Pd−NHC−MIL-101(Cr)

Suzuki-Miyaura

[41]

6

Pd@[Ni(H2BDP-SO3)2]

Suzuki-Miyaura

[42]

7

Pd/Y-MOF

Suzuki-Miyaura, Sonogashira

[43]

8

PdII@Cu(BDC)/2-Py-SI

Heck

[44]

9

Pd@MIL-101

Heck Suzuki

[45]

10

Pd(II)-UiO-67

Heck Suzuki

[46]

11

Pd@MIL-101-NH2 Pd@MIL-88B-NH2

Heck

[47]

Figure 25.14  The Suzuki-Miyaura coupling of 4-chloroanisole catalyzed by Pd@MIL-101-Cr.

Figure 25.15  The Suzuki-Miyaura coupling of bromobenzene derivatives catalyzed by Pd@MIL-101-Cr.

into the corresponding biphenyl in only 84%. These differences were explained by the different solubility of these compounds in water. Zhou investigated the mechanism of the Heck C–C coupling catalyzed by Pd@MIL-101-NH2 and Pd@MIL-88B-NH2 in the water medium (Figure 25.16) [47]. They demonstrated that in the absence of H2O, Pd remained in an oxidation state +2. However, in the presence of water and olefin, Pd was reduced to Pd(0), which represents the active species that initiates the catalytic cycle. COFs materials possess well-defined porous skeletons which can bind metal nanoparticles. Additionally, COFs are insoluble in organic solvents and in water, and are attractive supports for heterogeneous catalysts. Surprisingly, they have only been scarcely used for C-C coupling thus far. The first example of using COF material in the Suzuki-Miyaura reaction in an aqueous medium was presented by Esteves et al. (Figure 25.17) [49]. They used Pd(OAc)2@COF-300 as a catalyst in

547

548

25  The Role of Water in Reactions Catalyzed by Transition Metals

Figure 25.16  Formation of the catalytically active form of Pd catalyst.

Figure 25.17  Influence of water on the Suzuki-Miyaura reaction rate.

reaction between bromobenzene and phenylboronic acid in methanol and obtained 90% yield during 180 minutes. The use of the MeOH:H2O mixture (1:1) led to an 80 % yield already after 20 minutes. Unfortunately, the authors did not perform the reaction in water alone, however, clearly, the reaction was improved by the addition of water. In another work, the COF-based material Pd@COF-QA was used for the first time in Suzuki– Miyaura cross-coupling in only water [50]. In the reaction between iodobenzene and phenylboronic acid, 99% yield of biphenyl was obtained after six hours. Under the same conditions much less reactive chlorobenzene was converted in 95%, indicating the high efficiency of this catalyst in water. 25.2.6.2  Hydrogenation Reactions

Jiang et al. developed an efficient heterogeneous catalyst system based on Pd nanoparticles supported on MIL-101-Cr and nanoparticles (Figure 25.18) [51]. It was used for phenol hydrogenation in water and excellent selectivity >99.9% to cyclohexanone was obtained in remarkably mild reaction conditions, including low pressure (0.1 MPa) and temperature as low as 25 °C. It was emphasized that cyclohexanone could easily be separated from water, which is attractive for eventual practical applications. An interesting example of the fundamental role of water in the transfer hydrogenation of phenylacetylene was reported by Trzeciak and Augustyniak (Figure 25.19) [52]. They used as [Pd(2-pymo)2]n as a catalyst, which is a unique MOFs containing Pd(II) nodes. Mechanistic studies indicated that water is a donor of one hydrogen atom, and the second hydrogen is generated from the hydrolysis of NH3BH3. The same catalyst is active in the dehydrogenation of NH3BH3. It is worth mentioning that the presence of hydrophobic cages Figure 25.18  The hydrogenation in [Pd(2-pymo)2]n limits the contact of NH3BH3 with Pd in a of phenol to cyclohexanone.

25.2  The Role of Water in Heterogeneous Catalytic Systems

Figure 25.19  Reactions of NH3BH3, hydrogen release and transfer hydrogenation, catalyzed by [Pd(2-pymo)2]n.

Figure 25.20  Activation of formate on UiO-66-Zr in hydrodehalogenation catalyzed by Pd@UiO-66.

water solution. This only resulted in a partial reduction of Pd(II) to Pd(0) and lowered the yield of H2 formed from the hydrolysis of NH3BH3. Recently, Olsbye et al. showed the positive influence of H2O on CO2 hydrogenation to methanol catalyzed by Pt@UiO-67 [53]. The obtained results indicated that water increased the selectivity to methanol because it promoted methanol desorption from the catalyst surface and inhibited the formation of CH4. Li et al. [54] evaluated the hydrodehalogenation of aryl and heteroaryl halides over Pd supported on the UiO-66 (Figure 25.20). In addition to ammonium formate being a hydrogen donor, the presence of water was required to trigger the reaction. The proposed catalytic cycle supposes that the first carboxyl from ammonium formate is activated on Pd active sites, and then the nucleophilic addition of water to the carbonyl groups with the H–O bond cleavage. As a result, two Pd–H active sites for hydrodehalogenation reaction are obtained. 25.2.6.3  Hydroamination

Cirujano et al. reported the synthesis of a defective nickel pyrazolate porous framework with cationic [Pd(NH3)4]2+ sites. This material was employed as a catalyst in hydroamination reactions (Figure 25.21) [55] Interestingly, in case there is no water in the catalytic system, the reaction goes through by the coordination of palladium sites to the triple bond of the substrate, and, finally,

549

550

25  The Role of Water in Reactions Catalyzed by Transition Metals

Figure 25.21  Mechanism of hydroamination and hydration over Pd@NiBDP catalyst.

indole was generated. However, water introduced into the system favored the hydration of the terminal alkyne, and 2-aminoacetophenone was obtained.

25.3  The Contribution of Water in Homogeneous Catalytic Systems 25.3.1  Oxidation and Epoxidation The contribution of water in the oxidation of alkanes was studied experimentally and theoretically for the [MeReO3]/H2O2/H2O–CH3CN system. According to this scheme, water is the H-transfer promoter and oxidation occurs with the contribution of HOO� and HO� radicals, formed in the water assisted process [56]. Thus, the Re(VII) complex, MeReO2(OO)(H2O), formed in the reaction of MeReO3 with H2O2, reacted with the second H2O2 molecule to give MeReO2(OO)(H2O)(H2O2) (Figure 25.22). Next, the reaction of this complex with H2O resulted in a cleavage of the Re-OOH bond and the formation of the HOO� radical and further reactions with H2O2 and H2O generate HO� radicals involved in the oxidation process. It was found, that the coordination of water to the Re catalyst decreased the activation energy of radical formation and, in consequence, the oxidation reaction can occur under mild conditions, at room temperature. The epoxidation of olefins by hydrogen peroxide catalyzed by the same complex, MeReO3, was accelerated in the presence of H2O, which acted as a co-catalyst. This improving effect was explained by the contribution of water in all steps of the catalytic process in agreement with results of DFT calculations [57]. In particular, the formation of cyclic transition states facilitates the proton transfer from H2O2 to the oxo or hydroxo ligand. As a result, peroxo complexes are formed and they react next with cyclohexene to form epoxide (Figure 25.23).

25.3.2  The Hydrogenation of Carbonyl Compounds and CO2 Ab initio molecular dynamic calculations performed for the model Ru-catalysts in the transfer hydrogenation of formaldehyde, show the difference between reaction mechanisms operating in an aqueous solution, in methanol or in the gas phase. The formation of hydrogen bonds by water

25.3  The Contribution of Water in Homogeneous Catalytic Systems

Figure 25.22  Formation of HOO� and HO� radicals in reaction with MeReO2(OO)(H2O) complex.

Figure 25.23  Contribution of water in olefin epoxidation with Re catalyst.

Figure 25.24  Ru-catalyzed transfer hydrogenation of formaldehyde in water.

and formaldehyde facilitated a proton transfer from a water molecule to substrate. A Ru-methoxide intermediate created at this stage was stabilized by three water molecules (Figure 25.24). In contrast, methanol interacted mainly with the catalyst [58]. Two energetic profiles were calculated for acetone hydrogenation catalyzed by [RhH2(PR3)2S2]+ in the presence of water [59]. The first case was based on a reductive elimination pathway and the second, on outer sphere hydrogenation. It was found that the energetic span for the first path is lower and, consequently, it can be considered the energetically preferred route. The presence of two water molecules significantly reduced the relative energy of the Rh-intermediate from 27.1 kcal mol–1 to 10.9 kcal mol–1 (Figure 25.25). The accelerating influence of water on the Ru-catalyzed hydrogenation of CO2 was studied using the DFT method. It was found that the formation of hydrogen bonds between water and CO2 facilitated the nucleophilic attack of the H ligand on CO2 by lowering the activation barrier [60]. The water modulated interaction of CO2 with the Ru-H species lead to the insertion product. It is worth noting that in the reaction carried out in the presence of water, the active species is

551

552

25  The Role of Water in Reactions Catalyzed by Transition Metals

cis-Ru(H)2(PMe3)3(H2O)2, whereas cis-Ru(H)2(PMe3)3 is catalytically active in the reaction performed without water. Interestingly, when the reaction was carried out in water, CO2 interacted with hydride and aqua ligands but not with the Ru center.

25.3.3  The Cyclotrimerization of Alkynes Water played an important role in the Pt(II) catalyzed cycloisomerization of enynes [61]. This reaction is mainly catalyzed by Pt and Au complexes and different mechanisms were proposed for these two types of catalysts. For Au, a linear coordination of alkyne to Au(I) is expected rather than the formation of metallacycle. In contrast, metallacycles are typical intermediates in the Pt catalyzed cycloisomerization of enynes. It was found that in this reaction water facilitated product formation and to explain this effect Echavarren et al. proposed that water prevented the chelating coordination of the 1,6-enyne to the Pt catalyst involving the coordination of 1,6-enyne via its double and triple bonds. This assumption was next supported by electrospray ionization-mass spectrometry (ESI-MS) studies and theoretical calculations. The Pt(II) anionic complex with the diphosphine ligand, used as a precatalyst, formed an adduct with the cyclic reaction product identified by ESI MS (m/z 394.6) using CAR/CID sequences (CAR = collision-activated reaction, CID = collision-induced dissociation). What was important, the amount of this adduct significantly increased (by factor c. 2.6), in the presence of H2O compared to CHCl3. In conclusion, the presence of the aqua ligand favored the coordination of 1,6-enyne to Pt via a triple bond and limited the formation of chelate and directed reaction selectivity toward cycloisomerization (Figure 25.26).

Figure 25.25  Contribution of water in ketone hydrogenation.

Figure 25.26  The cyclotrimerization of 1,6-enynes.

25.3  The Contribution of Water in Homogeneous Catalytic Systems

25.3.4  The Isomerization of Allylic Alcohols The isomerization of allylic alcohols presents a simple atom-economical procedure for the synthesis of aldehydes or ketones which can be performed in water. In the review article published by Romerosa et al., different aspects of this reaction were discussed including the presentation of catalysts and mechanistic approaches. Considering catalysts of isomerization, ruthenium is the most frequently used and several versions of the reaction mechanism for Ru complexes were proposed [62, 63]. One of these mechanisms, proposed for the neutral Ru-Cl complex, is based on the π-allylhydride scheme. In the first step, alcohol coordinates to the Ru catalyst via a double bond. Next, the β-proton from the π-allylic fragment is transferred to the metal, forming the Ru-H species. The rearrangement of protons in this intermediate produces enol transformed next to aldehyde. In this classic mechanism water was not involved (Figure 25.27). In 2011 Valera-Alvares et al. presented studies on the isomerization of allylic alcohols catalyzed by Ru complexes [{Ru(η3:η3-C10H16)Cl(μ-Cl)}2], [Ru(η3:η3-C10H16)Cl2(L)], (L=CO, PR3, CNR, NCR, and [Ru(η3:η3-C10H16)(NCMe)2]SbF6 (C10H16 = 2,7-dimethylocta-2,6-diene-1,8-diyl) in water and in THF solutions [64]. For both reactions, the Gibbs free energy profiles were constructed and it was found that the coordination of water to Ru lowered the energy barrier by c. 7 kcal mol–1. It was also characteristic that in all intermediates included in the catalytic process, Cl and H2O ligands remained coordinated to Ru (Figure 25.28). Another mechanism was proposed for the same reaction in a basic medium in which catalytic efficiency was significantly higher than in neutral conditions (Figure 25.29). In this case, water

Figure 25.27  The isomerization of allylic alcohols.

Figure 25.28  Ru intermediates formed during isomerization of allyl alcohol under neutral conditions.

Figure 25.29  Ru intermediates formed during isomerization of allyl alcohol under basic conditions.

553

554

25  The Role of Water in Reactions Catalyzed by Transition Metals

Figure 25.30  The isomerization of allyl alcohol in water with Ru-pyrazole catalyst.

contributed directly in the reaction as a donor of a proton transferred to an enolate ligand. Ru species involved in the catalytic cycle contained aqua or hydroxo ligands. Gimeno et al. employed a Ru(IV) catalyst bearing the pyrazole-type ligand, [Ru(η3:η3-C10H16) Cl2L] (C10H16 = 2,7-dimethylocta-2,6-diene-1,8-diyl, L = pyrazole ligand) in the isomerization of allylic alcohols in water [65]. They proposed the mechanism of this reaction with the contribution of water and pyrazole ligands, while the Cl ligand was not involved in an active intermediate (Figure 25.30). The studies performed for the most active complex show that water easily substituted the Cl ligand and formed Ru-OH species after the proton transfer to an N-atom of pyrazole. The substrate coordinated to Ru via an oxygen atom and formed the hydrogen bond with an OH ligand. Next, a bifunctional concerted mechanism occurred with a hydride transfer to Ru and a proton transfer to the OH ligand. The as formed Ru intermediate contained aqua and acrolein ligands. The transfer of two hydrogens from Ru-H and N-H groups to the C=C bond of acrolein was the final step of this reaction leading to the expected ketone. The hydrogen transfer step was investigated in detail considering two pathways with a different role of water. The authors concluded that the presented mechanism of isomerization with the active contribution of water can be competitive to the chloride route in the classic mechanism. The new mechanism also explains that reaction in water is faster than in tetrahydrofuran (THF). The effect of water was mentioned in the studies of the isomerization of allylic alcohols catalyzed by the water soluble [Rucp(H2O)(PTA)2](CF3SO3) complex (PTA = 1,3,5-triaza-7-phosphaadamantane) [66]. An increase of the reaction efficiency, observed in this reaction, was explained using the results of neutron scattering and theoretical calculations. The most interesting finding was the formation of a three-membered water chain connecting hydrogen atoms of the OH group of allylic alcohol with the N-atom of PTA. A similar water chain was also found for the second isomer of an Ru- intermediate, [Rucp(η2-CH2=CH-CH2OH)(PTA)2]+. The interaction with the water chain resulted in the increased stabilization of the isomer involved in the further reaction. The improving effect of water was additionally evidenced by the fact that catalytic results were much worse in MeOH than in water. This is because in MeOH hydrogen bonds do not stabilize the most favorable conformation of substrate, which can rotate around the C–C bond. In summary, the specific solvation of catalytic intermediates was responsible for the high efficiency of allyl alcohol isomerization in water.

25.3.5  Hydroarylation with Boron Compounds Very promising synthetic pathways can be realized due to the reactivity of diboron compounds with water. For example, diboron compounds mediate a palladium-catalyzed transfer of hydride from the water molecule to unsaturated substrates, such as alkenes and alkynes. In these reactions, the O–H bond in water is split on the highly oxophilic boron center and acts as a hydrogen donor [67].

25.3  The Contribution of Water in Homogeneous Catalytic Systems

Figure 25.31  Plausible mechanism of hydroarylation catalyzed by Pd(OAc)2(PCy3)2.

Figure 25.32  The hydroarylation of alkynes with D2O.

The hydroarylation of alkynes presents another reaction which can be successfully performed with water as the hydride source. In this reaction, an internal alkyne is transformed to the olefin with the addition of aryl and hydride to the triple bond. Prabhu et. al., reported an example of such a process catalyzed by Pd(OAc)2/Pcy3 with the diboron compound B2pin2 [B2pin2 = bis(pinacolato) diboron] used for water activation and PhB(OH)2 served as a donor of the phenyl group [68]. The reaction mechanism involved in the formation of the hydrido species of the H-Pd-OAc type in the reaction of Pd(OAc)2(PCy3)2 with B2pin2 and H2O (Figure 25.31). It is worth noting that Pd(OAc)2(PCy3)2(Bpin) reacted with H2O forming the catalytically active species, HPd(OAc) (PCy3)2. Water also contributed in another step of the catalytic process, namely the transmetalation of the Pd precursor with B2pin2 leading to pinB-[Pd]-OAc and AcO-Bpin. The as formed AcO-Bpin reacted with H2O forming HOAc, which is employed in the end of the catalytic cycle to recover catalytically active H-[Pd]-OAc from [Pd(0)(PCy3)2]. The function of water as a hydrogen donor was also evidenced in the hydroarylation of alkynes performed in water with the arylating agent Na[BPh4], catalyzed by PdCl2(im)2 (im = substituted imidazole) catalysts with different imidazole ligands [69]. The test experiment, performed with D2O instead of H2O showed the presence of D exclusively in the vinyl position of arylated olefin despite the presence of non-deuterated HOAc in this reaction (Figure 25.32).

25.3.6  Hydroformylation An unexpected increase of hydroformylation selectivity was achieved using the Rh(acac)(CO)2 and π-acceptor P(NC4H4)3 ligand in the presence of water. The hydroformylation of 1-butene carried out in a toluene/water mixed solvent system provided n/iso as high as 46 (at 8 bar H2/CO) or 51 (at 6 bar H2/CO). For comparison, in the absence of water the highest n/iso value was 22. Similarly, in the hydroformylation of propene with the P(NC4H4)3 ligand n/iso was 27.1 while without water it was 18.2. It was supposed that water interacted with the Rh-H catalytic intermediate and influenced a migratory insertion step to form a linear alkyl complex and next a linear aldehyde (Figure 25.33).

555

556

25  The Role of Water in Reactions Catalyzed by Transition Metals

Figure 25.33  The proposed interactions of water with Rh-H and Rh-alkyl intermediates during selective hydroformylation.

Not only selectivity but also the reaction rate increased in the presence of water [70]. This can be demonstrated by a comparison of TON values increasing from 127 to 255 after the addition of water to toluene in the hydroformylation of propene with Rh(acac)(CO)2/P(NC4H4)3. An increase of TON values was also observed for PPh3, PCy3 and PPh(NC4H4)2 in the same conditions.

25.4  Conclusions In conclusion, it is clear that water used as the solvent in the catalytic reactions is not an innocent component of these systems. On the contrary, by forming hydrogen bonds with the catalyst and reactants, water can significantly affect the activation stage and increase the rate of the reaction. Theoretical models developed for the interaction between water and catalyst, developed for heterogeneous systems, explain well the phenomenon of acceleration of the reaction. In homogeneous systems, water often becomes a ligand coordinated to the metal center. It can undergo decomposition (e.g. with the formation of radicals that participate in the subsequent steps of the reaction). The course of these processes depends on the type of metal and the coordinated ligands and it would therefore be more difficult to elaborate a uniform description of these reactions. There is no doubt, however, that it is possible to model the efficiency and selectivity of catalytic reactions in a very wide range by conducting them in water.

References 1 Anastas, P.T. (ed.) (2010). Reactions in water. In: Handbook of Green Chemistry, vol. 5 (ed. C.-J. Li), Wiley VCH Verlag GmbH & Co. KGaA. 2 Gawande, M.B., Bonifacio, V.D.B., Luque, R. et al. (2013). Chem. Soc. Rev. 42: 5522–5551. 3 Butler, R.N. and Coyne, A.G. (2010). Chem. Rev. 110: 6302–6337. 4 Rideout, D.C. and Breslow, R. (1980). J. Am. Chem. Soc. 102: 7816–7817. 5 Breslow, R. (1991). Acc. Chem. Res. 24: 159. 6 Gajewski, J.J. (1997). Acc. Chem. Res. 30: 219–225. 7 Narayan, S., Muldoon, J., Finn, M.G. et al. (2005). Angew. Chem. Int. Ed. 44: 3275–3279. 8 Chandra, A. and Fokin, V.V. (2009). Chem. Rev. 109: 725–748. 9 Jung, Y. and Marcus, R.A. (2007). J. Am. Chem. Soc. 129: 5492–5502. 10 Karhan, K., Khaliullin, R.Z., and Kuhne, T.D. (2014). J. Chem. Phys. 141: 22D528. 11 Kitanosono, T. and Kobayashi, S. (2020). Chem. Eur. J. 26: 9408–9429. 12 Chang, C.-R., Huang, Z.-Q., and Li, J. (2016). WIREs Comput. Mol. Sci. 6: 679–693.

References

13 Davies, P.R. (2016). Top. Catal. 59: 671–677. 14 Roberts, M.W. (2014). Catal. Lett. 144: 767–776. 15 Micek-Ilnicka, A. (2009). J. Mol. Catal. A: Chem. 308: 1–14. 16 Li, G., Wang, B., and Resasco, D.E. (2020). ACS Catal. 10: 1294–1309. 17 Meng, Q., Cao, D., Zhao, G. et al. (2017). Appl.Catal.B: Env. 212: 15–22. 18 Ma, R., Wu, X.P., Tong, T. et al. (2017). ACS Catal. 7 (1): 333–337. 19 Li, X., Wang, X., Roy, K. et al. (2020). ACS Catal. 10 (10): 5783–5792. 20 Hensley, A.J., Wang, Y., Mei, D., and McEwen, J.S. (2018). ACS Catal. 8 (3): 2200–2208. 21 Andrushkevich, T.V. and Ovchinnikova, E.V. (2020). Mol. Catal. 484: 110734. 22 Yang, X., Wang, X., and Qiu, J. (2010). Appl. Catal. A: General 382: 131–137. 23 Mullen, G.M., Gong, J., Yan, T. et al. (2013). Top. Catal. 56 (15–17): 1499–1511. 24 Bernasconi, L., Kazaryan, A., Belanzoni, P., and Baerends, E.J. (2017). ACS Catal. 7 (6): 4018–4025. 25 Bray, J.T., Ford, M.J., Karadakov, P.B. et al. (2019). React. Chem.Eng. 4 (1): 122–130. 26 Pentsak, E.O. and Ananikov, V.P. (2019). Eur.J. Org. Chem. 26 (2019): 4239–4247. 27 Cummings, S.P., Le, T.N., Fernandez, G.E. et al. (2016). J. Am. Chem. Soc. 138 (19): 6107–6110. 28 Kamal, A., Srinivasulu, V., Seshadri, B.N. et al. (2012). Green Chem. 14 (9): 2513–2522. 29 Dai, Y., Gao, X., Chu, X. et al. (2018). J. Catal. 364: 192–203. 30 Anderson, J.A., McKenna, F.M., Linares-Solano, A., and Wells, R.P. (2007). Catal. Lett. 119 (1): 16–20. 31 Bereta, T., Mieczyńska, E., Ronka, S. et al. (2021). New J. Chem. 45 (11): 5023–5028. 32 Alsalahi, W. and Trzeciak, A.M. (2018). Chem. Select 3: 1727–1736. 33 Pascanu, V., Miera Gonzalez, G., Inge Ken, A., and Matute-Martin, B. (2019). J. Am. Chem. Soc. 141: 7223–7234. 34 Geng, K., He, T., Liu, R. et al. (2020). Chem. Rev. 120: 8814–8933. 35 Freund, R., Zaremba, O., Arnauts, G. et al. (2021). Angew. Chem. Int. Ed. 60: 23975–24001. 36 Wang, C., Liu, X., Demir, N.K. et al. (2016). Chem. Soc. Rev. 45: 5107–5134. 37 Yuan, B., Pan, Y., Li, Y. et al. (2010). Angew. Chem. Int. Ed. 49: 4054–4058. 38 Carson, F., Pascanu, V., Bermejo Gómez, A. et al. (2015). Chem. Eur. J. 21: 10896–10902. 39 Kardanpour, R., Tangestaninejad, S., Mirkhani, V. et al. (2014). J. Organomet. Chem. 761: 127–133. 40 Singha Roya, A., Mondal, J., Banerjee, B. et al. (2014). Appl. Catal. A: Gen. 469: 320–327. 41 Niknam, E., Panahi, F., and Khalafi-Nezhad, A. (2020). Appl. Organomet. Chem. 34: 5470. 42 Augustyniak, A.W., Zawartka, W., Navarro, J.A.R., and Trzeciak, A.M. (2016). Dalton Trans. 45: 13525–13531. 43 Huang, J., Wang, W., and Li, H. (2013). ACS Catal. 3: 1526–1536. 44 Alamgholiloo, H., Rostamnia, S., Hassankhani, A. et al. (2018). Appl. Organomet. Chem. 32: 4539. 45 Shang, N., Gao, S., Zhou, X. et al. (2014). RSC Adv. 4: 54487–54493. 46 Chen, L., Rangan, S., Li, J. et al. (2014). Green Chem. 16: 3978–3985. 47 Yuan, N., Pascanu, V., Huang, Z. et al. (2018). J. Am. Chem. Soc. 140 (26): 8206–8217. 48 Lima, C.F.R.A.C., Rodrigues, A.S.M.C., Silva, V.L.M. et al. (2014). ChemCatChem 6: 1291–1302. 49 Goncalves, R.S.B., de Oliveira, A.B.V., Sindra, H.C. et al. (2016). ChemCatChem 8: 743–750. 50 Wang, J.-C., Liu, C.-X., Kan, X. et al. (2020). Green Chem. 22: 1150–1155. 51 Liu, H.L., Li, Y.W., Luque, R., and Jiang, H.F.A. (2011). Adv. Synth. Catal. 353: 3107–3113. 52 Augustyniak, A.W. and Trzeciak, A.M. (2022). Inorganica Chim. Acta 538: 120977/1–120977/6. 53 Gutterød, E.S., Øien-Ødegaard, S., Bossers, K. et al. (2017). Ind. Eng. Chem. Res. 56 (45): 13206–13218. 54 Tonga, L., Songa, X., Jiang, Y. et al. (2022). Int. J. Hydrog. Energy. 47: 15753–15763.

557

558

25  The Role of Water in Reactions Catalyzed by Transition Metals

55 Cirujano, F.G., López-Maya, E., Navarro, J.A.R., and De Vos, D.E. (2018). Top. Catal. 61: 1414–1423. 56 Kuznetsov, M.L. and Pombeiro, A.J. (2009). Inorg. Chem. 48 (1): 307–318. 57 Goldsmith, B.R., Hwang, T., Seritan, S. et al. (2015). J. Am. Chem. Soc. 137 (30): 9604–9616. 58 Pavlova, A. and Meijer, E.J. (2012). ChemPhysChem. 13 (15): 3492–3496. 59 Polo, V., Schrock, R.R., and Oro, L.A. (2016). Chem. Comm. 52 (96): 13881–13884. 60 Ohnishi, Y.Y., Nakao, Y., Sato, H., and Sakaki, S. (2006). Organomet. 25 (14): 3352–3363. 61 Baumgarten, S., Lesage, D., Gandon, V. et al. (2009). ChemCatChem. 1 (1): 138–143. 62 Lorenzo-Luis, P., Romerosa, A., and Serrano-Ruiz, M. (2012). ACS Catal. 2 (6): 1079–1086. 63 Scalambra, F., Lorenzo-Luis, P., de los Rios, I., and Romerosa, A. (2019). Coord. Chem. Rev. 393: 118–148. 64 Varela‐Álvarez, A., Sordo, J.A., Piedra, E. et al. (2011). Chem. Eur. J. 17 (38): 10583–10599. 65 Bellarosa, L., Díez, J., Gimeno, J. et al. (2012). Highly efficient redox isomerisation of allylic alcohols catalysed by pyrazole‐based ruthenium (IV) complexes in water: mechanisms of bifunctional catalysis in water. Chem. Eur. J. 18 (25): 7749–7765. 66 Scalambra, F., Holzmann, N., Bernasconi, L. et al. (2018). Water participation in catalysis: an atomistic approach to solvent effects in the catalytic isomerization of allylic alcohols. ACS Catal. 8 (5): 3812–3819. 67 Neeve, E.G. and Geier, S. (2016). Diboron(4) compounds: from structural curiosity to synthetic workhorse. Chem. Rev. 116: 9091–9161. 68 Rao, S., Joy, M.N., and Prabhu, K.R. (2018). Employing water as the hydride source in synthesis: a case study of diboron mediated alkyne hydroarylation. J. Org. Chem. 83 (22): 13707–13715. 69 Kocięcka, P. and Trzeciak, A.M. (2020). Efficient hydroarylation of terminal alkynes with sodium tetraphenylborate performed in water under mild conditions. Appl. Catal. A: General 589: 117243. 70 Alsalahi, W., Grzybek, R., and Trzeciak, A.M. (2017). N-Pyrrolylphosphines as ligands for highly regioselective rhodium-catalyzed 1-butene hydroformylation: effect of water on the reaction selectivity. Catal. Sci. Technol. 7: 3097–3103.

559

26 Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts Skyler Markham1, Debbie C. Crans1,2, and Bruce Atwater2 1 2

Dept. Chemistry, Colorado State University, Fort Collins, Colorado Fort Hays State University, 600 Park Street, Hays, Kansas

26.1  Introduction One of the most important reactions in organic chemistry to date is the cross-coupling reaction. Its importance can be demonstrated in the many products that are prepared using these reactions (including discodermolide [1], polmacoxib (Acelax) [2], and sonidegib [2]) and in the fact that it won the Nobel Prize in 2010 [3]. Of particular utility to industry are the Suzuki-Miyaura, Sonogoshira, and Buchwald-Hartwig cross couplings [4]. With these three reactions occurring in 40% of the industrial publications in 2014, it is apparent industry heavily relies on these crosscoupling reactions [4]. There have been over 45,000 papers compiled in Web of Science concerning Suzuki-Miyaura cross-coupling reactions that have primarily focused on ligands, metal catalyst, reaction type, and organoboron species. The importance of the palladium (Pd) based catalysts is one of the common factors across these reactions, as illustrated in the Suzuki-Miyaura cross-­ coupling cycle (Figure 26.1) [5]. However, the toxicity of these catalysts and the limited availability of these catalysts is often ignored [6]. In this review, we examine present literature focused on more sustainable catalysts along with exploring the utility of speciation for future catalyst design for use in Suzuki-Miyaura cross-coupling reactions.

26.2  The First Cross-coupling Reaction The first coupling between an alkyl halide and organoboron reagent was reported by Suzuki and Miyaura in 1981, Figure 26.2 [7]. While using metal catalysts to form carbon-carbon bonds was well-known, none had demonstrated a reaction suitable for frequent use in industry [8]. The use of relatively non-toxic boron compounds complemented by their simple and efficient synthesis allowed for extensive use in both industrial and academic applications [8].

Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 2, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

560

26  Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts

Ph

[Pd0L2] Reductive Elimination

L

Oxidative Addition L

L

L Pd II Ph

Ph

X

Ph

Pd II

Pd II X L

L

RO

Trans Metallation L B(OH)2OR

Ph

Pd II OR L

B(OH)2

Figure 26.1  General Suzuki-Miyaura Cross-Coupling Catalytic Cycle [5] / American Chemical Society.

B(OH)2 +

Br R

Pd(PPh3)4 aq. Na2CO3

R

Figure 26.2  First Suzuki-type cross-coupling reported by Suzuki and Miyaura [7] / Taylor & Francis.

26.3  Phosphine Ligands for Catalysts of Cross-Coupling Reactions Since the initial report by Suzuki and Miyaura, there have been major advances to the reaction. One major area is in improving the ligands surrounding the catalyst leading to two major ligand classes: phosphines and N-heterocyclic carbenes (NHC), Figure 26.3. Phosphine based ligands have a long and rich history for Pd-catalyzed cross-coupling reactions [7, 9]. Both aromatic and aliphatic phosphine ligands have been used in cross-coupling reactions which demonstrates the ligands require both electronic and steric properties of such substituents to be successful, Figure 26.4 [10–15]. The success of these ligands comes as no surprise, as they are electron rich which increases the metal center’s reactivity for oxidative addition [16]. The ligands also readily dissociate and reassociate with Pd(II) which can allow a coordination site to open or close to facilitate oxidative insertion or reductive elimination at the metal center [17–19]. Phosphine ligands have been diversified over time to enhance reactivity, including developing both sterically hindered aliphatic groups as well as heteroatom based groups. The first study demonstrating a wide substrate scope using [1,1ʹ-bis(diphenylphosphino)ferrocene]dichloro-palladium (II) (Pd(dppf)Cl2) was in 1988 by Thompson et al. [20]. Some of the more recent developments into phosphine ligands belong to the Phos class (Figure 26.4) which have been further modified by multiple groups including Buchwald et al. and Chan et al. [21, 22]. This class has been rapidly growing to include several ligands derivatives, all of which have demonstrated their utility across multiple cross-coupling reactions, Figure 26.4. The Phos ligand design is based on the Phos scaffold (Figure 26.4). The R groups bound to the phosphorus will increase electron density at the phosphorus, allowing an increased rate of oxidative addition. Reductive elimination will also be increased with large R groups (R = t-Bu and

26.3  Phosphine Ligands for Catalysts of Cross-Coupling Reactions Ph P P

R1 Ph

Fe P Ph

X

PR2 R2

R4

R

Ph

N

X N

R

R3

Figure 26.3  Examples of two major classes of ligands, phosphines and carbenes. From left to right: three examples of phosphine class, triphenylphosphine, ferrocenylphosphine ligand (dppf), and Phos ligand; and from the carbene class, N-heterocyclic carbene.

Figure 26.4  Various phoshine (Phos) ligands commonly used in cross-coupling reactions beginning with the general Phos scaffold (where R can be alkyl, aryl, H, or a heteroatom, R3 ≠ H typically for ease of synthesis). Then we follow by showing ligands in a historical order: JohnPhos [10], DavePhos [11], Xphos [12], Sphos [13], t-BuXPhos [14] and BretPhos [15].

Cy, shown in Figure 26.4). The steric strain caused by R1 next to the phosphorus will force a conformational change that pushed PR2 over the lower aryl ring and improves the rate of reductive elimination. The lower aryl ring increases the ligands size, slowing down the oxidation of the ligand by O2. This ring also allows stabilization of the Pd complex through Pd-arene interactions. In addition, R2 and R4 increase stability by preventing cyclometallation [23]. Overall, the main goal of catalyst development was working toward a universal catalyst system. For a universal catalyst to be achieved, several criteria must be met: i) possess a broad substrate scope; ii) operate at low catalyst loadings; and iii) operate at or near ambient temperature [13]. These goals have led Buchwald and others to further refine their catalysts to develop new and more advanced catalysts [24, 25]. Although a true universal catalyst has not yet been reported, there have been several advances in catalyst design that have given rise to multiple catalysts which meet most of the goals outlined previously [10–15]. The large number of ligands that have been synthesized has allowed for a systematic understanding of the effects of substituents on the ligand and coordination of the ligand class to the Pd in the catalyst [15]. This high degree of understanding of the ligand has allowed Buchwald and others to further their journey in search of a truly universal ligand for the (Pd)II catalyst for cross-coupling reactions. Through the work of Buchwald et al.to find a universal catalyst, the XPhos ligand was developed (Figure 26.4) [12]. The XPhos catalyst system allowed for heteroaromatic couplings in high yields (80% or larger) at mild temperatures (up to 40 °C) and short reaction times (30 minutes to two hours) for several different ligands. This is a step in the direction of developing the new general catalyst design for heteroaromatic coupling reactions and the variety of ligands and their conditions and yields summarized in Figure 26.5 [26]. A major advantage of the phosphine ligands over other ligand classes is their ability to be readily interchanged with other phosphine based ligands. The interchangeable nature of the ligands on

561

562

26  Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts

Figure 26.5  Regioselectivity in product distribution based on phosphine (Phos) structures [26] / American Chemical Society. Two catalytic systems are shown (A and B).

the Pd(II) allows a researcher to use similar reaction conditions thereby allowing catalyst flexibility in similar applications, Figure 26.5. The ability to vary the ligand while keeping the conditions identical makes these catalyst systems highly attractive for high-throughput screenings in industrial and academic settings [26]. These variable features of the phosphine-based ligand systems make them prominent ligands in effective catalysts in cross-coupling reactions. The second major class of ligand family for Pd-catalyzed Suzuki-Miyaura cross-coupling reactions is the NHC) ligand, shown in Figure 26.6 [27, 28]. This ligand family generates two general subclasses of Pd coordination complexes: the Organ PEPPSI class of complexes (left) [28, 29] and Nolan’s class of complexes (right) [29], Figure 26.6. NHCs have a history as effective ligands for organometallic transformations [30–32] and organocatalytic transformations [33], as well as for cross-coupling reactions [29, 34]. These two classes of complexes differ by the ternary ligand system attached to the Pd center. For the PEPPSI ligands, the Pd(II) center typically has the NHC coordinated anti to a pyridine ligand and two halide ligands to complete the coordination sphere [28]. For the Nolan family of catalysts, the NHC ligand is accompanied by a halide and an η-three bound allyl anion species trans to the NHC group [35]. Both of these catalyst families have demonstrated

26.3  Phosphine Ligands for Catalysts of Cross-Coupling Reactions

excellent performance in various cross-coupling reactions [36–38]. Of particular note is their ability to catalyze Suzuki-Miyaura reactions for a variety of boron co-catalyst species [28, 39–43]. The various positions in the coordination complex of these catalysts are highly tunable and alters the reactivity of the catalysts. The groups on the backbone of the NHC ring typically force the aromatic groups on the nitrogens closer to the Pd(II) center [28]. This impacts the catalysts in two major ways. First, it accelerates reductive elimination of the transmetallated species through the steric crowding of the metal center. The second major advantage is that it forces the alkyl groups on the aromatic rings into a position over the Pd(II) center [44, 45]. This positioning of the alkyl groups over the Pd(II) center thereby further accelerates reductive elimination due to steric crowding of the metal center. It also has the added benefit of limiting the formation of agostic interactions with β-hydrogens on the transmetallated species which limits β-hydride elimination allowing for some of the catalysts to be using in secondary C(sp3)-C(sp2) couplings, Figure 26.7 [44]. The latter reaction has not so far been observed with Suzuki-Miyaura reactions [46–48]. One major advantage of NHC catalysts relative to the phosphorus-based ligands is the high degree of stability demonstrated by the pre-catalyst NHC species. Greater catalyst stability allows for convenient storage of the catalysts under ambient benchtop conditions [49]. The PEPPSI R

N R

Cl

R

N Pd Cl

R

N

R R

N

R

N Pd

R

Cl Cl

Figure 26.6  Two classes of N-Heterocyclic carbene (NHC) ligands (NHC) catalysts: pyridine enhanced precatalyst preparation stabilization and initiation (PEPPSI) catalyst on the left, and Nolan Catalyst on the right [27, 28].

Figure 26.7  Agostic interactions due to pyridine enhanced precatalyst preparation stabilization and initiation (PEPPSI) catalysts [44] / John Wiley & Sons.

563

564

26  Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts

Ar

X

+

Ar’

B(OH)2

Conditions

Ar’ N

O N (93%, X = Cl)

Ar

S

S

N

S (88%, X = Br)

Figure 26.8  Compounds synthesized using pyridine enhanced precatalyst preparation stabilization and initiation (PEPPSI) catalysts [36] / Thieme Medical Publishing Group.

(98%, X = Br) OMe

F3C

OMe OMe

(85%, X = Cl)

(93%, X = Br)

catalyst robustness has been routinely demonstrated. For example, they have been used on a solid support, and they have been recycled in flow reactions as well [50]. The PEPPSI catalyst has been used to generate a wide array of products, Figure 26.8 [36]. Unlike the phosphorus family of ligands, the NHC ligands are more on track toward the development of a true universal Pd(II) catalyst that can be utilized in all cross-coupling reactions. The main drawback for this class of coordination complex is that if the catalyst fails, a new catalyst must be prepared rather than merely altering one ligand on the coordination complex as done with the phosphorus ligands. However, this is remedied by the wide substrate scope demonstrated by the catalysts generated from the NHC ligands along with the large number of different catalysts that have been developed based on these ligands [40, 44, 45].

26.4  Speciation Many Pd catalysts, particularly those formed from Phos ligands, are rather stable and readily support loss of ligands which allows the complexes to engage as catalysts under relevant reaction conditions. However, many Pd catalysts are coordination complexes and thus defined as a metal ion surrounded by ligands which are formed in equilibrium reactions. This is the case even if the ligand is carbonbased which will ultimately lead to C-C bond formation in the cross-coupling reaction. At this time it appears that the properties of the palladium catalysts used in the cross-couplings may not be as well understood as initially presumed [51, 52]. For example, it has been suggested that the active catalyst may not be a homogenous Pd(II) species which carries out the reaction but rather a palladium nanoparticle [51, 52]. Mechanistic work continues to be reported which adds to the discussion of the reaction mechanism [51, 52]. Hence as the development of the system moves toward green and sustainable catalysts, more consideration of the structure and electronic properties of the catalyst and associated speciation chemistry would be required examples of which are described below.

26.5  Palladium Nanoparticle Catalysts and Speciation Increasing interest in the nature of the specific Pd species that carries out the cross-coupling reactions suggest several alternative options for the central metal in the catalyst exist. This is of particular importance to industry as homogenous catalysts are often expensive and unrecoverable

26.5  Palladium Nanoparticle Catalysts and Speciation

Figure 26.9  Schematic representation of different types of nanoparticles (NPs) divided into organic, hybrid, and inorganic NPs. Ref [54] / MDPI / CC BY 4.0.

[53]. Nanoparticles provide an alternative solution to these problems. For example, the Pd could be supported by any of the nanoparticle types shown in Figure 26.9 [54]. Nanoparticles can be readily be recovered through physical separations [55]. However, there many different nanoparticles (NPs) which vary in shape, size, structure, and composition and which have different catalytic properties. In Figure 26.9 [54], We show organic, hybrid, and inorganic NPs in circular, spherical, and elongated shapes. Shape, structure, size, and molecular composition will affect the properties of the materials as supports for the catalysts and varies types of NP-coatings will further provide flexibility in use of the NPs. Examples of different reactions utilizing such nanoparticles are discussed later and summarized in Figure 26.10. One of the first nanoparticle catalysts employed for this process was palladium on carbon (Pd/C), Figure 26.10 [56]. Initial work with this catalyst system demonstrated that it can be readily employed with good to excellent yields, and it is the most common catalyst system used under Suzuki-Miyaura coupling conditions [56]. Subsequently, the Pd/C catalyst was utilized in a onepot process to form biaryl aniline species from aryl boronic acids and halogenated nitroaromatics Use of Pd/C B(OH)2

X R1

+ X = Br, I

2M K2CO3/DME 80 oC

R2 Y

Br +

Pd/C, PPh3

B(OH)2

R2

R1

one pot: 1) Pd/C (10%) Na2CO3 reflux

NH

2) H2 (1atm) R2CHO, rt

O2N

R2

Use of nanospheres X R1

+ R2 X = Cl, Br, I

B(OH)2

Pd NPs@carbon nanospheres (0.2 mol%) R1

Figure 26.10  Various reactions with different nanosupports [55–57].

R2

565

566

26  Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts

in good to excellent yields [57]. This two-step process involved utilizing the supported catalysts to carry-out a cross-coupling between the boronic acid and the halogenated aromatic substrate. The nitro group was then reduced to an aniline under standard conditions followed by a subsequent reductive amination to give a secondary amine. This process demonstrates the utility and power of the support nanoparticle catalysts for carrying out multiple sequential reactions in one pot with the same catalyst. In addition, the catalysts for these reactions were recovered by a simple filtration [57]. This thereby allows for the catalysts to be recycled for another reaction sequence; however, this was only suggested and not demonstrated by the authors this in their papers [56, 57]. Carbon based supports have also been investigated but have been met with mixed levels of success, Figure 26.10 [55]. With Pd/C supported catalysts, leaching of a Pd species to the bulk solution is possible where observed cross-coupling activity can be attributed to the homogeneous Pd species [58]. Leaching was also reported from other supported catalyst systems [55], and these observations raise the possibility that nanomaterial based catalysts are exerting cross-coupling activity through a homogenous catalyst and/or from a catalyst located at the surface of the nanoparticle. A recent study by Zeng et al. deposited Pd catalysts onto a preformed iron oxide nanoparticle and then coated the nanoparticles with the microporous Stöber silica coating of varying pore sizes [51]. Zeng utilized filters with differing pore sizes to separate any soluble catalyst from the nanoparticle catalysts to determine whether the cross-coupling reaction was catalyzed by soluble or surface active catalysts. In addition, they varied the size of the halide and the boron species to determine whether the leached material or the Pd from the surface was the most effective cross-coupling catalyst [51]. They demonstrated that their catalyst had good reactivity with small boronic acids and halides, but the system showed no activity when tested with larger halides and a larger boronic acid than those commonly used. However, when a small aryl halide was employed with a large boronic acid they were able to achieve the same level of conversion demonstrated with the small halide and small boronic acid [51]. Their results suggest the cross-­ coupling reactions must be happening in bulk solution due to leached Pd catalyst. Therefore, any product that is produced must originate from cross-coupling reactions in the bulk solution rather than on the nanoparticles themselves [51]. This data lends credence to the notion that these reactions occur with homogenous catalysts and greatly contributed to the debate whether soluble or surface associated Pd species are the most active cross-coupling catalysts [52]. Numerous reports regarding the advantages of heterogeneous vs. homogeneous catalysts have been made particularly in industrial settings [51, 52, 58]. Extensive literature is available regarding the complex physical and chemical phenomena involved in such processes [51, 52]. The investigations into nanoparticle catalysts have revealed that such approaches have considerable potential [55]. However, Pd coated nanoparticle catalysts suffers from the same problems as coordination complex catalysts, such that in solution the catalyst engages in speciation equilibria, and more than one species is present in solution. In the following section we will address speciation of the catalysts and its impact on cross-coupling reactions.

26.6  Speciation of Palladium (Pd) Catalysts Given the great variety of reaction parameters that are critical to the success of the cross-coupling reactions, we here will describe a few selected cases when formation of different Pd species affect the stereo- and regiochemistry as well as yield of product formation. The speciation of Pd(OAc)2 is particularly relevant because Pd(OAc)2 is often used as a Pd catalyst precursor by forming an active Pd catalyst under some conditions of cross-coupling reactions. Pd(OAc)2 is reduced in situ to form catalytically active Pd(0) species that can catalyze

26.6  Speciation of Palladium (Pd) Catalysts

cross-­coupling reactions at ambient temperatures [58, 59]. Pd(OAc)2 is also used in ligand free (“ligandless”) Suzuki-Miyaura reactions [58] however, its structure is important to the function. Of the Pd-material Trimeric [60] and polymeric [61] structures of Pd(OAc)2 have been reported in the solid state; however, the solution structure of Pd(OAc)2 is sensitive to its solvent environment. A trimer has been reported in benzene [62], acetic acid [63], chloroform [64] and methanol [64]. A linear dimer has been reported in chloroform-acetic acid [65], whereas a monomer has been reported in N-methyl-2-pyrrolidinone (NMP) [66]. Studies have now shown that the Pd atom on the trimer is rapidly subjected to a nucleophilic attack by H2O, resulting in the loss of acetic acid and followed by a slower conversion to the monomeric species. The dissociation of the trimeric structure to the monomeric structure increases as the dipole moment of the solvent increases [58]. For example, in a solution of 40 mM Pd(OAc)2 in toluene, 71% was trimer, 21 % was monomer and 2 % particles, whereas in DMF, there was 56 % of the trimer, 42% monomer and 8 % particles [58]. Pd coordination complexes, including those used as catalysts, are dependent on three factors: (i) the oxidation state of the Pd, (ii) the ligand and the ratio of Pd to ligand, and (iii) other reaction conditions such as the presence of O2, water, solvent, temperature, and reaction time. Phosphine ligands are strong ligands and often remain coordinated to the complex even at low Pd: ligand ratios. Ligands such as acetate (OAc−) more readily dissociate as shown in Eq. 26.1 and 26.2.

Pd(OAc)2 + 3 PPh3  trans- Pd(OAc)2 (PPh3 )2  + PPh3  

(26.1)



Pd(OAc)2 + PPh3  [Pd(µ 2 -OAc) (κ-OAc )(PPh3 )]2

(26.2)

Specifically in the case of solutions of Pd(OAc)2 and triphenylphosphine (PPh3), the formation of the species depend on the Pd(OAc)2:PPh3 ratio when other reaction conditions remain the same [52]. When PPh3 is in threefold or more excess of Pd(OAc)2, the major species in solution is as shown in Eq (1) is the trans-[Pd(OAc)2(PPh3)2]. When the Pd(OAc)2 is about the same concentration as PPh3, the major species in solution is the dimer [Pd(μ2-OAc) (k-OAc)(PPh3)]2, as shown in Eq (2). This difference as been explored and found to result in a change in cross-coupling site selectivity in dihalogenated heteroarenes as shown in Figure 26.11 [52].

Figure 26.11  Different Pd species arising from different ratios of Pd(OAc)2/nPPh3 (PPh3 = triphenylphosphine) result in different cross-coupling selectivities under cross-coupling conditions. Adapted from Ref [52].

567

568

26  Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts

Approaches toward the preparation of sp2- and sp3-hybridized carbon frameworks using Suzuki– Miyaura cross-coupling reactions include a strategy for the chemoselective synthesis and conversion of boronic esters. This approach provides direct access to reactive sp2-hybridized ­pinacol protected boronic acid (BPin) products by controlling boron speciation during the cross-­ coupling reaction, as illustrated in Figure 26.12 [67]. Suzuki–Miyaura cross-coupling of an aryl BPin with a conjunctive haloaryl methyl-iminodiacetic protected boronic acid (BMIDA) generates an intermediate biaryl BMIDA [67, 68]. The byproduct from the Suzuki–Miyaura coupling reaction is HOBPin. Through appropriate control of the basicity in the biphasic reaction system these species can be hydrolyzed, and the resulting biaryl boronic acid and pinacol can be driven toward formation of the aryl BPin derivative. Accordingly, under this protocol a protected boronic acid affords reactive boronic ester products with no need for deprotection reactions. Solution speciation of boronic acids can be chemoselectively controlled to enable the formal homologation of boronic acid pinacol esters [69]. The reaction is tolerant of aryl and vinyl functionality as both the pinacol donor and acceptor, respectively, and enables streamlining of the iterative cross-coupling reaction as well as a method for controlled oligomerization. Another reported approach involves manipulating speciation so that a different Pd intermediate within the catalytic cycle becomes the resting state through starvation of a reagent, Figure 26.13

Bpin

Ar1

Ar2

Ar2

BMIDA +

HO

Bpin

Ar1

BMIDA

X

HO Bpin Ar1

Ar2 Ar1

Ar2

B(OH)2

Me

Me

+ HO

Me

Me

Figure 26.12  Chemoselective boronic ester synthesis by controlled speciation [67] / Thieme Medical Publishing Group.

Ar

ArX Pd0

Path A: Slow addition of reagent Path B: Bulk addition of reagent

Rate Determining Step

X PdII Ar

Ar PdII X

Path A

N2

Path B

Decomposition and Deletion

Figure 26.13  Catalytic cycle for the cross-coupling of diazo nucleophiles with slow addition of reagent (Path A) and bulk addition of reagent (Path B) [70] / American Chemical Society.

26.7  Alternative Metal Catalysts

[70]. This approach was successfully reported for the diazo cross-coupling reaction by slow addition of the diazo reagent. These conditions modified speciation within the catalytic cycle of the Pd intermediate so the oxidative addition Pd(II)ArCl intermediate becomes the resting state. By adding the diazo reagent slowly, the reaction with the diazo reagent became the resting state in the catalytic cycle. This strategy greatly expands the scope of the diazo cross-coupling reaction. Finally, we highlight a defined heterogeneous Pd catalyst based on a phosphine-metal–organic framework (P-MOF) ligand system [71]. This system catalyzes Suzuki reactions under exceptionally mild conditions and still displayed higher selectivity than that achieved using Pd(PPh3)4, a standard homogeneous catalyst. The Pd-P-MOF catalyst converted a wide range of substrates and that, when formed, Pd nanoparticles promote hydrodehalogenation. The Pd reactivity is maximized by having excess phosphine ligand present in the system. The main drawbacks of these systems are the limited stability and recyclability of the MOF framework under the crosscoupling reaction conditions. We conclude that, understanding speciation is becoming an increasingly important factor and tool for improving the reaction outcome of cross-coupling reactions.

26.7  Alternative Metal Catalysts 26.7.1  Nickel Although Pd catalysts have received the majority of the attention for Suzuki-Miyaura coupling reactions in the literature, other metals have also been explored [67]. Currently, there is an interest in the field to move toward utilizing earth abundant and “green sustainable” metals to catalyze the cross-coupling reactions. The first major alternative metal that has been employed in cross-coupling reactions is nickel (Ni) [72]. In general, Ni’s main advantage is its earth abundance and significantly cheaper to produce compared to Pd. Ni-catalyzed Suzuki-Miyaura couplings often employ phosphine or diamine ligands to carry out the reactions, Figure 26.14 [73–75] This catalyst pairing allows for several interesting transformations which are difficult to carry out with traditional Pd-catalyzed coupling reactions. Of particular interest is the ability of Ni-catalysts to form quaternary centers through the coupling of an sp2 hybridized organoboron species with an sp3 hybridized halide, Table 26.1 [74]. In addition, Ni can also be utilized to form chiral centers via the coupling of two sp3 carbons with acceptable levels of stereospecificity [75, 76]. Although the stereoinduction is relatively modest, it remains significant as this kind of coupling has been difficult to accomplish with Pd catalysts [77]. Although Ni-complexes have some distinct advantages over Pd catalysts, unfortunately there are major drawbacks which limit their utility. Ni-catalysts frequently induce radical reactions during the cross-coupling reactions [78]. This is because Ni(I) is more readily accessible than Pd(I) [79]. Consequently, Ni can induce radical reactions which can interfere with the desired cross-coupling reactions [80]. However, this radical formation is the sub-reaction that supports the formation of the quaternary carbon centers (outlined previously) and thus expanding the applicability of the resulting products [75–77]. Since Ni is a first-row transition metal ion, compared to the second row Pd atom, the Ni-coordination complexes are likely to be less labile than the Pd complex, and the Ni-speciation chemistry is less complex. Therefore these cross-coupling catalysts can be a good alternative to the Pd catalysts when optimized.

569

570

26  Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts

t-Bu

t-Bu

P N

N

Ph

Ph

MeHN

NHMe

Figure 26.14  Common ligands used for Ni-catalyzed Suzuki cross-couplings: phosphine ligands, bipyridine ligands, and diamine [73–75]. Phosphine [68–70].

Table 26.1  Cross-coupling of tertiary centers using Ni [72].

26.7.2  Cobalt Cobalt (Co) has recently received significant attention as an alternative metal to Pd. Co, as an essential element, has significantly lower toxicity than Pd and is much more earth abundant [81]. Co was first employed in cross-coupling reactions in the early part of the 20th century [82, 83]. However, the interest in these systems were limited until recently when an emphasis on the development of more earth abundant systems became important [84, 85].

26.7  Alternative Metal Catalysts

Recent work on cobalt catalysts has focused on multiple types of cross-coupling [86, 87] including Suzuki-Miyaura couplings utilizing pincer type ligands, Figure 26.15 [85, 88, 89]. These systems have allowed for the formation of C(sp2)–C(sp3) bonds utilizing a diamine ligand in acceptable yields with several examples with varied functionality shown in Table 26.2[85]. Although this required a relatively high level of catalyst loading, this is not as prohibitive for Co as it is for Pd due to the lower cost and less toxic properties of many Co-complexes [81]. Other reports showed the utilization of a set of Co(II) catalysts to carry out the coupling of B(sp2) boronic acids with C(sp2) electrophiles [88]. These reactions proceeded with much lower levels of catalyst loading with good to excellent yields of the desired products [88]. The investigations into Co as a cross-coupling catalyst have opened new avenues for exploration, Table 26.2 [81]. Although parts of the mechanism of the Suzuki-Miyaura cross-coupling reaction are still under investigation [90, 91], the use of Co catalysts allows for additional methods of investigation of the cross-coupling reactions in general. Although cobalt is an NMR active nucleus, the Co-catalyst involved in the cross-coupling reaction [Co(II)], is paramagnetic and of low sensitivity. It may provide additional characterization of the inorganic portion of the reaction as well as the kinetics assessment of the organic portion of the reaction. New information of how the SuzukiMiyaura cross-coupling reaction proceeds with earth abundant catalysts is likely to be helpful for future catalyst design and evaluation of the performance of the corresponding Pd catalysts and potentially lower the catalyst loadings. Ph PiPr2 N Co Cl PiPr2

N N

N + CoCl2

H Me H Br N Co Br N H H Me

Figure 26.15  Common ligands used for Co catalyzed cross-coupling reactions [85, 88, 89]. Table 26.2  C(sp2)-C(sp3) coupling catalyzed by Co [81].

571

572

26  Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts

26.8  Speciation of Nickel and Cobalt Catalysts Ni-pre-catalysts are a sustainable alternative to Pd pre-catalysts for the Suzuki−Miyaura crosscoupling reaction [80]. Aryl sulfamates are phenolic derivatives which can act as a directing group for the pre-functionalization of the electrophilic aromatic backbone prior to cross-coupling. Coordination complexes of Ni(0), Ni(I), and Ni(II) can form from the pre-catalyst complex, (dppf) Ni(o-tol)(Cl) (where dppf = 1,1′ bis(diphenylphosphino)-ferrocene), for Suzuki−Miyaura coupling reactions involving aryl sulfamates and boronic acids (Figure 26.16). Catalysts formed from Ni, aryl sulfamates, and boronic acids can function effectively at low catalyst loading and at milder reaction conditions than reported Pd systems. The reactions of dppf-Ni(0) with alkyl halides proceeds through three-coordinate Ni(0) intermediates, such as [Ni(dppf)(L)] [92]. The effect of added ligand (L) on catalyst speciation and the rates of reactions of [Ni(COD)(dppf)] with alkyl halides have been investigated. Trends in reactivity with regard to steric bulk and electron donation of monodentate ligands was found to decrease the reaction rate. The halide abstraction step is generally irreversible and the subsequent recombination of a Ni(I) complex with an alkyl halide have a significant effect on the overall rate of the reaction and some ligands form very stable [Ni(dppf)(L)2] species. This study established that different ligands increase the rate of the reaction between [Ni(COD)(dppf)] and a model alkyl bromide giving a 200-fold spread in rate constants. Another study examined the cross-coupling reactions of Ni(0)(dppf) complexes with alkyl halides which proceeded through three-coordinate nickel(0) intermediates in the form [Ni(dppf) (L)], Figure 26.17 [92]. A series of monodentate ligands (L) are found to affect catalyst speciation; overall sterically bulky and electron donating ligands decrease the reaction rates of [Ni(COD) (dppf)] with alkyl halides. Some ligands form very stable [Ni(dppf)(L)2] species. The yields of prototypical NI(0)(dppf)-catalyzed Kumada cross-coupling reactions of alkyl halides are significantly improved by the addition of free ligands, which demonstrates speciation provides another important variable to consider when optimizing nickel-catalyzed reactions of alkyl halides.

OSO2NMe2

OSO2NMe2 F3C

MeO

EWG + 1 equiv.

2.5mol% [Ni] 4.5 equiv. K3PO4

CF3

OMe

toluene, 80 oC 24 hr

EDG

MeO

B(OH)2

MeO

MeO [Ni] = P

EWG/EDG Product Ratio

NiII Cl

1.86 +/– 0.04

NiI Cl

1.66 +/– 0.11

P P Ni0 P P

1.67+/– 0.11

P P P

Figure 26.16  Preferential product Formation for SMC reactions catalyzed by (dppf)Ni(II)(o-tol)(Cl), (dppf)Ni(I) (Cl), (dppf)2Ni(0) with aryl sulfamates of differing electronic properties and only 1 equiv. of boronic acid [80].

26.8  Speciation of Nickel and Cobalt Catalysts

Figure 26.17  Excess ligand affects speciation and reaction rates from [92].

Optimizing the transmetalation reaction from a nucleophile to a metal catalyst led to the first Co-catalyzed Suzuki-Miyaura cross-coupling reaction: Cross-coupling between aryl triflate electrophiles and heteroaryl boron nucleophiles catalyzed by a Co complex, Figure 26.18 [93]. The reaction was catalyzed by a new class of high-spin, tetrahedral, bis(phosphino)pyridine Co(I) alkoxide or aryloxide complexes converting to a low-spin square planar cobalt(I) catalytic intermediate. The reactivity of methyl- or ethylalkoxide Co(I) complexes facilitated transmethylation at ambient temperature, but the concurrent β-hydride elimination reaction made these complexes unsuitable as cross-coupling catalysts. The higher stability of isopropyl or CH(Ph)CH3 alkoxide Co(I) complexes facilitated transmethylation at ambient temperature and were found to be suitable as cross-coupling catalysts. The favorable speciation chemistry of the first-row cobalt complexes over state-of-the-art second row Pd complexes allowed optimizing the cross-coupling reaction conditions to result in the first Co catalyzed C(sp2)-C(sp2) bond formation between neutral pinacolate boron nucleophiles and aryl triflate electrophiles [93]. Since 2016, exploration into the speciation chemistry of Co-catalysts and the applications of a range of Co-catalyst for the Suzuki-Miyaura cross-coupling reactions have dramatically expanded. At this time many more Co-catalysts including halides, NHC-ligands, peptides, and pre-catalysts, as well as a variety of nucleophiles and electrophiles, have been successfully employed.

Cobalt–catalyzed Suzuki–Miyaura cross coupling... OTf + PinB

O

O Co 85% yield

transmetalation N1

2-BfBpin

O1

Co1

N1 P2

Co1

P1 O1

C20

P2

P1

S=0 S=1

...enabled by transmetalation with neutral boron reagents.

Figure 26.18  Cobalt-catalyzed Suzuki-Miyaura cross-coupling reaction. Ref [93] / John Wiley & Sons.

573

574

26  Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts

These recent studies have demonstrated that potential of the earth abundant metal, Co, as catalyst for the Suzuki-Miyaura cross-coupling reaction has materialized as a viable alternative for Pd-catalyzed reaction conditions in the future [88, 94–97].

26.9  Cross-coupling Reactions and Sustainability: Summary and the Future In this review, we have described the popular cross-coupling reaction, its history, and pertinent efforts toward developing the global catalyst system as well as selected current developments toward a green catalyst system. Specifically, we have highlighted advances using Ni and Co based catalysts as representatives for the numerous earth abundant metals that are currently investigated as greener catalyst for the cross-coupling reaction. In addition to describing the suitable substrates and scope of the reactions, we show that analyzing the reactions mechanistically particularly by considering the associated speciation chemistry of the catalysts under the reaction conditions can be used to effectively improve coupling yields. Recent studies have demonstrated that simple strategies in addition to those commonly used including the addition of a small excess of catalysts ligand and continuous addition of reagents are methods that will improve yields even with earth abundant metal catalysts. Hence, we believe that speciation principles can assist the future catalyst design particularly as the systems move into more green and sustainable solvents that will be more suitable for larger scales and milder reaction conditions important for specific industrial processes.

References 1 Marshall, J.A. and Johns, B.A. (1998). J. Org. Chem. 63: 7885–7892. 2 Flick, A.C. et al. (2017). J. Med. Chem. 60 (15): 6480–6515. 3 Nobel Prizes 2010: Richard F. Heck / Ei-ichi Negishi / Akira Suzuki. Angew. Chem. Int. Ed. 49: 8300. 4 Brown, D.G. and Boström, J. (2016). J. Med. Chem. 59: 4443–4458. 5 Suzuki, A. and Miyaura, N. (1995). Chem. Rev. 95 (7): 2457–2483. 6 Hosseini, M.-J. et al. (2016). Metallomics 8 (2): 252–259. 7 Miyaura, N., Yanagi, T., and Suzuki, A. (1981). Synth. Commun. 11: 513. 8 Johanssen Seechurn, C.C.C., Kitching, M.O., Colacot, T.J., and Snieckus, V. (2012). Angew. Chem. Int. Ed. 51: 5062. 9 Dieck, H.A. and Heck, R.F. (1974). J. Am. Chem. Soc. 96 (4): 1133. 10 Old, D.W., Wolfe, J.P., and Buchwald, S.L. (1998). J. Am. Chem. Soc. 120 (37): 9722–9723. 11 Wolfe, J.P., Singer, R.A., Yang, B.H., and Buchwald, S.L. (1999). J. Am. Chem. Soc. 121 (41): 9550–9561. 12 Huang, X., Anderson, K., Zin, D. et al. (2003). J. Am. Chem. Soc. 125 (22): 6653–6655. 13 Walker, S.D., Barder, T.E., Martinelli, J.R., and Buchwald, S.L. (2004). Angew. Chem. Int. Ed. 43: 1871–1876. 14 Burgos, C.H., Barder, T.E., Huang, X., and Buchwald, S.L. (2006). Angew. Chem. Int. Ed. 45 (26): 4321–4326. 15 Fors, B.P., Watson, D.A., Biscoe, M.R., and Buchwald, S.L. (2008). J. Am. Chem. Soc. 130 (41): 13552–13554.

References

16 Pickett, T.E., Roca, F.X., and Richards, C.J. (2003). J. Org. Chem. 68: 2592. 17 Marion, N. and Nolan, S.P. (2008). Acc. Chem. Res. 41: 1440. 18 Fleckenstein, C.A. and Plenio, H. (2010). Chem. Soc. Rev. 39: 694. 19 Littke, A.F. and Fu, G. (2002). Angew. Chem. Int. Ed. 41: 4176. 20 Thompson, W.J., Jones, J.H., Lyle, P.A., and Thies, J.E. (1988). J. Org. Chem. 53: 2052. 21 Sather, A.C., Lee, H.G., De La Rosa, V.Y. et al. (2015). J. Am. Chem. Soc. 137 (41): 13433–13438. 22 Wu, J. and Chan, A.S.C. (2006). Acc. Chem. Res. 39: 711–720. 23 Martin, R. and Buchwald, S.L. (2008). Acc. Chem. Res. 41 (11): 1461–1473. 24 Dai, W. and Zhang, Y. (2005). Tet. Lett. 46: 1377–1381. 25 Kinzel, T., Zhang, Y., and Buchwald, S.L. (2010). J. Am. Chem. Soc. 132 (40):14073–14075. 26 Yang, Y. and Buchwald, S.L. (2013). J. Am. Chem. Soc. 135: 10642–10645. 27 Hillier, A.C. and Nolan, S.P. (2002). Platin. Met. Rev. 46 (2): 50–64. 28 Hadei, N., Kantchev, E.A.B., O’Brie, C., and Organ, M.G. (2005). Org. Lett. 7 (17): 3805–3807. 29 Froese, R.D.J., Lombardi, C., Pompeo, M. et al. (2017). Acc. Chem. Res. 50 (9): 2244–2253. 30 Jalaj, M., Hammouti, B., Touzani, R. et al. (2020). Mater. Today Proc. 31: S122–S129. 31 Herrmann, W.A. (2002). Angew. Chem. Int. Ed. 41 (8): 1290–1309. 32 Voloshkin, V.A., Tzourasa, N.V., and Nolan, S.P. (2021). Daltons Trans. 50: 12058–12068. 33 Heravi, M.M., Zadsirjan, V., Kafshdarzadeh, K., and Amiri, Z. (2020). Asian J. Org. Chem. 9 (12): 1999–2034. 34 Fortmana, G.C. and Nolan, S.P. (2011). Chem. Soc. Rev. 40: 5151–5169. 35 Hiller, A.C., Grasa, G.A., Viciu, M.S. et al. (2002). J. Organomet. Chem. 653: 69–82. 36 Organ, M.G., Chass, G.A., Fang, D.-C. et al. (2008). Synthesis 17: 2776–2797. 37 Valente, C., Pompeo, M., Sayah, M., and Organ, M.G. (2014). Org. Process Res. Dev. 18 (1): 180–190. 38 Zhao, Q., Meng, G., Nolan, S.P., and Szostak, M. (2020). Chem. Rev. 120 (4): 1981–2048. 39 Zhang, C., Huang, J., Trudell, M.L., and Nolan, S.P. (1999). J. Org. Chem. 64 (11): 3804–3805. 40 Farmer, J., Hunter, H.N., and Organ, M.G. (2012). J. Am. Chem. Soc. 134 (42): 17470–17473. 41 O’Brien, C.J., Kantchev, E.A.B., Valente, C. et al. (2006). Chem. Eur. J. 12: 4743–4748. 42 Marion, N., Navarro, O., Mei, J. et al. (2006). J. Am. Chem. Soc. 128 (12): 4101–4111. 43 Diebolt, O., Braunstein, P., Nolan, S.P., and Cazin, C.S.J. (2008). Chem. Commun. 3190–3192. 44 Pompeo, M., Froese, R.D.J., Hadei, N., and Organ, M.G. (2012). Angew. Chem. Int. Ed. 51: 11354–11357. 45 Atwater, B., Chandrasoma, N., Mitchell, D. et al. (2015). Angew. Chem. Int. Ed. 127: 9638–9642. 46 Yamamoto, Y., Takada, S., and Miyaura, N. (2006). Chem. Lett. 35: 704. 47 Yamamoto, Y., Takada, S., and Miyaura, N. (2006). Chem. Lett. 35: 1368. 48 Yamamoto, Y., Takada, S., and Miyaura, N. (2009). Organometallics 28: 152. 49 Valente, C., Belowich, M.E., Hadei, N., and Organ, M.G. (2010). Eur. J. Org. Chem. 2010 (26): 4343–4354. 50 Price, G.A., Hassan, A., Chandrasoma, N. et al. (2017). Angew. Chem. Int. Ed. 129 (43): 13532–13535,4. 51 Sun, B., Ning, L., and Zeng, H.C. (2020). J. Am. Chem. Soc. 142: 13823–13832. 52 Scott, N.W.J., Ford, M.J., Jeddi, N. et al. (2021). J. Am. Chem. Soc. 143 (25): 9682–9693. 53 Pagliaro, M., Pandarus, V., Ciriminna, R. et al. (2012). ChemCatChem. 4: 432–445. 54 Silva, S., Almeida, A.J., and Vale, N. (2019). Biomolecules 9: 22. 55 Hong, K. et al. (2020). ACS Appl. Nano Mater. 3: 2070–2103. 56 Marck, G., Villiger, A., and Buchecker, R. (1994). Tet. Lett. 35 (20): 3277–3280. 57 Pedersen, L., Mady, M.F., and Sydnes, M.O. (2013). Tet. Lett. 54: 4772–4775. 58 Adrio, L.A., Nguyen, B.N., Guilera, G. et al. (2012). Catal. Sci. Technol. 2: 316–323.

575

576

26  Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts

59 Deng, Y., Gong, L., Mi, A. et al. (2003). Synthesis 2003 (3): 337–339. 60 Skapski, A.C. and Smart, M.J.L. (1970). J. Chem. Soc. D 658. 61 Kirik, S.D. (2004). Acta. Crystallogr. Sec. C. Cryst. Struct. Commun. 60: M449–M450. 62 Stephenson, T.A., Morehouse, S.M., Powell, A.R. et al. (1965). J. Chem. Soc. (0): 3632–3640. 63 Pandey, R.N. et al. (1974). Can. J. Chem. 52: 1241–1247. 64 Bukhmutov, V.I., Berry, J.F., Cotton, F.A. et al. (2005). Daltons Trans. (11): 1989–1992. 65 Stoyanov, E.S. (2000). J. Struc. Chem. 41: 440–445. 66 Evans, J., O’Neill, L., Kambhampati, V.L. et al. (2002). J. Chem. Soc. Dalton Trans. (10): 2207–2212. 67 Fyfe, J.W.B. and Watson, A.J.B. (2015). Synlett 26: 1139–1144. 68 Fyfe, J.W.B., Valverde, E., Seath, C.P. et al. (2015). Chem. Eur. J. 21: 8951–8964. 69 Fyfe, J.W.B., Seath, C.P., and Watson, A.J.B. (2014). Angew. Chem. Int. Ed. 53: 12077–12080. 70 Sullivan, R.J., Freure, G.P.R., and Newman, S.G. (2019). ACS Catal 9: 5623–5630. 71 Cartagenova, D., Bachmann, S., Püntener, K. et al. (2022). Catal. Sci. Technol. 12: 954. 72 Tamao, K., Sumitani, K., and Kumada, M. (1972). J. Am. Chem. Soc. 94: 4374–4376. 73 Percec, V., Bae, J.Y., and Hill, D.H. (1995). J. Org. Chem. 60: 1060. 74 Zultanski, S. and Fu, G. (2013). J. Am. Chem. Soc. 135: 624–627. 75 Lu, Z., Wilsily, A., and Fu, G. (2011). J. Am. Chem. Soc. 133 (21): 8154–8157. 76 Zultanski, S. and Fu, G. (2011). J. Am. Chem. Soc. 133 (39): 15362–15364. 77 Cammidge, A.N. and Crépy, K.V.L. (2000). Chem. Comm. (18): 1723–1724. 78 Li, Y., Luo, Y., Peng, L. et al. (2020). Nat. Commun. 11: 417. 79 Lin, Q., Dawson, G., and Diao, T. (2021). Synlett 32: 1606–1620. 80 Mohadjer Beromi, M., Nova, A., Balcells, D. et al. (2017). J. Am. Chem. Soc. 139: 922–936. 81 Leyssens, L., Vinck, B., Van Der Straeten, C. et al. (2017). Toxicology 387 (15): 43–56. 82 Gilman, H. and Lichtenwalter, M.J. (1939). J. Am. Chem. Soc. 61: 957. 83 Kharasch, M.S. and Fuchs, C.F. (1941). J. Am. Chem. Soc. 63: 2316. 84 Uemara, S. and Fukuzawa, S. (1982). Tet. Lett. 23: 1181. 85 Ludwig, J., Simmons, E., Wisniewski, S., and Chirik, P. (2021). Org. Lett. 23: 625–630. 86 Gosmini, C., Bégoiun, J.-M., and Moncomble, A. (2008). Chem. Comm. (28): 3221–3223. 87 Hammann, J.M., Hoffmayer, M.S., Lutter, F.H. et al. (2017). Synthesis 49 (17): 3887–3894. 88 Asghar, S., Tailor, S.B., Elorriaga, D., and Bedford, R.B. (2017). Angew. Chem. Int. Ed. 56: 16367. 89 Obligacion, J.V., Semproni, S.P., Pappas, I., and Chirik, P.J. (2016). J. Am. Chem. Soc. 138: 10645–10653. 90 Gülak, S., Stepanek, O., Malberg, J. et al. (2013). Chem. Sci. 4: 776. 91 Kreyenschmidt, F., Meurer, S.E., and Koszinowski, K. (2019). Chem. Eur. J. 25: 5912–5921. 92 Greaves, M.E., Ronson, T.O., Maseras, F., and Nelson, D.J. (2021). Organometallics 40: 1997–2007. 93 Neely, J.M., Bezdek, M.J., and Chirik, P. (2016). ACS Cent. Sci. 2 (12): 935–942. 94 Arevalo, R. and Chirik, P.J. (2019). J. Am. Chem. Soc. 141 (23): 9106–9123. 95 Mills, L.R., Gygi, D., Ludwig, J.R. et al. (2022). ACS Catal 12: 1905–1918. 96 Tailor, S.B., Manzotti, M., Smith, G.J. et al. (2021). ACS Catal. 11 (7): 3856–3866. 97 Piontek, A., Ochędzan-Siodłak, W., Bisz, E., and Szostak, M. (2020). ChemCatChem. 13 (1): 202–205.

577

27 Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions Ana P. Carvalho1,2, Angela Martins2,3, Filomena Martins1,2, Nelson Nunes2,3, and Rúben Elvas-Leitão2,3 1

Departamento de Química e Bioquímica, Faculdade de Ciências da Universidade de Lisboa, Campo Grande, Lisboa, Portugal Centro de Química Estrutural, Institute of Molecular Sciences, Faculdade de Ciências, Universidade de Lisboa, Campo Grande, Lisboa, Portugal 3 DEQ, Instituto Superior de Engenharia de Lisboa, IPL, R. Conselheiro Emídio Navarro, Lisboa, Portugal 2

27.1  Introduction Friedel-Crafts reactions are among the best-known types of organic reaction of aromatic molecules. Their name derives from the two chemists who first studied them c. 1877, the French Charles Friedel and the American James Mason Crafts [1, 2] These reactions can be included in a broader classification; that is, electrophilic aromatic substitutions, in which the aromatic rings can be substituted by halogen, nitro groups, alkyl groups, and others. Because aromatic molecules are much less reactive compared with simple alkanes, electrophilic aromatic substitutions require a catalyst or specific experimental conditions to increase the electrophilic nature of the substituent. Classical examples of this strategy include the use of FeBr3 in the bromination of benzene or the mixture of concentrated HNO3 and H2SO4 to generate NO2+ used in aromatic nitration. Starting from the original Friedel and Crafts papers, knowledge regarding these reactions has expanded over the years, leading to the publication of numerous books and papers on the subject. One of the most important works was the four-volume book Friedel-Crafts and Related Reactions, edited by George Olah [3], in which the author stated that different types of reactions could be considered as Friedel-Crafts type if taking place under the catalytic effect of a Lewis acid type or of protic acids [4]. Despite the strict definition of Friedel-Craft reactions, the term is still used in a broader sense and includes different processes, even biological pathways. An example is vitamin K1 biosynthesis where, in the absence of a catalyst, the carbocation electrophile is formed by the dissociation of an organodiphosphate in a Friedel-Crafts look alike reaction. Friedel-Crafts reactions can be divided in two types: alkylations and acylations and, as the name implies, the difference is based on the type of electrophile substrate, an alkyl or an acyl group. The original methods used to perform these reactions combined AlCl3 as catalyst and an alkyl halide or an acyl halide, respectively, for alkylations or acylations.

Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 2, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

578

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions

However, Friedel-Crafts reactions currently include those with a range of different alkylating agents, namely: activated and unactivated alkenes, alkynes, paraffins, alcohols, ethers, carbonyl, and other acylating agents such as carboxylic acids, carboxylic acid derivatives, esters, and anhydrides [3]; as well as the use of different Lewis acid catalysts such as AlBr3, FeCl3, SbCl5, TiCl4, ZnCl2, SnCl4, and BF3 and of strong Brӧnsted acids such as HF, H3PO4, and H2SO4. Progress in developing these types of reactions lead, additionally, to the current use heterogeneous catalysts, as described further ahead in this chapter, and even of asymmetric catalysts. The simplified mechanism for the classic acylation reaction (analogous to alkylation) is depicted in Figure 27.1 showing its main steps: carbocation formation with assistance of AlCl3, attack of the aromatic ring leading to the formation of a C–C bond and a new carbocation intermediate, and, finally, proton loss with the formation of the neutral acylated substitution product. The formation of acylium ions is favored in polar solvents or when using anhydrides esters with strong proton acids. Despite the synthetic opportunities provided by this class of reactions, namely the creation of C–C bonds, there are several intrinsic limitations to Friedel-Crafts reactions: aromatic and vinyl halides usually do not react, and the aromatic ring should not have electron-withdrawing groups. Another major challenge for these reactions is in finding ways to reduce possible carbocation rearrangements and limiting the number of substitutions in the aromatic substrate. However, these constraints are not problematic in acylation reactions, since being the acyl carbocation more stabilized rearrangements do not occur. Moreover, as the aromatic acylation products are less reactive, this reaction produces essentially monosubstituted products. In fact, the most efficient way to introduce a -CH2- group in a benzene ring is to perform an acylation, produce a ketone, and afterwards reduce the carbonyl group. Traditional Friedel-Crafts acylation reactions have also some drawbacks compared to alkylations, as they require equimolar quantities of catalysts (e.g. AlCl3) due to the formation of an acid/base complex of the acyl halide. Both Friedel-Crafts alkylations and acylations are important synthetic tools used in industry, as they are used to prepare chemical feedstock, synthetic intermediates, and fine chemicals [5]. The large

Figure 27.1  Acylation general mechanism.

27.1  Introduction

number of reactions also means that different chemical, engineering, environmental factors, and economic factors can affect them. Because the focus of this chapter is on acylation reactions, only their practical challenges will be detailed next. The most important products of Friedel-Crafts acylations are aromatic ketones for which synthetic methods can use different procedures, solvents, acylating agents, and catalysts. Table 27.1 illustrates several examples that demonstrate the diverse nature of the obtained ketones and their uses, according to the different experimental options taken. The choice of catalyst plays a major role in Friedel-Crafts acylation reactions. The traditional catalyst is AlCl3, and less active catalysts include BF3 and SnCl4. Other strong acids such as H2SO4, HClO4, and HO[P(OH)(O)O]nH can also be used when the acylating agents are anhydrides, esters, or carboxylic acids. One of the main problems of traditional catalysts is the large stoichiometric amounts needed, that is a threefold amount in the case of anhydrides (i.e. these catalysts act more like reagents). Separation is also a drawback, and it is even more problematic since these catalysts cannot be recycled. Additionally, they are corrosive, making the separation difficult and creating the need to neutralize and dispose of large quantities of waste. Economic and environmental problems of homogeneous catalysts have pushed the development of heterogeneous acylation catalysts. These materials are strong solid acids and different alternatives are possible: modified clays, solid superacids, surface mounted acids, and Nafion are all examples of these materials; however, zeolites are the most important ones, and their use will be detailed further in the text. It is worth mentioning that commercial scale processes for the acylation of aromatic compounds (combined with acetic anhydride) have been already developed by Rhône-Poulenc (now Rhodia) using the Beta zeolite (BEA) structure [7]. Other variables, such as the solvents used, can also have an impact on the obtained products by influencing regioselectivity [8]. It is also possible to find in the literature examples of acylating Table 27.1  Aromatic ketones obtained by Friedel-Crafts acylation reactions [6]. Acylating agent

Aromatic compound

Product

Uses

acetic anhydride

benzene

acetophenone

perfumes, plasticizer pharmaceutical, solvent,

acetic anhydride

toluene

4-methylacetophenone

perfumes

acetic anhydride

anisole

4-methoxyacetophenone

perfumes

acetic anhydride

isobutylbenzene

4-isobutylactophenone

pharmaceuticals

dichloroacetyl chloride

1,2-dichloro benzene

α,α,2,4-tetrachloro acetophenone

insecticides

chlorobutyroyl chloride fluorobenzene

chloropropyl 4-fluorophenyl ketone

pharmaceuticals

tetrachloromethane

benzene

benzophenone

pharmaceuticals, insecticides, perfumes

benzoyl chloride

benzene

benzophenone

pharmaceuticals, insecticides, perfumes

phosgene

N,N-dimethyl aniline

4,4’-bis-dimethyl aminobenzophenone

dyes

phthalic anhydride

benzene

2-benzoylbenzoic acid

anthraquinone

579

580

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions

agents acting as solvents in reactions that are considered by some authors as solvent-free acylation reactions [9, 10], or even the resort to more unusual solvents such as ionic liquids [11–13]. Experimental procedures also play a role when working with certain solvents in acyl halide acylations. Three different methodologies are commonly used [6]: reagents are all mixed and cooled and the catalyst added slowly (Elbs method); the acylating agent is added to a cooled solution/ suspension of catalyst, with the substrate added afterwards (Perrier method); or the aromatic substrate is the solvent and the acylating agents is added slowly (Bouveault method). Regarding acylating agents, as stated previously, different compounds can be used in Friedel-Crafts reactions and the selection is obviously dependent on the required substituents (R). However, other aspects should be taken into account, namely the relative reactivity of the acylating agents. Roughly, the reactivity order is the following: [RCO]+[BF4]−≈[RCO]+[ClO4]−>RCOO−>SO3H>RCOX>(RCO)2 O>RCO2R’ >RCONR’2. This order is not absolute because different R substituents influence reactivity, and different acyl halides may also show different reactivities: RCOI>RCOBr>RCOCl>RCOF [6]. The great flexibility of Friedel-Crafts acylation reactions is also connected to the variety of substrates that can be used. There are numerous examples of substrates used in the production of aromatic ketones including benzene and benzene derivatives, among which toluene and aromatic amines should be considered. Also important, in this regard, is the acylation reactions of polynuclear aromatic compounds, viz. naphthalene, biphenyls, anthracene, and phenanthrene. Additionally, heteroaromatic compounds like furan, thiophene, and pyrrole, which are normally activated, can be acylated in milder conditions eventually with significantly lower environmental impact. If one uses a broader definition of Friedel-Crafts acylation, it is also possible to include substrates such as olefins. Friedel-Crafts acylations have been part of numerous synthetic routes for multiple molecules. Some of these routes are still used at present, whereas others have been replaced by more economic or environmentally friendly ones. An example is the already mentioned acetophenone, a product that was commercially manufactured using Friedel-Crafts acylation of benzene by acetic ­anhydride at 30 °C, catalyzed by AlCl3 in an 85% yield reaction. However, acetophenone is now produced industrially as a by-product of the oxidation of ethylbenzene or cumene. The opposite occurs with the synthesis of anthraquinone from phthalic anhydride and benzene, which uses a Friedel-Crafts reaction to produce o-benzoylbenzoic acid in the first step. This method was first described by Heller in 1906 and is still used despite the development of a newer process developed by BASF using styrene (which has also drawbacks). The number of products currently manufactured using synthesis steps involving Friedel-Crafts acylations is massive. The synthesis of tolmetin (1-methyl-5-(4-methylbenzoyl)-1H-pyrrole-2-­ acetic acid) an anti-inflammatory, analgesic–antipyretic medicine, and Naproxen ((+)-6-methoxyα-methyl-2-naphthaleneacetic acid), an anti-inflammatory agent, are only two of the many examples of different products that could be listed here.

27.2  Zeolites and Hierarchical Zeolites Zeolites are solid materials known for more than 250 years that became important catalysts for heterogeneous processes only after the discoveries made by Milton and Barrer in the middle of the last century [14, 15]. In fact, the studies of these scientists were the first ones to establish the synthesis protocols to obtain zeolitic structures, some found in nature and others without a natural counterpart, making the use of these solids in industrial applications possible. At present, according to the Structure Commission of the International Zeolite Association, the number of synthetic zeolites is higher than 240, as over 67 natural structures have been identified [16].

27.2  Zeolites and Hierarchical Zeolites

Structurally, zeolites are very organized frameworks resulting from the different forms by which SiO4 and AlO4− tetrahedrons are linked, sharing one or more oxygens to create cavities and channels with openings in the range of microporosity (>CH3COOH-BEA≈HBEA. According to the authors, the reason for this trend may be that some Brӧnsted acid sites, which were covered by non-acidic amorphous substances on the external surface, would be exposed during the treatment with HCl and HNO3. In the presence of CH3COOH, a weak organic acid, only a small part of the amorphous substance was removed. Upon deactivation, mainly by deposition of carbonaceous deposits, the zeolite samples were dried and calcined at 450 °C for five hours. The reusability of BEA based catalysts was demonstrated after nine catalytic cycles. Indeed, the average lifetime of H-BEA was kept at 2911 minutes before the deactivated point (considered when the conversion of thiophene was below 90%), during the nine catalytic cycles, which was roughly identical to the capacity of CH3COOH-BEA sample (3061 min). As expected, the HCl-BEA and HNO3BEA showed superior reusability performance, with 4863 and 4893 minutes, respectively. Koehle et al. [76] investigated the catalytic behavior of Brӧnsted and Lewis acid zeolite catalyst in the acylation of methylfuran in the presence of AA at 110 °C at 14 bar. The reaction was performed using two acid BEA zeolites, with high and low alumina contents, (Si/Al = 23 and 138, respectively). The reaction was also carried out in the presence of various Lewis acid BEA zeolites, that is, Sn, Zr, Hf, or Ti loaded BEA. The authors showed that the highest reaction rate was obtained with the BEA (Si/Al = 23) whereas the higher turnover frequency was found for BEA (Si/Al = 138). Among Lewis acid zeolites Sn-BEA shows higher turnover frequency when compared to the other metal loaded zeolites. Electronic structure calculations showed that, the presence of both Brönsted and Lewis acid sites, follow the same mechanism (i.e. the classic addition-elimination aromatic electrophilic substitution). An unexpected finding that resulted from the authors calculations was that the most favorable catalytic pathway in Lewis rich zeolite Sn-BEA sample also involves a Brӧnsted acid catalysis step carried out by the silanol groups present on the material. The operation conditions, particularly the role of the heating source was explored by Shekara et al. [77] who studied solventless liquid-phase acylation of p-cresol with different acylating agents, namely, acetic, propionic, butyric, hexanoic, octanoic, and decanoic acids, using BEA zeolite (Si/Al = 30) under conventional and microwave heating. A significant increase in activity was observed when the reaction occurred under microwave radiation. In fact, under conventional heating, the conversion with all acylating agents was less than 20%, whereas under microwave heating the conversions ranged between 50 to 80%. Moreover, the characterization of the microwave spent catalysts showed the absence of coke/coke precursors in contrast to catalysts used under conventional heating. The authors pointed out that, under conventional heating, the initial inhibition and coke formation occurred due to the preferential adsorption of the acylating agent. As p-cresol is chiefly responsible for heating up the reaction mixture, the heat-up energy released

27.3  Zeolites and Hierarchical Zeolites as Catalysts for Friedel Crafts Acylation Reactions

when the reaction occurs under microwave radiation might be affecting the adsorption of the acylating agent, suppressing the formation of coke/coke precursors. This study opened new perspectives to alternative heating sources in Friedel-Crafts acylation reaction, aiming to suppress one of the major drawbacks in the use of zeolitic materials (i.e. the fast deactivation due to the formation of coke/coke precursors that become trapped inside the zeolite porosity). To overcome the disadvantages inherent to the use of purely microporous zeolites, hierarchical materials started to appear in the literature from the beginning of the 21st century, as referred before. One of the first examples on the application of hierarchical zeolites in Friedel-Crafts acylation reactions was presented by Derouane et al. in 2004 [78]. The authors prepared nano-sized H-BEA, starting from a commercial H-BEA zeolite with Si/Al = 18, through the confined space synthesis (CSS) method, where the zeolite was synthesized within the pores of a carbon black matrix, that is, a hard templated bottom-up method. Nano-sized H-BEA (n-H-BEA), with crystal sizes in the range 0.01–0.02 μm, and agglomerates of about 5 μm were obtained. The catalytic behavior of parent and modified zeolites was studied in the Friedel-Crafts acylation of AN by AA at 90 °C for five hours, where it was shown that n-H-BEA sample has a superior catalytic performance since the zeolite nanocrystals decrease the diffusional constraints that limit the egression of the voluminous product, p-methoxyacetophenone (p-MOAP) from the small crystall zeolites with very short diffusion path. At the time, the authors anticipated the future application of hierarchical zeolite materials, comprising nano-sized and micro+mesoporous materials, in many fine and specialty chemical organic synthesis reactions, when one of the products is strongly adsorbed inside the zeolite native micropores. The optimization of Brӧnsted and Lewis acidity as a consequence of the generation of intracrystalline mesoporosity in HZSM-5 zeolite (MFI structure) was studied by Silva et al. [79]. The authors performed desilication treatments on commercial MFI zeolite (Si/Al = 12 and 23) with NaOH, changing the base concentration, temperature, and treatment time. The characterization of the modified samples showed a significant modification on the textural parameters, estimated from N2 adsorption isotherms (Table 27.2). For example, sample Z23-at5, treated with a 0.2 mol L–1 NaOH solution for 90 minutes at 85 °C reached a mesoporous volume of 0.22 cm3 g–1 and an external surface area of 160 m2 g–1 against 0.02 cm3 g–1 and 12 m2 g–1, respectively, for the parent Z23 (HZSM-5,

Table 27.2  Alkaline treatment conditions, Si/Al ratio, and textural properties of parent and desilicated ZSM-5 zeolite [79]. Porous volume (cm3 g–1)

Treatment conditions

Sample

[NaOH] (mol L–1)

Z23

−

T (°C)

−

t (min)

−

Si/Al1

RC (%)

23

100

External area (m2 g–1)

12

micro

meso

total2

0.15

0.02

0.17

Z23-at1

0.2

65

30

18

108

14

0.16

0.02

0.19

Z23-at2

0.2

65

48

17

107

21

0.16

0.04

0.19

Z23-at3

0.2

85

30

17

101

64

0.15

0.09

0.24

Z23-at4

0.2

85

48

14

87

136

0.12

0.17

0.29

Z23-at5

0.2

85

90

14

78

160

0.10

0.22

0.32

Z23-at6

0.4

85

48

9

48

172

0.07

0.36

0.53

Z23-at7

0.4

85

90

8

43

149

0.07

0.34

0.41

1)  Determined by energy-dispersive X-ray spectroscopy (EDX), 2)  total pore volume estimated at p/p0=0.95.

589

590

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions

Si/Al = 23). Nevertheless, some loss of microporosity was also noted, with 0.10 cm3 g–1 for the treated sample against 0.15 cm3 g–1 for the parent material. Modifications on the profile of the acid sites also occurred, with a decrease in the number of catalytically strong active sites (ht-AS), and a pronounced generation of Lewis acid sites, resulting in a substantial decrease in Brӧnsted to Lewis acid ratio (B/L). Again, for Z23-at5 sample the B/L ratio was 3.2 against 21 for the parent Z23. The catalytic performance was studied in the acylation of AN with AA at 100 °C for two hours using a ratio of AA to AN of 2:1. Figure 27.7a presents the specific activity, calculated as the ratio of the main product, 4-methoxyacetophenone (4-MAP) per ht-acid site (denominated as 4-MAP/ht-AS). As can be observed, the development of mesoporosity caused a substantial increase in the specific activity, reaching a maximum with Z23-at5 sample. It is worth noticing that the sample with the highest mesoporous volume, Z23-at7, presents a decrease in the specific activity, which can be attributed to a high decrease in microporous volume (see Table 27.2) as well as a B/L ratio of 2.7, which is lower than that for all other samples. The most promising sample, Z23-at5, was selected for reusability assays. Thus, the sample was recovered after each cycle, and calcined at 600 °C under air atmosphere (5° C min–1) for five hours. As can be observed from Figure 27.7b, sample Z23-at5 retained approximately 90% of the specific activity for 4-MAP for at least three catalytic cycles. Some loss of specific activity can be ascribed to some loss of active sites during the regeneration treatments. To explore the influence of the textural modification performed on desilicated zeolites, Aleixo et al. [46] evaluated the catalytic behavior of MCM-22 (MWW structure), a less assayed zeolite in Friedel Crafts acylation reactions. The parent zeolite was desilicated for 45 minutes at 50 °C with 0.05 or 0.1 mol L–1 NaOH solutions (samples MCM-22/0.05 and MCM-22/0.1, respectively), ­followed by an acid treatment (AT) with 0.1 mol L–1 HCl at 70 °C for six hours [41], giving ­MCM-22/0.05/AT and MCM-22/0.1/AT samples. The characterization of the samples through low temperature N2

Figure 27.7  Specific activity (4-MAP/ht-AS) for anisole (AN) acylation with acetic anhydride (AA) at 100 °C for two hours using a ratio of AA to AN of 2:1 for fresh catalyst (a) parent (Z23) and desilicated samples and (b) regeneration cycles for desilicated Z23-at5 sample. Ref [79] / Elsevier.

27.3  Zeolites and Hierarchical Zeolites as Catalysts for Friedel Crafts Acylation Reactions

adsorption isotherms showed that the development of mesoporosity occurs when 0.05 mol L–1 NaOH solution is used (0.22 cm3 g–1 against 0.15 cm3 g–1 for the parent zeolite) and is not affected by the acid treatment nor, significantly, by the increase of the base concentration. On the other hand, a decrease of the microporous volume is noted as NaOH concentration increases, becoming more pronounced upon acid treatment, in accordance with some loss of crystallinity. The acidity characterization, quantified by the integration of FTIR spectra of pyridine chemisorbed on Brӧnsted and Lewis acid sites, showed an important decreased on Brӧnsted acidity for MCM-22/0.05 sample due to the partial occlusion caused by deposition of fragments at the pore apertures, which are removed upon acid treatment, giving an even higher concentration of Brönsted acid sites when compared with the parent material. In the case of MCM-22/0.1/AT sample, the use of a more concentrated NaOH solution caused a simultaneous extraction of Si and Al from the framework that is accentuated by the acid treatment. For the catalytic assays, furan, pyrrole, and AN were used as substrates and AA was used as acylating agent. The experiments were performed for about 1 h at 60 °C using a molar ratio AA/substrate of 5 and 150 mg catalyst. According to the reaction mechanism in the case of furan and AN, the reaction is almost selective toward isomers 2-acethylfuran and p-MOAP, respectively, because the α-product is preferred because of the greater stability of the α-intermediate structure. For pyrrole substrate, two isomers were obtained, which can be attributed to a higher electrophilic susceptibility of carbon-β relative to carbon-α [65]. To quantify the catalytic behavior of parent and hierarchical catalyst, by calculating the kinetic parameters for the reaction, the authors proposed a simplified equation for the LangmuirHinshelwood model [42, 46]. Considering the reaction scheme, represented as

A (ads) + S(ads) → P(ads)

where A stands for the acylating agent, S for the substrate and P for the acylated product(s), the reactions occur in the presence of a large excess of acylating agent, to avoid a premature deactivation of the zeolite by the acylated product. In addition, the secondary product, acetic acid, is so marginally adsorbed that it can be discarded. Considering these assumptions, the reaction rate, r, can be expressed upon simplification, as

r≅

k  A  S

2

( A  + S + Kr P )



(27.1)

where k represents the rate constant of the rate determining step and Kr represents the ratio between the adsorption equilibrium constant of the product(s) and the normalized equilibrium constant of the reagents. Figure 27.8 exemplifies the evolution of product yield as a function of reaction time for the acylation of furan by AA. As can be seen in this figure, a significant increase in product yield in the first minutes of reaction occurred, followed by a significant attenuation. The obtained results show the effect of desilication treatment where a decrease in product yield is observed as NaOH concentration increases from 0.05 to 0.1 mol L–1. On the other hand, the effect of HCl treatment led, in both cases, to an increase in product yield, which is especially relevant in the first minutes of reaction (see insert in Figure 27.8). Table 27.3 summarizes the kinetic parameters obtained from the non-linear regression treatment applied to the proposed Langmuir-Hinshelwood model equation as well as the turnover frequencies, calculated as the ratio between the kinetic constant and the concentration of accessible Bronsted acid sites [80]. From all three substrates, furan was selected as a representative example since it shows, for all samples, higher rate constant, and TOF.

591

592

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions

Figure 27.8  Product yield as a function of reaction time for the acylation of furan by acetic anhydride (AA). Lines represent calculated values resulting from the application of the kinetic model. Ref [46] / Elsevier.

Table 27.3  Summary of kinetics (rate constant, k, and turnover frequency, TOF), adsorption parameters, and statistical figures of merit for the catalytic reaction using furan as substrate [46]. Furan

MCM-22

MCM-22/0.05

MCM-22/0.05/AT

MCM-22/0.1

MCM-22/0.1/AT

k (mmol min–1 g–1)

14.3 ± 0.2

13.6 ± 0.3

178.8 ± 1.9

5.1 ± 0.1

23.8 ± 1.9

Kr

5.3 ± 0.3

8.8 ± 0.6

113.9 ± 4.9

5.8 ± 0.74

12.6 ± 2.5

0.992

0.989

0.999

0.955

0.924

2

R

sfit

0.044

0.056

0.187

0.029

0.311

F

1269

761

11239

171

97

TOF (min–1)

93.5

117.2

1027.6

48.6

171.2

2

R = determination coefficient; sfit= standard deviation of fit; F = Fisher-Snedecor statistics.

As depicted, desilicated samples show lower rate constants and TOFs when compared with the parent MCM-22, which can be ascribed to the accumulation of extra-framework material at the pore mouths of the zeolite, or even some loss of active sites, especially in the case of MCM-22/0.1 sample. For the acid treated samples, a significant increase in rate constant and TOF is observed. For example, for MCM-22/0.05/AT sample, k is about 12 times and Kr is more than 20 times higher when compared with the parent zeolite. For MCM-22/0.1/AT sample the increase in the rate constant is much more moderate, as well as that in Kr, due to the loss of active sites during the desilication treatment. Interestingly, in the case of AN, a more voluminous molecule, Kr value is even lower when compared with the parent MCM-22, which was attributed to the possible interconnection of the two internal pore systems that fastens product desorption.

27.3  Zeolites and Hierarchical Zeolites as Catalysts for Friedel Crafts Acylation Reactions

Aleixo et al. continued to study the effect of alkaline and alkaline+acid treatment on Friedel Crafts acylation reactions with small molecules as substrates, applying the same methodology, but now using BEA zeolite with distinct Si/Al ratios (12.5 and 32) [42]. In this case, the alkaline treatment was made with 0.1 mol L–1 NaOH solution at 60 °C for 30 minutes. The subsequent acid washing was performed using the experimental conditions previously mentioned [41, 46]. The catalytic experiments were performed at 60 °C, with a molar ratio AA/substrate of 5 and 150 mg catalyst, using furan, AN, and pyrrole as substrates and AA as an acylating agent. The characterization of parent and modified samples showed that the Si/Al ratio of the parent material strongly influenced the catalytic behavior of desilicated and desilicated+acid treated samples. In the case of the BEA12.5 (SI/Al = 12.5) sample, desilication caused a decrease in products’ yield and rate constant. Upon acid treatment, the unblocking of the zeolite porosity and the recovery of acid sites, that even slightly exceeds the parent material, lead to an increase in product yield and rate constant. On the other hand, for BEA32 (Si/Al = 32) a dissimilar behavior was noted since a continuous loss of acid sites occurred during the desilication and continued during the acid treatment. However, the desilicated+acid treated sample, BEA32-D-AT, presented the highest product yield in the case of furan and AN substrates, which can be assigned to the textural modifications, namely, the existence of larger mesopores that were developed in this sample. The effect of size and chemical functionalities of substrate molecules was further explored by Elvas-Leitão et al. [81] on BEA modified zeolites through desilication and desilication+acid treatments, modified according to the procedure reported in a previous study [42]. The authors performed the catalytic assays at 60 °C, using benzofuran, thiophene, benzothiophene, and indole and compared the catalytic performance with the previously studied smaller furan, AN, and pyrrole substrates [42]. The relative size of the studied substrate molecules can be compared using the corresponding van der Waals volumes, giving the following order: benzothiophene>benzofuran> indole>furan>tiophene>pyrrole. The kinetic curves and the parameters obtained using the simplified Langmuir-Hinshelwood model, proposed before [42, 46], show a clear relation between the size of substate molecules and their reactivity order, being in all cases significantly higher for tiophene and furan and

80

% Acetylated Product Yield

70 60

Furan Benzofuran Pyrrole Indole Thiophene Benzothiophen Anisole

50 40 30 20 10 0

0

10

20

30 Time (min)

40

50

60

Figure 27.9  Product yield as function of reaction time using BEA-D sample catalyst [81] / MDPI / CC BY 4.0.

593

594

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions

substantially lower for the bulkier benzothiophene and benzofurane, as can be seen for the desilicated sample, BEA-D, shown as a representative example in Figure 27.9. When comparing the catalytic results for the three catalysts: parent, desilicated and desilicated+acid treated, the obtained results only show slight differences. In fact, the textural and chemical modifications performed on modified zeolite materials did not show in a first inspection major differences in the reactivity, neither in order or magnitude. This could only be rationalized through the mathematical analysis of the results, carried out by quantitative structure-property relationships (QSPR), as discussed later. The use of hierarchical zeolites prepared with a surfactant templated method was reported by Linares et al. [82] where cethyltrimethylammonium bromide (CTAB) surfactant, in the presence of NH4OH, was employed to modify USY zeolite used as heterogeneous catalyst in Friedel-Crafts alkylation of indole using alcohols as acylating agents for the synthesis of compounds of pharmaceutical interest. Recently, Martins et al. [47] modified Y zeolite using two surfactants with different sizes: CTAB and the less explored DTAB, with a slightly shorter carbon chain length, in the presence of NH4OH. The suspensions were heated at 150 °C under autogenous pressure, changing the duration of the treatments from 6 to 48 hours. The characterization of the modified materials showed that the duration of the treatment influences the final properties of the materials. For the samples treated with DTAB (HY_D series), the six hour treatment (HY_D6) showed a significant loss of crystallinity and a decrease in the microporous volume characteristic of the zeolite structure. When the duration of the treatment was extended to 12, 24, or 48 hours, a continued enlargement of the pores occurred, ascribed to the transformation of smaller into larger micropores, in addition to the development of small diameter mesopores, along with some progressive loss of crystallinity and acidity. The size of the surfactant molecule used in the treatments also affected the zeolite properties. For samples modified for the same duration of treatments (i.e. 12 hours), the zeolite modified with the larger surfactant molecule, CTAB, showed a higher volume of larger mesopores. However, a less crystalline material was obtained, with 77% degree of crystallinity for HY_C12 against 95 % for HY_D12, as well as lower concentration of Brӧnsted acid sites, with 0.392 mmol g–1 and 0.464 mmol g–1 for HY_C12 and HY_D12, respectively. The catalytic behavior of parent and surfactant templated samples was studied in the acylation of furane, using AA as acylating agent, at 60 °C, with a molar ratio AA/furane of 5. Figure 27.10 shows the product yield as a function of reaction time, considering that the reaction is almost completely selective into 2-acethylfuran. A close inspection of Figure 27.10, discloses a sharp increase in product yield, for all samples, in the first 10 min of reaction, followed by a slope attenuation. In the case of DTAB treated samples (Figure 27.10a), in the initial stage of the reaction all modified samples present higher yield, when compared with HY material. For reaction times longer than 20 minutes, HY_D48 sample behaves distinctly, since a plateau is observed whereas a small increase is still evident for the parent and the other treated samples, attributed to the occurrence of secondary reactions that occluded the sample micro+mesoporosity. Accordingly, the best catalytic performance was obtained for HY_D24 because larger treatment time produces larger mesopores that lead to deactivation phenomena. For the samples treated with CTAB, HY_C12 sample displays a significant improvement in product since the larger pores present in this sample fasten the reaction kinetics (rate constant and TOF are two times higher when compared with HY-D12 sample). However, when the treatment time increases to 24 hours, an inversion occurs, indicating the occurrence of secondary transformations that cause diffusional limitations. The results obtained in this study clearly show that a careful optimization is needed to tune the required amount and size of mesopores needed to improve the catalytic performance for each particular reaction.

27.3  Zeolites and Hierarchical Zeolites as Catalysts for Friedel Crafts Acylation Reactions

Figure 27.10  Product yield as a function of reaction time for parent HY and modified dodecyltrimethylammonium bromide (DTAB) (a) and cetyltrimethylammonium bromide (CTAB) (b) for a 6 to 48 hour treatment. Ref [47] / Elsevier.

Hierarchical zeolites where the mesoporosity is generated during the synthesis (bottom-up procedures) were also explored in Friedel-Crafts acylation reactions. Bohstrӧm and Holmberg [83], prepared hierarchical ZSM-5 (mesoporous HZSM-5) following the procedure previously described by Wang et al. [84], using poly(diallyldimethyammonium chloride) as a soft template for mesoporosity development. Mesoporous HZSM-5 showed the same Si/Al ratio as the parent material as well as the same diffraction patterns. The textural characterization revealed the maintenance of microporous volume but the mesoporosity volume almost doubled from 0.07 to 0.12 cm3 g–1. The catalytic performance of micro and mesoporous HZSM-5 sample was studied in the acylation of 2-methylindole with AA, using a 2:1 molar ratio of AA:2-methylindole, at 85, 100 and 115 °C during 24 hours. The reaction over the microporous zeolite was regiospecific, giving acetylation only in the 3-position of 2-methylindole, whereas in the case of mesoporous sample a mixture of two isomers, in a 3:2 ratio, was obtained. A kinetic model, based on the Eley-Rideal reaction mechanism, allowed the determination of rate constants adsorption energy and activation energy of the reaction. In the case of microporous catalyst higher rate constant and lower activation energy was found, as compared with the mesoporous sample, but the mesoporous HZSM-5 gave lower adsorption energy and a higher yield at the highest temperature studied (89 % against 65 % for the microporous catalyst). Thus, the catalyst with higher content in mesopores has an improved resistance to deactivation and poisoning, thus overcoming what is considered as a main disadvantage in the application of zeolite materials in liquid-phase reactions. The effects of mesopore generation in ZSM-5 zeolite, applied in liquid-phase acylation of bulky aromatic compounds was explored by Kim et al. [85]. The authors prepared MFI zeolite nanosponge, obtained through a seed-assistant hydrothermal synthesis using C22H45–N+(CH3)2–C6H12–N+(CH3)2– C6H13. The zeolite nanosponge showed a disordered network of 2.5 nm thick zeolite layers, with very large external surface (460 m2 g–1 for MFI nanosponge against 60 m2 g–1 for conventional MFI zeolite), and a large volume of mesopores (0.5 cm3 g–1), with a narrow distribution of diameters centered at about 4 nm, which is comparable to mesoporous MCM-41 that was used as reference material (mesopore volume of 0.6 cm3 g–1). In addition, MFI nanosponge contains a high amount of strong Brӧnsted acid sites located at the external surface, 53 μmol g–1, which largely exceeds what is usually observed for commercial bulk zeolites (6 and 7 μmol g–1 for bulk BEA and MFI, respectively).

595

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions (b)

100

MFI nanosponge bulk beta Al-MCM-41 bulk MFI

80

100

MFI nanosponge bulk beta Al-MCM-41 bulk MFI

80 Anisole conversion (%)

(a)

1-Methoxynaphthalene conversion (%)

596

60

40

60

40

20

20

0 0

5 10 15 Reaction time (h)

20

0

5 10 15 Reaction time (h)

20

Figure 27.11  Conversion of 1-methoxinaphtalene (a) and anisole (AN) (b) as a function of reaction time. Reaction conditions for 1-methoxynapthalene (1-MM): 1 mmol of 1-methoxinaphalene, 2 mmol of acetic anhydride (AA), 4 mL of nitrobenzene, 50 mg of catalyst, T = 120 °C; for AN: 4 mmol AN, 8 mmol AA, 4 mol of nitrobenzene, 50 mg of catalyst, T = 120 °C. Ref [85] / Elsevier.

The catalytic behavior of MFI nanosponge was compared with other porous catalysts: bulk (conventional) MFI and BEA zeolites and mesoporous Al-MCM-41. The catalytic performance was studied using a series of substrate molecules, namely AN, methoxynaphthalene, ­naphthalene, and dimethylbenzene, in the presence of AA as the acylating agent and nitrobenzene as the solvent. The experiments were performed at temperatures ranging from 120 to 160 °C, depending on the substrate used. Figure 27.11 shows the catalytic conversion for 1-methoxynaphtalene (1-MM) and AN. As shown, the catalytic performance of MFI nanosponge stands out, especially in the case of the bulkier substrate 1-MM. However, it must be emphasized that the high mesoporous volume cannot be the only parameter to explain the superior performance of MFI nanosponge since it is comparable with Al-MCM-41 and this catalyst reaches only a maximum conversion of about 10%. The role of the catalytically active Brönsted acid sites located at the external surface can have a determinant role since in this case the diffusional limitations are absent. In the case of a less voluminous molecule, AN, this effect is less pronounced and when using bulk BEA, a large pore zeolite with 12 membered oxygen rings (12-MR), a moderate conversion (around 10%) is observed since AN may access the active sites located inside the pores, which is not observed for medium pore (10-MR) bulk MFI. This study opens new perspectives to the development of new methodologies to produce hierarchical zeolites with high external surface and rich acidity. In fact, the active sites located at the external surface are often neglected due to shape selective effects that are highlighted in the case of zeolites but are useless when voluminous molecules are used as reagents, due to diffusional constraints. Thus, these new methods of synthesis/modifications are an elegant way to overcome this limitation often referred by Eric Derouane in liquid phase reactions using zeolites as heterogeneous catalysts [63, 64].

27.4  Understanding Friedel-Crafts Acylation Reactions through Quantitative Structure-Property Relationships

27.4  Understanding Friedel-Crafts Acylation Reactions through Quantitative Structure-Property Relationships Quantitative structure-property/activity relationships (QSPR/QSAR) include, in general terms, all statistical and mathematical methods by which different types of properties can be related with the structural features of a series of compounds. This relation can be expressed as follows:

P = f ( D1 , D2 , D3 … Di )

(27.2)

where P (or a function of P, such as ln (P) or 1/P) is the property of interest (e.g. biological activity, toxicity, rate or equilibrium constants, solubility, chromatographic retention coefficients, etc.) and Di represents the molecular descriptors that encode the compounds 1D, 2D or 3D structural characteristics. The mathematical relationship between the property (P) and the descriptors (Di), the so-called model equation, can be either linear or nonlinear. Molecular descriptors are experimentally determined quantities, either macroscopic or microscopic (in the sense of probing a given interaction at a molecular level), or, instead, theoretically derived quantities. There are thousands of molecular descriptors referenced in the literature which differ in dimensionality and nature [86]. Although the selection of descriptors is a difficult task, the following criteria should be used: relevancy, absence of redundancy, number (they should be as few as possible to avoid chance correlations), and interpretability. QSPR/QSAR are these days one of the most important methodologies for modeling and prediction purposes in various areas of knowledge from fundamental physical chemistry to medicinal chemistry, environmental and material sciences and nanotechnologies [87–99]. QSPR/QSAR methods can be essentially divided in two large groups: regression and classification methods, being the former used to find, as referred, quantitative correlations between chemical, biological, toxicological information and structure, and the latter to visualize similarity and clustering within the compounds’ space [100]. Among regression methods, multiple linear regression analysis (MLR), principal components regression (PCR), partial least squares (PLS), and artificial neural networks (ANN) have been used the most [92, 93, 100–103], with MLR occupying a prominent place in certain areas. On the side of classification methods, linear discriminant analysis (LDA), principal component analysis (PCA), decision trees (DT), random forests (RF), k-nearest neighbor (kNN), logistic regression (LR), Kohonen self-organizing maps (SOM), and support vector machines (SVM, which can be used either for classification or for regression purposes), are the most frequently cited [92, 100, 101]. All of these methodologies are considered traditional machine learning (ML) algorithms. ML is an important sub-group of artificial intelligence (AI) technology [104, 105]. Deep learning (DL) approaches (i.e. ANN with several hidden processing layers) are a more recent development in AI technology and have lately gathered special attention due to their potential ability to describe nonlinear input-output relationships and for having a more powerful generalization capability. DL techniques are considered an interesting complement to conventional ML strategies which are rather dependent on the use of human-crafted molecular descriptors [104, 105]. In the context of this chapter, we will, however, only address traditional ML algorithms, whose advantages and limitations have been thoroughly reviewed elsewhere [100–102, 104, 105]. Independently of the process and the methodology chosen, QSPR/QSAR studies always have one of two purposes (or both): either the identification of the key factors that determine a certain response (interpretative ability) or the prediction of a system’s behavior when convenient

597

598

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions

structural modifications are designed and introduced in a molecule or material, prior to any synthesis, microbiological assay, or experimental measurement (predictive ability). This dichotomy between interpretability and predictability, so well-illustrated in a recent paper by Fujita and Winkler [103], should always be taken into consideration when choosing a particular methodology. There is no method ideal for all situations. The choice depends on the nature of the process being modeled (and on the quality and amount of data) and on the objective of the study: if the main drive is to understand/interpret/rationalize a certain behavior, then the researcher should select a linear method such as MLR or PLS; if, on the contrary, the researcher is mainly interested to accurately predict a given response (property), then a non-linear method like ANN appears as a more convenient alternative (Figure 27.12). In case one wants to predict a compound’s class (e.g. active vs. inactive) and the activity information is scarce, classification methods are the most suitable. On the other hand, if there is enough property information (e.g. IC50 or k values) for a series of congeneric compounds for which it is possible to derive or experimentally obtain a set of molecular descriptors, then a regression method is the best option for quantitative prediction [100, 101]. QSPR have now been used for almost a century [106] to relate rate constants (k) of homogeneous kinetics in the liquid phase with macroscopic solvent properties such as dielectric constant, dipole moment, pH, refractive index, and other properties, or/and with microscopic solvent properties which are based in chemical probes capable of accounting for the possible role of Lewis acidity, Lewis basicity, dipolarity, polarizability, and other factors [107, 108]. Several model equations have been derived that allow a better understanding of the reaction mechanisms and/or the prediction of new k values for missing experimental values [109–111]. In the same way, equilibrium constants (K), solution enthalpies (∆sol H ), and many other physicochemical properties have also been correlated with different sets of parameters usually related to the same type of properties [112–114].

Figure 27.12  Trade-off between interpretability and predictability for the methods referred in the text. The markers’ size gives an indication of how much time is consumed by each method. Adapted from Ref [100].

27.4  Understanding Friedel-Crafts Acylation Reactions through Quantitative Structure-Property Relationships

Over time, different approaches have been proposed, the most notable one being the Abraham linear free energy relationships (LFER) model, that uses solute properties as independent variables to predict several important physicochemical and biological properties [115, 116]. Some of these QSPR models have also been applied to the study of catalytic effects, although, to the best of our knowledge, the number of studies resorting to QSPR analysis in the field of catalysis is still not very expressive [117–130] and this number is even more incipient when zeolites are used as catalysts [42, 81, 131–133]. The application of a QSPR strategy is, however, of undeniable interest when the experimental assessment of catalytic behavior is expensive, time-consuming or experimentally difficult [124], or when the focus is on the assessment (and rationalization) of the simultaneous effect of various factors [128, 131, 133]. With this in mind, a couple of years ago, we started a systematic study of the catalytic behavior of hierarchical MCM-22 and BEA zeolites in Friedel-Crafts acylation reactions carried out in mild conditions using, as already mentioned, AA as an acylating agent and various heteroaromatic rings of different sizes and chemical functionalities as substrates (namely, furan, benzofuran, thiophene, benzothiophene, pyrrole, indole, and AN) [42, 46, 81]. The idea was to probe the combined effect of changes in both substrate and catalyst (zeolite) properties over rate constants, k, and relative sorption equilibrium constants, Kr, through the use of a MLR-based QSPR strategy. The use of this approach assumes that k and Kr (or better ln k and ln Kr) can be expressed as linear combinations of independent and additive properties or descriptors, D, that in this case encode the structural features of both substrates and catalysts:

N

ln k (or ln K r ) = a0 + ∑ i=1 ai Di + x

(27.3)

In Eq. 27.3 x stands for the regression residuals, a0 is the regression independent term and ai the regression coefficients associated with each descriptor, Di, obtained by minimizing the sum of the squared residuals. The order of magnitude of the regression coefficients of the normalized descriptors reflects their relative importance, whereas the corresponding signs can be interpreted in terms of positive or negative contributions to the response variable. Experimental data (i.e. product yield vs. time) for the referred systems were first fitted to Eq. 27.1 (vd. 27.3) using a nonlinear regression method which allows the estimation of k and Kr values and of the corresponding coefficient parameters. Additionally, a specific software was used to calculate the parameter’s uncertainties. Before performing any MLR-QSPR like those we carried out in our studies, the following general aspects are always verified: the number of parameters (descriptors) should not exceed the number of data points to avoid overfitting and thus the possibility of chance correlations (as a rule of thumb, there should always be at least 3–5 points per each parameter); intercorrelations among descriptors should be checked, so that pairs of parameters for which the determination coefficient, r2, of the correlation matrix exceeds the cut-off value of 0.5 are excluded, thus preventing the presence of more than one parameter describing the same type of interaction in the same model equation (redundancy). If enough data points are available, the initial set should additionally be divided in two independent sub-sets, one to establish the model equation, the so-called training set, and the other, the test set or prediction set, to test the derived model, allowing therefore an external validation procedure to assess the model’s predictive ability. Test and training sets should have similar sizes and should be representative of the whole chemical/structural space covered by the descriptors and also of the dependent variable’s space (same variability); test set points should not be beyond training set points to avoid predictions outside the model’s applicability domain. Regressions can be carried out either by a forward- or a backward-stepwise procedure, achieved by either starting with equations containing a single descriptor and adding up

599

600

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions

terms, one at a time, or the other way around, starting with equations with the “maximum” number of descriptors and then remove one by one. Descriptors are added, retained or disregarded in each step according to rigorous statistical criteria: parameters to be considered need to have a significance level (SL)>95%, and the adjusted multiple determination coefficient, R2adj, of each regression should be high and as close to 1 as possible. Also, models have to be examined for spurious points (outliers). As a rule of thumb, calculated values are kept if they are not more than 2 standard deviations (of the fit), sfit, from the corresponding experimental values and also if they are not leverage points (i.e. points that tend to unbalance the regression and whose presence must therefore be prevented) [88, 134]. External validation procedures should be used whenever one seeks to evaluate the predictive capability of a model. External validation, as well as internal validation (model’s robustness), must comply to stringent metrics which have been duly explained elsewhere [88, 93, 134–137]. Also, visual inspection of scatter plots of predicted vs. experimental data is strongly recommended as a complimentary indicator of good predictive power [137]. In our work, a set of descriptors frequently used to qualitatively analyze results in heterogeneous catalysis studies were used. These comprised descriptors associated with the zeolite properties, namely external area (Aext), micropore volume (Vmicro), mesopore volume (Vmeso), Brønsted acidity ([B]pyr), Lewis acidity ([L]pyr), silicon/aluminium ratio (Si/Al) and crystallinity (CXRD), and others linked to the substrate properties: molecular diameter (MDiam), van der Waals volume (VvdW), dipolarity, and Lewis acidity (as given by the normalized Dimroth and Reichardt’s parameter, ETN). A stepping forward regression procedure applied to ln (k or Kr) vs. descriptors in the case of the Friedel-Crafts acylations studies over hierarchical MCM-22 zeolites [46], showed that the Lewis acidity descriptor, [L]pyr, needed not be considered in any of the final models since its statistical relevance was much lower than that of [B]pyr when both uniparametric equations were compared. The observed high intercorrelation between [B]pyr and [L]pyr, would nevertheless eliminate the possibility of the simultaneous use of two descriptors in the same fitting equation. In the case of Kr, the complete regression procedure led for this zeolite to four equations with three descriptors and identical R2adj and sfit. From these, Eq. 27.4 was selected because its physicochemical interpretation was more straightforward and there was no real statistical justification to select one equation in detriment of any of the others.



ln K r = (−3.84 ± 0.08) + (0.24 ± 0.10) B pyr + (0.89 ± 0.11) VvdW + (1.30 ± 0.10) ETN

(100%)

(100%)

(100%)

(95%)

(27.4)

N = 10; R2 = 0.972; Ra2dj = 0.958; s fit = 0.25; F = 70 (in this fitting, two data points were identified as outliers and were excluded, and the model refitted). For the rate constant k, the best-found equation was Eq. 27.5.



ln k = (1.57 ± 0.20) + (0.87 ± 0.21) B pyr − (2.00 ± 0.23) VvdW − (0.47 ± 0.23) ETN

(100%)

(100%) 2

N = 12; R = 0.914;

(100%)

Ra2dj

(93%)



(27.5)

= 0.882; s fit = 0.69; F = 29

In these equations, N is the number of points, sfit the standard deviation of the fit, and F the Fisher–Snedecor statistics (the numbers in brackets underneath each equation term are the significance level of each coefficient).

27.4  Understanding Friedel-Crafts Acylation Reactions through Quantitative Structure-Property Relationships

The mesoporous volume was removed in both cases given its low significance level. This was not surprising because diffusion is not the rate limiting step and product adsorption must be nearly insensitive to this property, in agreement with the conclusions of a Friedel-Crafts alkylation study involving Beta zeolites that showed that mesoporous volume is only relevant for large substrates [138]. A close inspection of Eq. 27.4 shows that all descriptors contribute positively for Kr, being ETN the most important and [B]pyr, the less relevant. For k, Eq. 27.5 reveals that [B]pyr increases k while the size of the substrate and its dipolarity/Lewis acidity, as described by ETN, decreases it. This is again in line with Ref [138], where a good relationship between catalytic activity and [B]pyr was also recognized. The positive sign of the ETN in Eq. 27.4 shows that substrate dipolarity favors product adsorption relative to that of the reagent and, hence, active sites remain occupied by product molecules and make it more difficult for further reagent molecules to adsorb (i.e. contributing to lower k, as evidenced by the negative sign of this coefficient in Eq. 27.5). The same tendency is seen for the VvdW volume, which is thermodynamically favorable (increases Kr) and kinetically unfavorable (decreases k). The observed relative order of descriptors’ importance for Kr is: ETN > VvdW > [B]pyr and for k is VvdW > [B]pyr > ETN. Notably, VvdW and [B]pyr influence rate at least twice as much as they influence adsorption. On the contrary, the effect of the substrates’ polarity coefficient, measured by ETN, is three times higher in Kr than in k. The interesting aspect of this QSPR analysis is that it complements and corroborates the catalytic qualitative study already mentioned. Using a similar procedure, a second work describing results obtained for the same set of substrates but using a mesoporous BEA zeolite was published by the same authors [42]. In this case, ln k and ln Kr were correlated to a set of descriptors, in which seven related to zeolite properties (Aext, Vmicro, Vmeso, [B]pyr, [L]pyr, Si/Al and CXRD) and five to substrate properties (VvdW; ETN; f(ε), a function of the dielectric constant; g(nD), a function of the refractive index; and μ, the dipole moment). As in the previous study, the data set comprised 12 points limiting the maximum number of variables in the fitted equations to four. In this case, both final model equations (Eqs. 27.6 and 27.7) contained four descriptors as follows: l n k = (0.96 ± 0.03) −(0.28 ± 0.05)C XRD + (0.72 ± 0.08) Vmicro −(0.53 ± 0.05) B pyr − (0.61 ± 0.02) f (ε)               (100%)            (99.9%)                   (100%)                       (100%)                      (100%) N = 12; R2 = 0.992; R2adj = 0.986; s fit = 0.04; F = 195

(27.6)

and, K t = (0.22 ± .0.03) + (0.06 ± 0.03)Si/Al − (0.24 ± 0.03) Vmicro + (0.83 ± 0.03) Vvdw + (0.44 ± 0.03) ET N ln                  (100%)

(93%)                   2

(100%)        2

(100%)          

N = 12; R = 0.990; Radj = 0.984; s fit = 0.04; F = 165      

          (100%)   

(27.7)

Variable inflation factors (VIFs) were also computed and were all below seven, indicating that multicollinearity was not biasing the estimates of coefficients. Figure 27.13a shows a good correlation between predicted and observed ln k values. Open squares correspond to outliers, removed from the data set sequentially and the model refitted after each removal. Comparing results with those obtained for MCM-22, there is an important difference between these two zeolites: the inclusion, for BEA, of the degree of crystallinity. CXRD has a negative influence over ln k, indicating that the loss of crystallinity resulting from post-treatment

601

602

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions

modifications of BEA (selective extraction of Si or, in some cases, of Si and Al) accelerates the reaction. Vmicro has the most significant positive contribution to ln k which reinforces the evidence from the previous work that these reactions occur mainly inside micropores and the enhanced diffusion due to mesoporosity is not relevant to the reaction. Also, [B]pyr exhibits a surprisingly negative coefficient probably meaning that enhancing this type of acidity as compared to the departing structure is detrimental to the catalyst. This is very clear in the case of furan reacting within zeolite BEA-32-AT, for which a decrease of over 50% of [B]pyr leads to a considerable increase in k, compared to parent and desilicated samples. This suggests that even for the sample with the lowest Brønsted acidity there are enough acid sites to catalyze the reaction. Therefore, a higher amount of acid sites probably promotes secondary reactions and adsorption, followed by deactivation processes. Finally, f(ε), the substrates’ dipolarity, also decreases the rate constant, implying very strong adsorption of product molecules leading to the occupancy of active sites and thus preventing reagent adsorption. For Kr, Figure 27.13b displays a good correlation between predicted (by model Eq. (27.7)) and observed ln Kr values. In this case, three outliers (open squares) were detected and sequentially removed. Because VIF values were approximately unit, coefficients and their standard deviations are unbiased. As for MCM-22, ETN and VvdW also contribute positively to ln Kr, being VvdW the most important descriptor and reflecting the effect of product bulky molecules which diffuse slowly and become more adsorbed. Vmicro is also relevant, but contributes to the lowering of ln Kr whereas the Si/Al ratio has a very small positive contribution. The negative influence of Vmicro relates to the formation of mesopores from the destruction/enlargement of micropores. The occurrence of secondary reactions/adsorption phenomena mostly inside mesopores promotes the formation of bulky products that become imprisoned inside the mesopores and will not be able to diffuse. The positive sign of the ETN coefficient, means that the combination of the substrate’s HBA and dipolarity stimulates product adsorption relative to that of the substrate. The obtained equations allowed, once again, the establishment of relative orders of importance. For Kr: VvdW (+) > ETN(+) > Vmicro (−) ≫ Si/Al(+); for k: Vmicro (+) > f (ε)(−) > [B]pyr(−) > CXRD (−). With BEA zeolite, the QSPR strategy was again consistent with the characterization analysis and the catalytic results. The first two QSPR analysis were performed with small substrate molecules with low interest as APIs. In order to ensure a higher variability within each descriptor, the number of substrates was increased in the last of the three works given as examples [81], by adding

Figure 27.13  Plots of (a) ln kpred vs. ln kexp and (b) ln Kr pred vs. ln Kr exp.

27.4  Understanding Friedel-Crafts Acylation Reactions through Quantitative Structure-Property Relationships

molecules with different chemical functionalities and larger sizes. Qualitatively, the data only showed that the rate constants of thiophene and furan were significantly higher than those for the other substrates. Also, part of the descriptors previously used referred to the substrate molecules as if they were solvents in the process. In the third of these works, only descriptors pertaining to the substrates as such were included, this time aiming to account for a higher diversity and complexity of substrate molecules and to consider them as such and not as solvents. The developed models were based on the correlation of ln k and of ln Kr, with seven descriptors related to the zeolite properties (the same ones considered previously) and seven substrate properties, namely: surface area (σ), van der Waals volume (VvdW), density (ρ), molar volume (Vm), dipole moment (μ), viscosity (η), and surface tension (γ). Given the larger data set, and since some outliers were expected, fitted models went up to a maximum of five descriptors. MLR results were filtered according to the same sequence of criteria applied earlier, plus: (i) only equations for which R2adj ≥ 0.95 were selected; (ii) only model equations with two or less outliers were chosen; (iii) only model equations with the top 50% Fisher-Snedecor F-values were kept. This procedure led, both for ln k and ln Kr, to several, virtually statistically identical models, but it was possible to determine in both cases the preferred one using also as criteria, the standard deviation of the fit, the adjusted determination coefficient and the predicted determination coefficient, R2pred, to check the predictability of each model equation. R2pred = 1 − PRESS/SST, where n

n 2 PRESS = ∑ i=1 ( yi − y (i))2, SST = ∑ i=1 ( yi − y ) , and ŷ(i) are the predicted values of the response

of the ith observation using a model whose estimate is based on (n−1) data points excluding the ith data point. If R2adj − R2pred < 0.2, the model was considered to show good predictability. Solutions for which coefficients lacked physical meaning were also excluded. For ln k and ln Kr, the final equations were the following:

ln k =(1.06 ± 0.07) −(0.76 ± 0.05)VvdW −(0.52 ± 0.07)ρ +(0.14 ± 0.04) B pyr    

(100%)            (100%)                   1 (100%)                (100%)    

(27.8)

N = 21; R2 = 0.928; R2adj = 0.915; R2 pred = 0.907; s fit = 0.08; F = 73            and,



ln K r = (0.31 ± 0.02) + (0.60 ± 0.04)σ + (0.45 ± 0.04) r − (0.06 ± 0.02) Vmicro           

(100%)           (100%)                (100%)                (99%)      

(27.9)

N = 16; R2 = 0.963; R2adj = 0.953; R2 pred = 0.935; s fit = 0.08; F = 103           Because R2pred was quite close to R2adj, it was concluded that the models had good predictive ability within the covered variable’s space. Also, residuals were randomly distributed around the trendline, as seen in Figure 27.14a. The relative importance of the descriptors for k is this time VvdW(−)>ρ(−)>[B]pyr (+). In line with previous works, [B]pyr is important and increases k, although it is the least contributing descriptor, possibly because a greater difference of Brønsted acidity among zeolites is needed to further reveal its importance. The most important descriptor, VvdW, which accounts for the volume occupied by substrate molecules, is unfavorable to k, showing that the size of the substrate determines the ease of reagents diffusion inside zeolite pores. Furthermore, the weight of this descriptor suggests that a large electron cloud establishes weaker bonding with the catalytic surface. On the other hand, substrate density, ρ, diminishes k, since a larger density causes steric hindrance.

603

604

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions

Figure 27.14  Plots of (a) ln kpred vs. ln kexp and (b) ln Kr pred vs. ln Kr exp (� = outlier).

For Kr, again only three descriptors are needed to model the adsorption/desorption behavior, and the best equation now includes σ, the substrate’s surface area accessible to the zeolite, ρ, the substrate’s density, and again, only one zeolite parameter, Vmicro. The relative importance of descriptors is: σ (+) > ρ (+) > Vmicro (−) and, thus, the reagent’s accessible surface area and the density increase Kr, since desorption becomes more difficult due to stronger adsorption. Larger microporous volumes facilitate desorption relative to adsorption and, so, decreases Kr. Surprisingly, Lewis acidity, [L]pyr, has shown no significant role in the processes, although it spanned for 31% against only 14% for Brønsted acidity, [B]pyr. Since [B]pyr is strongly correlated with Vmicro, and σ is even more strongly correlated with VvdW, results show a somewhat inverse influence of similar parameters in Kr and in k: the same, or similar, parameters which diminish the rate constant also increase the difficulty in product desorption (larger Kr). This can be rationalized in terms of stronger adsorption forces that at the same time increase catalytic efficiency but also make active sites less accessible to new molecules. Figure 27.14 shows the fitted values for k and Kr. In summary, the results of these three studies show that the most relevant parameter in both k and Kr is related to the substrate molecular size. However, these parameters may be associated with different reaction steps, namely accessibility to micropores, diffusion capacity of molecules through the zeolitic atomic network and molecular polarizability, factors that were already shown to be relevant in the previous works using fewer and smaller model molecules. It would be very interesting to better assess the predictive ability of the models. This could be achieved by estimating k and Kr for a different set of substrates not used in these correlations, for which experimental values obtained in the same conditions (i.e. same zeolite, acylating agent and temperature) would exist, but up to date there are no reports of such values in the literature. The used multivariate QSPR approach, by taking into consideration, simultaneously, multiple changes in relevant properties of substrates and zeolites, produced in these three cases results that are consistent with data obtained from the structural characterization and from the catalytic tests. This strategy allows the identification of key properties that can be optimized through synthesis or post-synthesis treatments to design catalysts with improved catalytic performance.

27.5  Final Remarks The present work overviews recent developments of Friedel-Crafts acylation reactions using zeolitebased catalysts. These chemical transformations play an important role in the multi-step synthesis

References

Although BEA zeolite structure is already being used at an industrial level as a Friedel-Crafts heterogeneous catalyst, the large majority of the published works still focus on the use for these reactions of Lewis or strong protonic acids as homogeneous catalysts, in spite of the economic and environmental drawbacks associated with them. Zeolites have been extensively tested as acid catalysts in Friedel-Crafts acylations involving small substrates. To enlarge the applicability of zeolite-based catalysts to acylation of bulkier compounds and avoid severe deactivation phenomena en route, studies based on hierarchical zeolite structures but using larger substrates have been carried out with promising results. Additionally, to rationalize the combined effect of substrate and zeolite properties upon catalytic behavior, a MLR-based QSPR strategy was applied. This approach allowed the identification of key properties that could be used to improve catalytic performance. The search for better performing catalysts is indeed an important topic that continues to draw the attention of the scientific community. In summary, the data already available in the literature shows that hierarchical zeolite derived solids can be used as catalysts in Friedel-Crafts reactions, with clear environmental advantages over traditional homogeneous catalysts. Furthermore, the application of machine learning techniques provides useful information for a better catalyst design.

Acknowledgements This research was funded by Fundação para a Ciência e a Tecnologia (FCT) through UIDB/ 00100/2020. UIDP/00100/2020 and LA/P/0056/2020 and Instituto Politécnico de Lisboa (IPL) through Project IPL/2021/ZeoGemini ISEL.

References 1 Friedel, C. and Crafts, J.M. (1877). C. R. Acad. Sci. 84: 1392–1395. 2 Friedel, C. and Crafts, J.M. (1877). C. R. Acad. Sci. 84: 1450–1454. 3 Olah, G.A. (1963). Friedel-Crafts and Related Reactions. Geneva: Interscience Publishers. 4 Bandini, M. and Umani-Ronchi, A. (2009). Catalytic Asymmetric Friedel-Crafts Alkylations. Weinheim: Wiley-VCH. 5 Franck, H.-G. and Stadelhofer, J.W. (1988). Industrial Aromatic Chemistry, 1e. Berlin: Springer Berlin Heidelberg. 6 Röper, M., Gehrer, E., Narbeshuber, T., and Siegel, W. (2000). Acylation and alkylation. In: Ullmann’s Encyclopedia of Industrial Chemistry (ed. C. Ley), 337–382. Hoboken: John Wiley and Sons, Ltd. 7 Cavani, F., Centi, G., Perathoner, S., and Trifiró, F. (eds.) (2009). Sustainable Industrial Process. Principals, Tools and Industrial Examples. Weinheim: Wiley-VCH. 8 Short, W.F., Stormberg, H., and Wiles, A.E. (1936). J. Chem. Soc. 319–322. 9 Li, W., Jin, H., Yang, S. et al. (2019). Green Process. Synth. 8: 474–479. 10 Kusumaningsih, T., Prasetyo, W.E., and Firdaus, M. (2020). RSC Adv. 10: 31824–31837. 11 Earle, M.J., Seddon, K.R., Adams, C.J., and Roberts, G. (1998). Chem. Commun. 2097–2098. 12 Boon, J.A., Levisky, J.A., Pflug, J.L., and Wilkes, J.S. (2002). J. Org. Chem. 51: 480–483. 13 Ross, J. and Xiao, J. (2002). Green Chem. 4: 129–133. 14 Barrer, R.M. (1981). Zeolites 1: 130–140. 15 Guisnet, M. and Ribeiro, F.R. (2006). Les zeolithes: un nanomonde au service de la catalyse. Les Ullis: EDP Science.

605

606

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions

16 Structure Commission of the International Zeolite Association. www.iza-structure.org (accessed 28 June 2023). 17 Figueiredo, J.L. and Ramôa Ribeiro, F. (2007). Catálise Heterogénea, Fundação Calouste Gulbenkian, 2e. Lisbon. 18 Wang, W., Chang-Jun, L., and Wu, W. (2019). Catal. Sci. Technol. 9: 4162–4187. 19 Schwieger, W., Machoke, A.G., Weissenberger, T. et al. (2016). Chem. Soc. Rev. 45: 3353–3376. 20 Barrer, R.M. and Makki, M.B. (1964). Can. J. Chem. 42: 1481–1487. 21 Triantafillidis, C.S., Vlessidis, A.G., and Evmiridis, N.P. (2000). Ind. Eng. Chem. Res. 39: 307–319. 22 Xue, H., Huang, X., Zhan, E. et al. (2013). Catal. Commun. 37: 75–79. 23 De Baerdemaeker, T., Yilmaz, B., Müller, U. et al. (2013). J. Catal. 308: 73–81. 24 Ivanov, D.P., Pirutko, L.V., and Panov, G.I. (2014). J. Catal. 311: 424–432. 25 Kawai, T. (2014). Colloid Polym. Sci. 292: 533–538. 26 Janssen, A.H., Koster, A.J., and De Jong, K.P. (2002). J. Phys. Chem. B 106: 11905–11909. 27 You, Q., Wang, X., Wu, Y. et al. (2021). New J. Chem. 45: 10303–10314. 28 Xin, H., Li, X., Fang, Y. et al. (2014). J. Catal. 312: 204–215. 29 Kubota, Y., Inagaki, S., and Takechi, K. (2014). Catal. Today 226: 109–116. 30 Kim, J., Choi, M., and Ryoo, R. (2010). J. Catal. 269: 219–228. 31 Mériaudeau, P., Tuan, V.A., Nghiem, V.T. et al. (1999). J. Catal. 185: 378–385. 32 Groen, J.C., Hamminga, G.M., Moulijn, J.A., and Pérez-Ramírez, J. (2007). Phys. Chem. Chem. Phys. 9: 4822–4830. 33 Paixão, V., Carvalho, A.P., Rocha, J. et al. (2010). Microporous Mesoporous Mater. 131: 350–357. 34 Paixão, V., Monteiro, R., Andrade, M. et al. (2011). Appl. Catal. A Gen. 402: 59–68. 35 Verboekend, D., Mitchell, S., Milina, M. et al. (2011). J. Phys. Chem. C 115: 14193–14203. 36 Pérez-Ramírez, J., Christensen, C.H., Egeblad, K. et al. (2008). Chem. Soc. Rev. 37: 2530–2542. 37 Groen, J.C., Moulijn, J.A., and Pérez-Ramírez, J. (2007). Ind. Eng. Chem. Res. 46: 4193–4201. 38 Groen, J.C., Peffer, L.A.A., Moulijn, J.A., and Pérez-Ramírez, J. (2004). Colloids Surfaces A Physicochem. Eng. Asp. 241: 53–58. 39 Groen, J.C., Jansen, J.C., Moulijn, J.A., and Pérez-Ramírez, J. (2004). J. Phys. Chem. B 108: 13062–13065. 40 Ogura, M., Shinomiya, S.Y., Tateno, J. et al. (2000). Chem. Lett. 8: 882–883. 41 Machado, V., Rocha, J., Carvalho, A.P., and Martins, A. (2012). Appl. Catal. A Gen. 445–446: 329–338. 42 Aleixo, R., Elvas-Leitão, R., Martins, F. et al. (2019). Mol. Catal. 476: 110495. 43 Andrade, M.A. and Martins, L.M.D.R.S. (2021). Molecules 26: 1680. 44 Mendoza-Castro, M.J., Serrano, E., Linares, N., and García-Martínez, J. (2021). Adv. Mater. Interfaces 8: 1–23. 45 García-Martínez, J., Johnson, M., Valla, J. et al. (2012). Catal. Sci. Technol. 2: 987–994. 46 Aleixo, R., Elvas-Leitão, R., Martins, F. et al. (2017). Mol. Catal. 434: 175–183. 47 Martins, A., Neves, V., Moutinho, J. et al. (2021). Microporous Mesoporous Mater. 323: 111167. 48 Van-Dúnem, V., Carvalho, A.P., Martins, L.M.D.R.S., and Martins, A. (2018). ChemCatChem 10: 4058–4066. 49 Al-Ani, A., Haslam, J.J.C., Mordvinova, N.E. et al. (2019). Nanoscale Adv. 1: 2029–2039. 50 Ottaviani, D., Van-Dúnem, V., Carvalho, A.P. et al. (2020). Catal. Today 348: 37–44. 51 García-Martínez, J., Johnson, M., Valla, J. et al. (2012). Catal. Sci. Technol. 2: 987–994. 52 Corma, A., Fornés, V., Guil, J.M. et al. (2000). Microporous Mesoporous Mater. 38: 301–309. 53 Madsen, C. and Jacobsen, C.J.H. (1999). Chem. Commun. 673–674. 54 Jacobsen, C.J.H., Madsen, C., Houzvicka, J. et al. (2000). J. Am. Chem. Soc. 122: 7116–7117. 55 Valtchev, V.P., Smaihi, M., Faust, A.C., and Vidal, L. (2004). Chem. Mater. 16: 1350–1355.

References

56 Zhu, H., Liu, Z., Wang, Y. et al. (2008). Chem. Mater. 20: 1134–1139. 57 Holland, B.T., Abrams, L., and Stein, A. (1999). J. Am. Chem. Soc. 121: 4308–4309. 58 Serrano, D.P., Aguado, J., Escola, J.M. et al. (2006). Chem. Mater. 18: 2462–2464. 59 Narayanan, S., Vijaya, J.J., Sivasanker, S. et al. (2015). J. Porous Mater. 22: 907–918. 60 Fukasawa, T., Otsuka, K., Murakami, T. et al. (2021). Colloid Interface Sci. Commun. 42: 100430. 61 Venuto, P.B. and Landis, P.S. (1968). In: Advances in Catalysis (ed. D.D. Eley, H. Pines, and P.B. Weisz), vol. 18, 259–371. New York: Academic Press. 62 Derouane, E.G. (1998). J. Mol. Catal. A Chem. 134: 29–45. 63 Derouane, E.G., Dillon, C.J., Bethell, D., and Derouane-Abd Hamid, S.B. (1999). J. Catal. 187: 209–218. 64 Derouane, E.G., Crehan, G., Dillon, C.J. et al. (2000). J. Catal. 194: 410–423. 65 Álvaro, V.F.D., Brigas, A.F., Derouane, E.G. et al. (2009). J. Mol. Catal. A Chem. 305: 100–103. 66 Clerici, M.G. (2000). Top. Catal. 13: 373–386. 67 Sartori, G. and Maggi, R. (2009). Advances in Friedel-crafts Acylation Reactions: Catalytic and Green Processes. Boca raton: CRC Press. 68 Liang, J., Liang, Z., Zou, R., and Zhao, Y. (2017). Adv. Mater. 29: 1–21. 69 Nayak, Y.N., Nayak, S., Nadaf, Y.F. et al. (2019). Lett. Org. Chem. 17: 491–506. 70 Procházková, D., Kurfiřtová, L., and Pavlatová, J. (2012). Catal. Today 179: 78–84. 71 Yamazaki, T., Makihara, M., and Komura, K. (2017). J. Mol. Catal. A. Chem. 426: 170–176. 72 Makihara, M., Aoki, H., and Komura, K. (2018). Catal. Letters 148: 2974–2979. 73 Wei, H., Liu, K., Xie, S. et al. (2013). J. Catal. 307: 103–110. 74 Chen, Z., Feng, Y., Tong, T., and Zeng, A. (2014). Appl. Catal. A, Gen. 482: 92–98. 75 Matsukata, M., Ogura, M., Osaki, T. et al. (1999). Top Catal. 9: 77–92. 76 Koehle, M., Zhang, Z., Goulas, K.A. et al. (2018). Appl. Catal. A Gen. 564: 90–101. 77 Chandra Shekara, B.M., Jai Prakash, B.S., and Bhat, Y.S. (2012). J. Catal. 290: 101–107. 78 Derouane, E.G., Schmidt, I., Lachas, H., and Christensen, C.J.H. (2004). Catal. Letters 95: 13–17. 79 Silva, D.S.A., Castelblanco, W.N., Piva, D.H. et al. (2020). Mol. Catal. 492: 111026. 80 Guidotti, M., Coustard, J.M., Magnoux, P., and Guisnet, M. (2007). Pure Appl. Chem. 79: 1833–1838. 81 Elvas-Leitão, R., Martins, F., Borbinha, L. et al. (2020). Molecules 25: 5682. 82 Linares, N., Cirujano, F.G., De Vos, D.E., and García-Martínez, J. (2019). Chem. Commun. 55: 12869–12872. 83 Bohström, Z. and Holmberg, K. (2013). J. Mol. Catal. A Chem. 366: 64–73. 84 Wang, L., Zhang, Z., Yin, C. et al. (2010). Microporous Mesoporous Mater. 131: 58–67. 85 Kim, J.C., Cho, K., Lee, S., and Ryoo, R. (2015). Catal. Today 243: 103–108. 86 Todeschini, R. and Consonni, V. (2010). Molecular Descriptors for Chemoinformatics, Vol. 2. Hoboken: Wiley Blackwell. 87 Madzhidov, T.I., Rakhimbekova, A., Afonina, V.A. et al. (2021). Mendeleev Commun. 31: 769–780. 88 Martins, F., Santos, S., Ventura, C. et al. (2014). Eur. J. Med. Chem. 81: 119–138. 89 Winkler, D.A. (2016). Toxicol. Appl. Pharmacol. 299: 96–100. 90 Santana, R., Onieva, E., Zuluaga, R. et al. (2021). Curr. Top. Med. Chem. 21: 828–838. 91 Salahinejad, M. (2015). Curr. Top. Med. Chem. 15: 1868–1886. 92 Berhanu, W.M., Pillai, G.G., Oliferenko, A.A., and Katritzky, A.R. (2012). Chempluschem 77: 507–517. 93 Cherkasov, A., Muratov, E.N., Fourches, D. et al. (2014). J. Med. Chem. 57: 4977–5010. 94 Quintero, F.A., Patel, S.J., Muñoz, F., and Sam Mannan, M. (2012). Ind. Eng. Chem. Res. 51: 16101–16115. 95 Awfa, D., Ateia, M., Mendoza, D., and Yoshimura, C. (2021). ACS ES&T Water 1: 498–517. 96 Villaverde, J.J., Sevilla-Morán, B., López-Goti, C. et al. (2018). Sci. Total Environ. 634: 1530–1539.

607

608

27  Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions

97 Roy, J., Ghosh, S., Ojha, P.K., and Roy, K. (2019). Environ. Sci. Nano 6: 224–247. 98 Sakaguchi, K., Matsui, M., and Mizukami, F. (2005). Appl. Microbiol. Biotechnol. 67: 306–311. 99 Meier, M.A.R. and Barner-Kowollik, C. (2019). Adv. Mater. 31: 1806027. 100 Martins, F., Ventura, C., Santos, S., and Viveiros, M. (2014). Curr. Pharm. Des. 20: 4427–4454. 101 Yee, L.C. and Wei, Y.C. (2012). Current Modeling Methods Used in QSAR/QSPR. In: Statistical Modelling of Molecular Descriptors in QSAR/QSPR, Vol. 2 (eds. M. Dehmer, K. Varmuza, and D. Bonchev), 1–31. Hoboken: John Wiley and Sons, Ltd. 102 Yap, C., Li, H., Ji, Z., and Chen, Y. (2007). Mini-Reviews Med. Chem. 7: 1097–1107. 103 Fujita, T. and Winkler, D.A. (2016). J. Chem. Inf. Model. 56: 269–274. 104 Tripathi, N., Goshisht, M.K., Sahu, S.K., and Arora, C. (2021). Mol. Divers. 2021 253 25: 1643–1664. 105 Jiménez-Luna, J., Grisoni, F., Weskamp, N., and Schneider, G. (2021). Expert Opin. Drug Discov. 16: 949–959. 106 Hammett, L.P. (2002). Chem. Rev. 17: 125–136. 107 Gonçalves, R.M.C. and Albuquerque, L.M.P.C. (2001). J. Phys. Org. Chem. 14: 731–736. 108 Moreira, L., Reis, M., Elvas-Leitão, R. et al. (2019). J. Mol. Liq. 291: 1–8. 109 Abraham, M.H., Doherty, R.M., Kamlet, M.J. et al. (1987). J. Chem. Soc. Perkin Trans. 2: 913–920. 110 Laurence, C., Mansour, S., Vuluga, D., and Legros, J. (2020). J. Phys. Org. Chem. 33: e4067. 111 Aliaga, C., Domínguez, M., Rojas, P., and Rezende, M.C. (2020). J. Mol. Liq. 312: 113362. 112 Martins, F., Moreira, L., Nunes, N., and Leitão, R.E. (2010). J. Therm. Anal. Calorim. 100: 483–491. 113 Reis, M., Moreira, L., Nunes, N. et al. (2011). J. Therm. Anal. Calorim. 108: 761–767. 114 Monteiro, C., Ventura, C., and Martins, F. (2013). J. Environ. Manage. 122: 99–104. 115 Abraham, M.H. (1993). Chem. Soc. Rev. 22: 73–83. 116 Abraham, M.H., Ibrahim, A., and Zissimos, A.M. (2004). J. Chromatogr. A 1037: 29–47. 117 Andzelm, J.W., Alvarado-Swaisgood, A.E., Axe, F.U. et al. (1999). Catal. Today 50: 451–477. 118 Corma, A., Serra, J.M., Serna, P., and Moliner, M. (2005). J. Catal. 2: 335–341. 119 Begum, S. and Achary, P.G.R. (2018). Int. J. Quant. Struct. Relationships 3: 36–48. 120 Chasing, P., Maitarad, P., Wu, H. et al. (2018). Catalysts 8: 422. 121 Dos Santos, V.H.J.M., Pontin, D., Rambo, R.S., and Seferin, M. (2020). J. Am. Oil Chem. Soc. 97: 817–837. 122 Gensch, T., Dos Passos Gomes, G., Friederich, P. et al. (2022). J. Am. Chem. Soc. 144: 1205–1217. 123 Van Der Linden, J.B., Ras, E.J., Hooijschuur, S.M. et al. (2005). QSAR Comb. Sci. 24: 94–98. 124 Köhler, K., Heidenreich, R.G., Soomro, S.S., and Pröckl, S.S. (2008). Adv. Synth. Catal. 350: 2930–2936. 125 Hemmateenejad, B., Sanchooli, M., and Mehdipour, A. (2009). J. Phys. Org. Chem. 22: 613–618. 126 Aguado-Ullate, S., Guasch, L., Urbano-Cuadrado, M. et al. (2012). Catal. Sci. Technol. 2: 1694–1704. 127 Li, B., Dong, Y., and Ding, Z. (2013). J. Environ. Sci. 25: 1469–1476. 128 Rangarajan, S., Bhan, A., and Daoutidis, P. (2014). Appl. Catal. B Environ. 145: 149–160. 129 Shahin, A. and Ganji, S. (2016). Curr. Drug Discov. Technol. 13: 232–253. 130 Yang, W., Yi, J., Ma, Z., and Sun, W.H. (2017). Catal. Commun. 101: 40–43. 131 Prabhu, S., Murugan, G., Cary, M. et al. (2020). Mater. Res. Express 7: 055006. 132 Xie, P., Pu, T., Aranovich, G. et al. (2021). Nat. Catal. 4: 144–156. 133 Arockiaraj, M., Paul, D., Klavžar, S. et al. (2022). J. Mol. Struct. 1250: 131798. 134 Teixeira, C., Ventura, C., Gomes, J.R.B. et al. (2020). Molecules 25: 456. 135 (Q)SARs: guidance documents and reports. OCDE. https://www.oecd.org/fr/env/ess/risques/ guidancedocumentsandreportsrelatedtoqsars.htm (accessed 31 March 2022). 136 Chirico, N. and Gramatica, P. (2011). J. Chem. Inf. Model. 51: 2320–2335. 137 Chirico, N. and Gramatica, P. (2012). J. Chem. Inf. Model. 52: 2044–2058. 138 Wang, Y., Sun, Y., Lancelot, C. et al. (2015). Microporous Mesoporous Mater. 206: 42–51.

Catalysis for a Sustainable Environment

Catalysis for a Sustainable Environment Reactions, Processes and Applied Technologies Volume 3

Edited by Professor Armando J. L. Pombeiro Instituto Superior técnico Lisboa, Portugal

Dr. Manas Sutradhar

Universidade Lusófona de Humanidades e Tecnologias Faculdade de Engenharia Lisboa, Portugal

Professor Elisabete C. B. A. Alegria Instituto Politécnico de Lisboa Departamento de Engenharia Química Lisboa, Portugal

This edition first published 2024 © 2024 John Wiley and Sons Ltd All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/ permissions. The right of Armando J.L. Pombeiro, Manas Sutradhar, and Elisabete C.B.A. Alegria to be identified as the author of the editorial material in this work has been asserted in accordance with law. Registered Offices John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Editorial Office The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com. Wiley also publishes its books in a variety of electronic formats and by print-on-demand. Some content that appears in standard print versions of this book may not be available in other formats. Trademarks: Wiley and the Wiley logo are trademarks or registered trademarks of John Wiley & Sons, Inc. and/ or its affiliates in the United States and other countries and may not be used without written permission. All other trademarks are the property of their respective owners. John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book. Limit of Liability/Disclaimer of Warranty In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. A catalogue record for this book is available from the Library of Congress Hardback ISBN: 9781119870524; ePub ISBN: 9781119870630; ePDF ISBN: 9781119870623; oBook ISBN: 9781119870647 Cover image: © Sasha Fenix/Shutterstock Cover design by Wiley Set in 9.5/12.5pt STIXTwoText by Integra Software Services Pvt. Ltd, Pondicherry, India

v

Contents VOLUME 1 About the Editors  xiii Preface  xv

1

Introduction  1

Armando J.L. Pombeiro, Manas Sutradhar, and Elisabete C.B.A. Alegria

Part I Carbon Dioxide Utilization  5 2

Transition from Fossil-C to Renewable-C (Biomass and CO2) Driven by Hybrid Catalysis  7 Michele Aresta and Angela Dibenedetto

3

Synthesis of Acetic Acid Using Carbon Dioxide  25

Philippe Kalck

4

New Sustainable Chemicals and Materials Derived from CO2 and Bio-based Resources: A New Catalytic Challenge  35

Ana B. Paninho, Malgorzata E. Zakrzewska, Leticia R.C. Correa, Fátima Guedes da Silva, Luís C. Branco, and Ana V.M. Nunes

5

Sustainable Technologies in CO2 Utilization: The Production of Synthetic Natural Gas  55 M. Carmen Bacariza, José M. Lopes, and Carlos Henriques

6

Catalysis for Sustainable Aviation Fuels: Focus on Fischer-Tropsch Catalysis  73

Denzil Moodley, Thys Botha, Renier Crous, Jana Potgieter, Jacobus Visagie, Ryan Walmsley, and Cathy Dwyer

7

Sustainable Catalytic Conversion of CO2 into Urea and Its Derivatives  117

Maurizio Peruzzini, Fabrizio Mani, and Francesco Barzagli

vi

Contents

Part II 8

Transformation of Volatile Organic Compounds (VOCs)  139

Catalysis Abatement of NOx/VOCs Assisted by Ozone  141

Zhihua Wang and Fawei Lin

9

Catalytic Oxidation of VOCs to Value-added Compounds Under Mild Conditions  161

Elisabete C.B.A. Alegria, Manas Sutradhar, and Tannistha R. Barman

10

Catalytic Cyclohexane Oxyfunctionalization  181

Manas Sutradhar, Elisabete C.B.A. Alegria, M. Fátima C. Guedes da Silva, and Armando J.L. Pombeiro

Part III 11

Carbon-based Catalysis  207

Carbon-based Catalysts for Sustainable Chemical Processes  209 Katarzyna Morawa Eblagon, Raquel P. Rocha, M. Fernando R. Pereira, and José Luís Figueiredo

12

Carbon-based Catalysts as a Sustainable and Metal-free Tool for Gas-phase Industrial Oxidation Processes  225

Giulia Tuci, Andrea Rossin, Matteo Pugliesi, Housseinou Ba, Cuong Duong-Viet, Yuefeng Liu, Cuong Pham-Huu, and Giuliano Giambastiani

13

Hybrid Carbon-Metal Oxide Catalysts for Electrocatalysis, Biomass Valorization and, Wastewater Treatment: Cutting-Edge Solutions for a Sustainable World  247 Clara Pereira, Diana M. Fernandes, Andreia F. Peixoto, Marta Nunes, Bruno Jarrais, Iwona Kuźniarska-Biernacka, and Cristina Freire

VOLUME 2 About the Editors  xiii Preface  xv

Part IV

Coordination, Inorganic, and Bioinspired Catalysis  299

14

Hydroformylation Catalysts for the Synthesis of Fine Chemicals  301 Mariette M. Pereira, Rui M.B. Carrilho, Fábio M.S. Rodrigues, Lucas D. Dias, and Mário J.F. Calvete

15

Synthesis of New Polyolefins by Incorporation of New Comonomers  323 Kotohiro Nomura and Suphitchaya Kitphaitun

Contents

16

Catalytic Depolymerization of Plastic Waste  339

Noel Angel Espinosa-Jalapa and Amit Kumar

17

Bioinspired Selective Catalytic C-H Oxygenation, Halogenation, and Azidation of Steroids  369

Konstantin P. Bryliakov

18

Catalysis by Pincer Compounds and Their Contribution to Environmental and Sustainable Processes  389 Hugo Valdés and David Morales-Morales

19

Heterometallic Complexes: Novel Catalysts for Sophisticated Chemical Synthesis  409 Franco Scalambra, Ismael Francisco Díaz-Ortega, and Antonio Romerosa

20

Metal-Organic Frameworks in Tandem Catalysis  429

Anirban Karmakar and Armando J.L. Pombeiro

21

(Tetracarboxylate)bridged-di-transition Metal Complexes and Factors Impacting Their Carbene Transfer Reactivity  445

LiPing Xu, Adrian Varela-Alvarez, and Djamaladdin G. Musaev

22

Sustainable Cu-based Methods for Valuable Organic Scaffolds  461 Argyro Dolla, Dimitrios Andreou, Ethan Essenfeld, Jonathan Farhi, Ioannis N. Lykakis, and George E. Kostakis

23

Environmental Catalysis by Gold Nanoparticles  481 Sónia Alexandra Correia Carabineiro

24

Platinum Complexes for Selective Oxidations in Water  515 Alessandro Scarso, Paolo Sgarbossa, Roberta Bertani, and Giorgio Strukul

25

The Role of Water in Reactions Catalyzed by Transition Metals  537 A.W. Augustyniak and A.M. Trzeciak

26

Using Speciation to Gain Insight into Sustainable Coupling Reactions and Their Catalysts  559

Skyler Markham, Debbie C. Crans, and Bruce Atwater

27

Hierarchical Zeolites for Environmentally Friendly Friedel Crafts Acylation Reactions  577

Ana P. Carvalho, Angela Martins, Filomena Martins, Nelson Nunes, and Rúben Elvas-Leitão

vii

viii

Contents

VOLUME 3 About the Editors  xiii Preface  xv Part V

28

Organocatalysis  609

Sustainable Drug Substance Processes Enabled by Catalysis: Case Studies from the Roche Pipeline  611

28.1 28.2 28.2.1 28.2.2 28.2.3 28.2.4 28.3

Kurt Püntener, Stefan Hildbrand, Helmut Stahr, Andreas Schuster, Hans Iding, and Stephan Bachmann Introduction  611 Case Studies  612 Aleglitazar  612 Idasanutlin  619 Danoprevir  623 Ipatasertib  626 Conclusions  635 References  636

29

Supported Chiral Organocatalysts for Accessing Fine Chemicals  639

29.1 29.2 29.2.1 29.2.2 29.2.3 29.2.4 29.2.5 29.2.5.1 29.2.5.2 29.2.6 29.2.6.1 29.2.6.2 29.2.7 29.2.7.1 29.2.7.2 29.3

30 30.1 30.2 30.2.1

Ana C. Amorim and Anthony J. Burke Introduction  639 Organocatalyst Immobilizations  640 Proline Immobilizations  640 Diphenylprolinol Silyl Ether (Jørgensen-Hayashi Organocatalyst) Immobilizations  643 Organocatalysts Based on Immobilized Pyrrolidines  645 Organocatalysts Based on Immobilized Imidazolidinones  647 Other Amino Acid and Peptide Type Catalysts  648 Supported-primary Amino Acid Catalysts  649 Supported-peptide Derivative Catalysts  650 Immobilized Amino-Cinchona Based Organocatalysts  651 Cinchona Picolinamide Derivatives  651 Cinchona Squaramide Derivatives  652 Other Organocatalysts  653 Phosphoric Acid Catalysts  654 Isothiourea Catalysts  655 Conclusions  657 References  657

Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis  659

Kotohiro Nomura and Nor Wahida Binti Awang Introduction  659 Synthesis of Bio-Based Aliphatic Polyesters by Condensation Polymerization  660 Synthesis of Bio-Based Aliphatic Polyesters by Condensation Polymerization and Dehydrogenative Condensation  661

Contents

30.2.2 30.2.3

30.3

31

Synthesis of BioBasd Aliphatic Polyesters by Acyclic Diene Metathesis (ADMET) Polymerization and Subsequent Hydrogenation  663 One Pot Synthesis of Bio-Based Long Chain Aliphatic Polyesters by Tandem ADMET Polymerization and Hydrogenation. Depolymerization by Reaction with Ethylene  670 Concluding Remarks and Outlook  671 References  672

Modern Strategies for Electron Injection by Means of Organic Photocatalysts: Beyond Metallic Reagents  675

31.6

Takashi Koike Introduction  675 Basic and Advanced Concepts for 1e− Injection by Organic Photoredox Catalysis  675 Triarylamine-based Highly Reducing Organic Photocatalysts  677 Consecutive Photoinduced Electron Transfer (conPET) by Organic Photoredox Catalysis  681 Consecutive Photoinduced Electron Transfer (conPET) by the Combination of Organic Photocatalysis and Electrolysis  683 Summary and Outlook  685 References  685

32

Visible Light as an Alternative Energy Source in Enantioselective Catalysis  687

31.1 31.2 31.3 31.4 31.5

32.1 32.2 32.2.1 32.2.2 32.2.3 32.2.4 32.3 32.3.1 32.3.2 32.4 32.4.1 32.4.2 32.5

Ana Maria Faisca Phillips and Armando J.L. Pombeiro Introduction  687 Dual Chiral Organocatalysis and Photoredox Catalysis  690 Chiral Amines as Catalysts  690 N-Heterocyclic Carbenes (NHCs) as Catalysts  695 Chiral Phosphoric Acids as Catalysts  696 Miscellaneous  700 Metal Catalyzed Processes  702 Dual Transition Metal/Photoredox Catalysis  702 Dual Chiral Lewis Acid/Photoredox Catalysis  707 Chiral Photocatalysts  708 Chiral-at-Metal Photocatalysts  710 Organic Photocatalysts  711 Conclusions  712 Acknowledgements  712 References  712

Part VI 33 33.1 33.2 33.3

Catalysis for the Purification of Water and Liquid Fuels  717

Heterogeneous Photocatalysis for Wastewater Treatment: A Major Step Towards Environmental Sustainability  719 Shima Rahim Pouran and Aziz Habibi-Yangjeh Introduction  719 Heterogeneous Photocatalysis  720 Sustainable Photocatalysts  721

ix

x

Contents

33.3.1 33.3.1.1 33.3.1.2 33.3.2 33.4

Metal Oxide-based Photocatalysts  722 Magnetic Metal Oxide Semiconductors  725 Green Synthesis Routes  730 Carbonaceous Photocatalysts  732 Remarks and Future Perspectives  737 Acknowledgments  737 References  737

34

Sustainable Homogeneous Catalytic Oxidative Processes for the Desulfurization of Fuels  743

34.1 34.2 34.3 34.4 34.5 34.6 34.7 34.8 34.9 34.10

35

35.1 35.2 35.3 35.3.1 35.3.2 35.3.3 35.3.4 35.4 35.4.1 35.4.2 35.4.2.1 35.4.2.2 35.4.2.3 35.4.2.4 35.4.2.5 35.4.3 35.4.3.1 35.4.3.2 35.4.3.3 35.4.3.4

Federica Sabuzi, Giuseppe Pomarico, Pierluca Galloni, and Valeria Conte Introduction  743 Vanadium  743 Manganese  746 Iron  746 Cobalt  748 Molybdenum  749 Tungsten  750 Polyoxometalates  750 Ionic Liquids  751 Conclusions  753 References  753

Heterogeneous Catalytic Desulfurization of Liquid Fuels: The Present and the Future  757

Rui G. Faria, Alexandre Viana, Carlos M. Granadeiro, Luís Cunha-Silva, and Salete S. Balula Introduction  757 Hydrodesulfurization  758 Adsorptive Desulfurization  761 ADS with Carbon-based Materials  762 ADS with Zeolites  763 ADS with Mesoporous Silica  765 ADS with Metal-organic Frameworks  765 Oxidative Desulfurization  767 Oxidants for (EC)ODS  768 Heterogeneous Catalysts for (EC)ODS  768 (EC)ODS with Zeolites  768 (EC)ODS with Metal-organic Frameworks  769 (EC)ODS with Carbon-based Materials  771 (EC)ODS with Mesoporous Silicas  773 (EC)ODS with Titanate Nanotubes  773 (EC)ODS Catalyzed by Heterogeneous Polyoxometalates  774 Carbonaceous Composites  774 MOF Composites  775 Zeolite Composites  775 Mesoporous Silica Composites  776

Contents

35.5 35.6

(EC)ODS Catalyzed by Membranes  777 Future Perspectives  778 Acknowledgments  779 References  779

Part VII 36

Hydrogen Formation, Storage, and Utilization  783

Paraformaldehyde: Opportunities as a C1-Building Block and H2 Source for Sustainable Organic Synthesis  785

36.2.4 36.2.5 36.3 36.4 36.5

Ana Maria Faísca Phillips, Maximilian N. Kopylovich, Leandro Helgueira de Andrade, and Martin H.G. Prechtl Introduction  785 Carbonylation and Related Reactions  787 Alkoxycarbonylation of Olefins  788 Carbonylation of Aryl Halides  790 Cascade C–H Activation/carbonylation/cyclization Reactions and Related Processes: The Synthesis of Heterocycles  793 Hydroformylation of Alkenes  794 N-formylation  798 Methylation and Related Reactions  799 Hydrogen Generation and Transfer-hydrogenation Reactions  810 Summary and Outlook  815 Acknowledgement  815 References  815

37

Hydrogen Storage and Recovery with the Use of Chemical Batteries  819

36.1 36.2 36.2.1 36.2.2 36.2.3

37.1 37.2 37.3 37.4 37.5 37.6 37.7 37.7.1 37.7.2 37.8 37.9 37.9.1 37.9.2 37.9.3 37.9.4 37.9.5 37.10

Henrietta Horváth, Gábor Papp, Ágnes Kathó, and Ferenc Joó Introduction  819 Hydrogen as an Energy Storage Material  820 Chemical Hydrogen Storage  821 Liquid Organic Hydrogen Carriers  822 Definitions and Fundamental Questions  824 Catalysts Applied in Hydrogen Batteries  826 Formic Acid and Formate Salts as Storage Materials in Hydrogen Batteries  827 Formic Acid as a Hydrogen Storage Material  828 Formate Salts as Hydrogen Storage Materials  829 Catalysts and Reaction Conditions Potentially Applicable in Hydrogen Batteries Based on the Formate-bicarbonate Equilibrium  831 Functional Hydrogen Batteries  832 Hydrogen Batteries Based on CO2–formic Acid Cycles  833 Hydrogen Batteries Based on Formate–bicarbonate Cycles  834 Hydrogen Batteries Based on N-heterocyclic Compounds  836 Hydrogen Batteries Based on Alcohols  837 Hydrogen Batteries Based on Whole-cell Biocatalysis  839 Summary and Conclusions  840 Acknowledgements  840 References  841

xi

xii

Contents

38

Low-cost Co and Ni MOFs/CPs as Electrocatalysts for Water Splitting Toward Clean Energy-Technology  847

38.1 38.2 38.3 38.3.1 38.3.1.1 38.3.1.2 38.3.2 38.3.2.1 38.3.2.2 38.3.3 38.4

Anup Paul, Biljana Šljukić, and Armando J.L. Pombeiro Introduction  847 Fundamentals of Water Splitting Reactions  849 MOFs/CPs as Electrocatalysts for Water Splitting Reactions  852 Co MOFs and Derived Electrocatalysts for OER and HER  852 Co MOFs and Derived Electrocatalysts for OER  852 Co MOFs and Derived Composites for HER  857 Ni MOFs and Derived Composites for OER and HER  859 Ni MOFs and Derived Composites for the OER  859 Ni MOFs and Derived Composites for HER  863 Polyhomo and Heterometallic MOFs of Co(II) or Ni(II) for OER and HER  864 Conclusions  868 Acknowledgements  868 References  868



Index  871

xiii

About the Editors Armando Pombeiro is a Full Professor Jubilado at Instituto Superior Técnico, Universidade de Lisboa (ULisboa), former Distant Director at the People’s Friendship University of Russia (RUDN University), a Full Member of the Academy of Sciences of Lisbon (ASL), the President of the Scientific Council of the ASL, a Fellow of the European Academy of Sciences (EURASC), a Member of the Academia Europaea, founding President of the College of Chemistry of ULisboa, a former Coordinator of the Centro de Química Estrutural at ULisboa, Coordinator of the Coordination Chemistry and Catalysis group at ULisboa, and the founding Director of the doctoral Program in Catalysis and Sustainability at ULisboa. He has chaired major international conferences. His research addresses the activation of small molecules with industrial, environmental, or biological significance (including alkane functionalization, oxidation catalysis, and catalysis in unconventional conditions) as well as crystal engineering of coordination compounds, polynuclear and supramolecular structures (including MOFs), non-covalent interactions in synthesis, coordination compounds with bioactivity, molecular electrochemistry, and theoretical studies. He has authored or edited 10 books, (co-)authored ca. 1000 research publications, and registered ca. 40 patents. His work received over. 30,000 citations (over 12,000 citing articles), h-index ca. 80 (Web of Science). Among his honors, he was awarded an Honorary Professorship by St. Petersburg State University (Institute of Chemistry), an Invited Chair Professorship by National Taiwan University of Science & Technology, the inaugural SCF French-Portuguese Prize by the French Chemical Society, the Madinabeitia-Lourenço Prize by the Spanish Royal Chemical Society, and the Prizes of the Portuguese Chemical and Electrochemical Societies, the Scientific Prizes of ULisboa and Technical ULisboa, and the Vanadis Prize. Special issues of Coordination Chemistry Reviews and the Journal of Organometallic Chemistry were published in his honor. https://fenix.tecnico.ulisboa.pt/homepage/ist10897

xiv

About the Editors

Manas Sutradhar is an Assistant Professor at the Universidade Lusófona, Lisbon, Portugal and an integrated member at the Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa, Portugal. He was a post‐doctoral fellow at the Institute of Inorganic and Analytical Chemistry of Johannes Gutenberg University of Mainz, Germany and a researcher at the Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa. He has published 72 papers in international peer review journals (including three reviews + 1 reference module), giving him an h-index 28 (ISI Web of Knowledge) and more than 2250 citations. In addition, he has 11 book chapters in books with international circulation and one patent. He is one of the editors of the book Vanadium Catalysis, published by the Royal Society of Chemistry. His main areas of work include metal complexes with aroylhydrazones, oxidation catalysis of industrial importance and sustainable environmental significance, magnetic properties of metal complexes, and bio-active molecules. The major contributions of his research work are in the areas of vanadium chemistry and oxidation catalysis. He received the 2006 Young Scientist Award from the Indian Chemical Society, India and the Sir P. C. Ray Research Award (2006) from the University of Calcutta, India. https://orcid.org/0000-0003-3349-9154 Elisabete C.B.A. Alegria is an Adjunct Professor at the Chemical Engineering Department of the Instituto Superior de Engenharia de Lisboa (ISEL) of the Polytechnic Institute of Lisbon, Portugal. She is a researcher (Core Member) at the Centro de Química Estrutural (Coordination Chemistry and Catalysis Group). She has authored 86 papers in international peered review journals and has an h-index of 23 with over 1600 citations, four patents, five book chapters, and over 180 presentations at national and international scientific meetings. She was awarded an Honorary Distinction (2017–2020) for the Areas of Technology and Engineering (Scientific Prize IPL-CGD). She is an editorial board member, and has acted as a guest editor and reviewer for several scientific journals. Her main research interests include coordination and sustainable chemistry, homogeneous and supported catalysis, stimuli‐responsive catalytic systems, green synthesis of metallic nanoparticles for catalysis, and biomedical applications. She is also interested in mechanochemistry (synthesis and catalysis) and molecular electrochemistry. https://orcid.org/0000-0003-4060-1057

xv

Preface Aiming to change the world for the better, 17 Sustainable Development Goals (SDGs) were adopted by the United Nations (UN) Member States in 2015, as part of the UN 2030 Agenda for Sustainable Development that concerns social, economic, and environmental sustainability. Hence, a 15-year plan was set up to achieve these Goals and it is already into its second half. However, the world does not seem to be on a good track to reach those aims as it is immersed in the Covid-19 pandemic crisis and climate emergency, as well as economic and political uncertainties. Enormous efforts must be pursued to overcome these obstacles and chemical sciences should play a pivotal role. Catalysis is of particular importance as it constitutes the most relevant contribution of chemistry towards sustainable development. This is true even though the SDGs are integrated and action in one can affect others. For example, the importance of chemistry and particularly catalysis is evident in several SDGs. Goal 12, addresses “Responsible Consumption and Production Patterns” and is aligned with the circularity concept with sustainable loops or cycles (e.g., in recycle and reuse processes that are relevant within the UN Environmental Program). Goal 7 addresses “Affordable and Clean Energy” and relates to efforts to improve energy conversion processes, such as hydrogen evolution and oxygen evolution from water, that have a high environmental impact. Other SDGs in which chemistry and catalysis play an evident role with environmental significance include Goal 6 (“Clean Water and Sanitation”), Goal 9 (“Industry, Innovation and Infrastructure” 13 (“Climate Action”), Goal 14 (“Life Below Water”), and Goal 15 (“Life on Land”). The book is aligned with these SDGs by covering recent developments in various catalytic processes that are designed for a sustainable environment. It gathers skilful researchers from around the world to address the use of catalysis in various approaches, including homogeneous, supported, and heterogeneous catalyses as well as photo- and electrocatalysis by searching for innovative green chemistry routes from a sustainable environmental angle. It illustrates, in an authoritative way, state-of-the-art knowledge in relevant areas, presented from modern perspectives and viewpoints topics in coordination, inorganic, organic, organometallic, bioinorganic, pharmacological, and analytical chemistries as well as chemical engineering and materials science. The chapters are spread over seven main sections focused on Carbon Dioxide Utilization, Transformation of Volatile Organic Compound (VOCs), Carbon-based Catalysts, Coordination, Inorganic, and Bioinspired Catalysis, Organocatalysis, Catalysis for the Purification of Water and Liquid Fuels,and Hydrogen Formation, Storage, and Utilization. These sections are gathered together as a contribution towards the development of the challenging topic.

xvi

Preface

The book addresses topics in (i) activation of relevant small molecules with strong environmental impacts, (ii) catalytic synthesis of important added value organic compounds, and (iii) development of systems operating under environmentally benign and mild conditions toward the establishment of sustainable energy processes. This work is expected to be a reference for academic and research staff of universities and research institutions, including industrial laboratories. It is also addressed to post-doctoral, postgraduate, and undergraduate students (in the latter case as a supplemental text) working in chemical, chemical engineering, and related sciences. It should also provide inspiration for research topics for PhD and MSc theses, projects, and research lines, in addition to acting as an encouragement for the development of the overall field. The topic Catalysis for Sustainable Environment is very relevant in the context of modern research and is often implicit, although in a non-systematic and disconnected way, in many publications and in a number of initiatives such as international conferences. These include the XXII International Symposium on Homogeneous Catalysis (ISHC) that we organized (Lisbon, 2022) and that to some extent inspired some parts of this book. In contrast to the usual random inclusion of the topic in the literature and scientific events, the applications of catalytic reactions focused on a sustainable environment in a diversity of approaches are addressed in this book. The topic has also contributed to the significance of work that led to recent Nobel Prizes of Chemistry. In 2022, the Nobel Prize was awarded to Barry Sharpless, Morten Meldal, and Carolyn Bertozzi for the development of click chemistry and bioorthogonal chemistry. The set of criteria for a reaction or a process to meet in the context of click chemistry includes, among others, the operation under benign conditions such as those that are environmentally friendly (e.g., preferably under air and in water medium). In 2021, the Nobel Prize was awarded to Benjamin List and David W.C. MacMillan for the development of asymmetric organocatalysis, which relies on environmentally friendly organocatalysts. The book illustrates the connections of catalysis with a sustainable environment, as well as the richness and potential of modern catalysis and its relationships with other sciences (thus fostering interdisciplinarity) in pursuit of sustainability. At last, but not least, we should acknowledge the authors of the chapters for their relevant contributions, prepared during a particularly difficult pandemic period, as well as the publisher, John Wiley, for the support, patience, and understanding of the difficulties caused by the adverse circumstances we are experiencing nowadays and that constituted a high activation energy barrier that had to be overcome by all of us… a task that required rather active catalysts. We hope the readers will enjoy reading its chapters as much as we enjoyed editing this book. Armando Pombeiro Manas Sutradhar Elisabete Alegria

609

Part V Organocatalysis

611

28 Sustainable Drug Substance Processes Enabled by Catalysis Case Studies from the Roche Pipeline Kurt Püntener, Stefan Hildbrand, Helmut Stahr, Andreas Schuster, Hans Iding, and Stephan Bachmann Pharmaceutical Division, Synthetic Molecules Technical Development, Process Chemistry and Catalysis, F. Hoffmann-La Roche Ltd, Basel, Switzerland

28.1  Introduction As a pioneer in healthcare, Roche has been committed to improving lives since the company was founded 1896 in Basel (CH). Delivering new drugs for innovative healthcare solutions is a tremendous undertaking with a journey, starting from understanding a disease fundamentally through the discovery and development of the therapeutic all the way to the manufacture and administration of the product. Adding to this challenge nowadays is the increasing complexity of the chemical structures of the drug candidates, with most of them bearing multiple chiral centers, complex ring ­systems, and a manifold of different functional groups that were introduced for optimal drug targeting, drug metabolism and pharmokinetics (DMPK), and toxicological profiles. Furthermore, increasing pressure on the pharmaceutical industry to provide affordable drugs and ultimately cost-efficient therapeutic modalities to enable treatments be accessible to patients worldwide contributes to this challenge. Whereas cost-efficiency in times past was associated with the direct costof-goods or manufacturing cost, more recently it has become evident that this term must include additionally the environmental impact of the drug substance (DS) manufacture. At present, it is compellingly self-evident that natural resources are limited and represent a precious commodity for everyone. To this end, the chemical industry has created new metrics on how to measure the “greenness” of their chemical processes and, ultimately, to didactically guide chemists on essential process optimizations. As one of the most prominent metrics, the process mass index (PMI = kg of input materials (solvents and reagents) needed to manufacture 1 kg of DS) has made its way into use in the fine chemical, agricultural, and pharmaceutical industries [1]. This index is used to systematically monitor the sustainability of DS processes and is determined from the very first good laboratory practices (GLP) toxicology supply campaigns through to launch of supplies to highlight the impact that process improvements have had during the course the various drug development phases. Most importantly, the process steps and operations which demonstrate the highest potential for overall resource savings along the way direct as a consequence the process research and development activities. Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 3, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

612

28  Sustainable Drug Substance Processes Enabled by Catalysis

Figure 28.1  The structures of drug candidates 1 (aleglitazar), 2 (idasanutlin), 3 (danoprevir), and 4 (ipatasertib).

To encourage Roche chemists to constantly strive for most sustainable DS processes and to recognize the contributions they have made in this field, the Roche Environmental Awareness in Chemical Technology (REACT) award was introduced. On the basis of the green chemistry principles established by Anastas and Warner [2], the main selection criteria for winning the REACT award are the following: i) shift from avoid/high uncertain to recommended/usable solvents based on the Roche solvent selection guide based on ACS classifications [3]; ii) reduction of PMI factor; iii) introduction of new reaction types, leading to cleaner, more efficient reactions with high atom efficiency; and iv) reduction of energy consumption. The following four case studies outline Roche’s effort to establish most sustainable processes for its investigated drug candidates 1–4 (Figure 28.1). The project teams of danoprevir (3), ipatasertib (4), and idasanutlin (2) were distinguished with REACT awards were achieved back in 2010, 2011 and 2014 [4]. Both cutting-edge process chemistry and catalytic methodologies contributed equally to this success. Although the focus for this casebook chapter is the impact catalysis has had on establishing sustainable DS processes, the contributions from process chemistry have also been fundamental, particularly in delivering efficient syntheses of the substrates for the targeted catalytic transformations and the ensuing downstream chemistry. The reported case studies highlight these contributions and, from a holistic point of view, the attributes a DS process must fulfill to be considered as sustainable.

28.2  Case Studies 28.2.1  Aleglitazar Aleglitazar (RG1439) (1), a potent, balanced dual peroxisome proliferator activated PPARα/γ agonist, showed insulin-sensitizing and glucose-lowering activity and favorable effects on lipid profiles [5]. Clinical studies were run to determine whether 1

28.2  Case Studies

and mortality among patients with type 2 diabetes mellitus (T2DM) suffering from recent acute coronary syndrome (ACS) when this regimen was added to standard treatment. However, due to safety signals and lack of efficacy, the further development of 1 was halted in 2013. Aleglitazar (1) and edaglitazone (BM131258) (5) [6] shared formylbenzothiophene 6 as a common intermediate (Scheme 28.1). When the PPARα/γ program was initiated, 5 was in advanced

Scheme 28.1  1st and 2nd generation synthesis of 9 for the manufacture of 1 and 5 via formylbenzothiophene 6.

613

614

28  Sustainable Drug Substance Processes Enabled by Catalysis

clinical development and, as a consequence, the synthesis of 6 was already well established and the building block available in multi-kg quantities. Accordingly, both the Discovery Chemistry synthesis of 1 as well as any new envisaged route toward 1 could benefit from readily available 6 or an advanced intermediate thereof. The synthesis of 6 comprised a total of eight steps (Scheme 28.1) [7, 8]. In the main branch, 2-formylthiophene (7) was converted into racemic tetrabutylammonium 2-hydroxy-2-(4-hydroxyphenyl) acetate (10), which, after iron assisted reductive decarboxylation and coupling with oxazole mesylate 18, provided 6 in an overall 32% yield (four steps). Mesylate 18 itself was readily accessible in four steps (49% yield) from 3-oxovalerate 12. Key intermediate 9 in the synthesis of 6 was initially accessed following a procedure reported by Hidai et al [9]. In a palladium catalyzed cyclocarbonylation reaction, 3-(2-thienyl)allyl acetate (19) (prepared in three steps (27% yield) from 7) was treated under CO (100 bar) and, in the presence of Ac2O, Et3N and a catalytic amount of Pd(OAc)2/PPh3 in toluene, delivered the product, after the saponification of phenol 9, in 68% yield (two steps). The harsh reaction conditions employed in the cyclocarbonylation step and the high catalyst loading (2.5 mol%) called for significant process improvements to render the sequence feasible on a technical scale. A breakthrough was achieved when 1-(2-thienyl)allyl acetate (8) was employed in place of its regiosiomer 19 for the cyclocarbonylation reaction. The acetate isomer 8 was readily accessible (87% yield) by addition of vinyl magnesium bromide to 7 followed by acylation with acetic anhydride and the cyclocarbonylation could be carried out under much milder conditions at significantly lower catalyst loadings (0.07 mol%). Because trace amounts of 19 were detected by high-performance liquid chromatography (HPLC) during the conversion of 8 into 9, it is very likely that 8 isomerizes prior to cyclization into 19, indicating that the overall process presumably follows the mechanism as proposed by Hidai. Introducing 8 as a synthesis equivalent for 19 enabled three steps to be cut and furnished 9 in a four-fold higher overall yield (73 vs 18%). To introduce the aldehyde function into 6, benzothiophene 9 was treated with glyoxalic acid/ KOH followed by the addition of Bu3N to readily isolate the 7-hydroxyl carboxylate formed as its ammonium salt 10. In the next step, 10 was subjected to iron assisted reductive decarboxylation to deliver 4-hydroxyl-2-formyl-benzothoiophene (11) which, in the final process, was converted without isolation into 6 (83% yield) through coupling with oxazole mesylate 18. In four steps, 18 was obtained in 48% yield from methyl 3-oxovalerate 12. The ketoester was treated with trimethyl orthoformate in the presence of amberlyst-15, which delivered a 2:1 mixture of enolester 13 and its ketal 14. Without separation, subsequent bromination with 1,3-dibromo-5,5dimethylhydantoin/2,2′-azobis(2-methylpropionitrile) (DBH/AIBN) furnished the intermediate bromo enolester 15, which was converted (in the presence of benzamide and trace amounts of acid) into the phenyl oxazole 16. After sodium borohydride reduction, the alcohol 17 formed was converted into 18 via treatment with MsCl/Et3N. With 6 in hand as an advanced building block, the end game of the Discovery Chemistry synthesis became quite straightforward. This employed the Evans methodology based on boron mediated diastereoselective aldol reactions that promoted the selective introduction of the corresponding chiral centers into key intermediate 22 (Scheme 28.2). (S)-4-Benzyl-2-oxazolidinone (20) was deprotonated at −78 °C with butyl lithium and treated with methoxyacetyl chloride, furnishing N-acylated oxazolidinone 21 isolated in 94% yield after chromatographic purification. Subsequently, 21 was condensed with 6 in the presence of dibutylboron triflate at -78 °C, which provided a 91:6:3 mixture of the Evans-syn, Evans-anti, and non-Evans-syn aldol products 22. Because the nonEvans-syn product delivers the undesired enantiomer of 1 after the subsequent transformations, it was chromatographically removed from the crude reaction mixture. The obtained Evans-syn and

28.2  Case Studies

Scheme 28.2  Discovery Chemistry end game synthesis of 1.

Evans-anti aldol product mixture (91:6) was then treated with triethylsilane/trifluoracetic acid (TFA) to promote dehydroxylation and deliver diastereomerically pure 23. After chromatographic purification and subsequent saponification, 1 with >99.9:0.1 er and 99.9 area-% in an overall yield of 56% (from 20) was obtained. Although the Discovery Chemistry route was quite short, it suffered from i) the employment of three chromatographic purifications; ii) the need to run two steps under cryogenic conditions; iii) the use of dibutylboron triflate which was expensive and not available in large quantities at that time; and iv) the handling of stoichiometric amounts of a chiral auxiliary that was used to install the stereo-information in 1. However, to ensure timely DS supplies for initial toxicology and clinical studies, the Discovery Chemistry route was retained. Work was preliminarily focused to reduce the number of chromatographic purification steps to shorten lead times. Step one product 21 could be purified by crystallization rendering the first chromatography obsolete. The second chromatography, needed to remove the undesired non-Evans-syn aldol product and avoid contamination of 1 with its (R)enantiomer, was intentionally skipped as the er-value of crude 1 could be enhanced by crystallization downstream. A simple silica gel filtration, nevertheless, was required to achieve a substrate quality suitable for the next steps. Additionally, in the subsequent dehydroxylation step, the amount of SiO2 could be reduced by replacing the chromatography by a filtration instead delivering 23 in an acceptable dr 97:3 quality. After the final saponification, crude 1 was recovered with a corresponding 97:3 er. After crystallization from ethyl acetate pure 1 (>99.9 area-%) was isolated with >99.9:0.1 er in an overall yield of 45–50% from carbamate 20. According to this protocol, two batches of 1 (0.5 and 5.8 kg) were produced enabling a timely commencement of GLP tox and clinical phase 1 studies. Although the three chromatographic purifications could be reduced to two silica gel filtrations leading to a significant decrease in eluent solvent and SiO2 consumption, it was not possible to omit these waste intense treatments entirely. As a result, process research work was initiated to

615

616

28  Sustainable Drug Substance Processes Enabled by Catalysis

determine a new route toward 1 that would i) rely on the installment of the chiral center through catalytic methodologies; ii) make use of readily available 6; iii) be short to remain competitive with the existing route; and iv) deliver 1 in high yield and with a low PMI. Most promising endgame variants were considered to rely upon the asymmetric reduction of a propenoic acid derivative or of a racemic 2-methoxy-3-oxo-propanoic derivative under dynamic kinetic control [10]. The later approach indeed was successful to deliver syn-24 (cf. Scheme 28.3) from its 3-oxo precursor in high yield (99%) as well as selectivity (97:3 dr, >99.8:0.2 er) [11] and furnish the envisaged target compound 1 after reductive dehydroxylation and saponification. However, introducing such a racemic 2-methoxy-3-oxo-propanoic unit starting from 6 proved cumbersome and this approach was abandoned. Ultimately, prime emphasis was put on approaches that comprised the chiral reduction of 2-methoxypropenoic acid derivatives that would yield either 1 directly or furnish the target product after subsequent saponification of an ester thereof (Scheme 28.3). Such an approach had the highest potential to outperform the Discovery Chemistry route. On the other hand, there was a risk of failure as neither 2-methoxypropenoic acid nor its ester variants had been reported as substrates for asymmetric (transfer) hydrogenations or biocatalysis. Nonetheless, based on Noyori’s pioneering work on Ru-BINAP catalyzed asymmetric hydrogenation (AH) of α/β-unsaturated acids [12], as well as a few other reported rhodium and ruthenium based catalyst systems [13], a tailor-made catalyst system to promote hydrogenation of 2-methoxypropenoic acids or a derivative

Scheme 28.3  Investigated routes toward propenoic acid derivatives as substrates for asymmetric reductions.

28.2  Case Studies

thereof and to install the desired chirality in 1 seemed reasonable. Notably though, the efficiency of the reported systems at the time was greatly substrate dependent, and for most reported catalysts, a high catalyst loading, drastic conditions, or long reaction times were required for satisfactory results. As a saving grace, if the er in the asymmetric reduction of 2-methoxypropenoic acid derivatives were not particularly high, crystallization of 1 or a diastereomeric ammonium salt thereof would still have provided a viable option to upgrade the optical purity of the target product. To acquire a rapid proof-of-concept, propenoic ester 25 was prepared under Wittig conditions from 6 and (1,2-dimethoxy-2-oxoethyl)triphenylphosphonium chloride (DBU, THF, 22 °C, Z/Eratio 5:1). After chromatography, pure 25 was converted into 26 via saponification. Preliminary screening results with either enzymes or chiral metal catalysts under hydrogen pressure for the reduction of 25 or 26 clearly demonstrated that only the hydrogenation approach employing ruthenium or rhodium catalysts and acid 26 as a substrate delivered useful conversions and selectivities (up to 95:5 er) [14]. Furthermore, employing pure Z-acid led to constantly higher er-values than its E-isomer or mixtures thereof. Consequently, the selective preparation of pure 26 was needed to achieve the highest er. In addition, a high substrate quality was deemed crucial for ensuring low catalyst loadings in the AH step, to save catalyst costs and permit facile removal of residual metal contamination in 1. Initial work was focused on improving the yield and selectivity in the Wittig reaction. Changing the reaction conditions from DBU/THF to KOtBu/DMF and increasing the reaction temperature to 75 °C brought noticeable improvements. Ester 25 and its E-isomer were formed in a 7:1 ratio, whereupon 76% of pure 25 precipitated after cooling the reaction mixture. The mother liquor containing 25 and its E-isomer (ratio 1:6) was treated with 2-methyl-5-t-butylthiophenol/AIBN at 100 °C to isomerize the E- into its Z-isomer. An additional 8% of pure 25 (total yield 84%) were thereby obtained. A less waste intensive and technically more feasible route to pure 26 was identified by treating aldehyde 6 with methyl acetoxyacetate in the presence of LDA at -78 °C or alternatively with a TiCl4/Et3N mixture at 0 °C yielding the syn/anti aldol products 24 in ratios of 1:4 (94% yield) and 6:4 (75% yield) respectively, after chromatography. Beneficially, the crude syn/anti-24 mixture upon acid catalyzed 1,2-elimination of water furnished exclusively the thermodynamically more stable Z-ester 25 independent of the syn/anti ratio of employed. Finally, the more preferred LDA variant delivered 25 in one step (89% yield) without purification of the aldol intermediate, independent of whether sulfuric acid or pTsOH, employed later, was used to promote water elimination. With an efficient process for the selective synthesis of pure 25 and an optimized procedure for the subsequent saponification affording crystalline Z-acid 26 (95% yield) now in place, the optimization of the AH process was addressed. For this a wide range of rhodium, iridium and ruthenium catalysts were screened whereby, [Ru(OAc)2((S)TMBTP)] (27) [13b] and [Ru(OAc)2((S)-1-Naphtyl-MeOBIPHEP)] (28) [15] emerged as favorable hits. These two catalysts delivered crude 1 with 95:5 and 85:15 er respectively under the screening conditions (Scheme 28.4). Interestingly, none of the iridium catalysts tested provided any conversion. The few rhodium based catalysts that delivered decent er-values (up to 90:10 er) were finally dropped on account of the associated high metal cost and ultimately because of the superior performance of the significantly less expensive catalyst 27. In parallel, a process was also developed with catalyst 28 containing (S)-1-Naphtyl-MeOBIPHEP, a member of the Roche proprietary atropisomeric MeOBIPEP ligand family [16], to create a third party independent route. Key for high catalyst performance was the employment of 0.2 eq of NaOMe as base and a 3:2 solvent mixture of MeOH/DCM. Under these conditions, with catalyst 27 and substrate-to-catalyst ratios (S/C) of up to 10,000, a full conversion was achieved in six hours at 30 bar H2 and 40–60 °C, delivering crude 1 with 97:3 er. Following acidification with HCl and two crystallizations, acid

617

618

28  Sustainable Drug Substance Processes Enabled by Catalysis

Scheme 28.4  Asymmetric hydrogenation (AH) of 26 employing various catalyst types/reaction conditions and downstream operations.

1 was isolated in 70% yield and >99.9:0.1 er. As expected, the er-enhancement was rendered more efficient when crude 1 was first crystallized as its (S)-phenylethylamine (PEA) salt. After HCl treatment crude 1 with 99.5:0.5 er was recovered. A subsequent crystallization created pure 1 with >99.9:0.1 er on a lab scale with 20% higher yield (90%) compared to Variant A conditions. To streamline the hydrogenation process even further, (S)-PEA was employed in lieu of NaOMe as base without affecting reaction rate or selectivity. After hydrogenation, the remaining 0.8 eq of (S)PEA was added to deliver the proper salt constitution. In summary, 6.7 kg of 1 were produced in 72% yield (four steps from 6) for clinical phase 2 supply employing catalyst 27 (S/C 3000) and one (S)-PEA salt crystallization for ee-upgrading. In a follow-up campaign, 19.6 kg of 1 were obtained in 56% yield, this time benefitting from the more readily accessible Roche catalyst 28 (S/C 3000). Herein, a second (S)-PEA salt crystallization step was added to deliver enantiopure 1 (>99.9:0.1 er). Finally, both processes produced 1 in comparable purity (>99.9 area-%) and with residual ruthenium levels of 99.9:0.1 er. Although the er of crude 1 would have met the specifications already, the (S)-PEA crystallization was retained in the process as was the final crystallization. The first part provided the best control over the impurity profile and for the depletion of residual iridium and the second part obtain the desired polymorph (form A). With the higher quality of crude 1, the downstream crystallizations proceeded with improved yields and superior efficiency. In the registration campaign, enantiopure 1 was isolated with 62% yield (four steps from 6), >99.9 area-% purity and a double time-space yield. Consequently, the HCDP occupancy time for the manufacturing of 1 reached very acceptable levels. Overall, the new AH process enabled the preparation of 1 in a very robust and atom-efficient manner. The use of stoichiometric amounts of a chiral auxiliaries for setting up the chiral center and the three chromatographic purification steps (which created large volumes of solvent and silica gel wastes) could be skipped entirely. The new process benefitted greatly from the identification of 31 as the catalyst that furnished almost perfect ee-induction (>99.9:0.1 er). Furthermore, it also delivered the hydrogenation substrate 26 in a reaction sequence that, in contrast to the first Wittig approach, generated only very minimal amounts of waste and thereby contributed to a significantly improved PMI of 281 (from 6).

28.2.2  Idasanutlin Idasanutlin (RG7388) (2) is a potent oral inhibitor of the mouse double minute 2 homolog (MDM2) protein antagonist and is being investigated at Roche as a potential treatment for a variety of solid tumors and hematologic malignancies [20]. As depicted in Scheme 28.5, 2 contains a pyrrolidine carboxamide core with four contiguous stereocenters with an all-anti relative configuration (2R,3S,4R,5S). The original medicinal chemistry synthesis for 2 employed an azomethine ylide-based [3+2] cycloaddition reaction approach with electron-deficient olefins, a logical synthetic disconnection [21]. Idasanutlin (2) was prepared in six steps and 16% overall yield starting from 2-(4-chloro-2-fluorophenyl) acetonitrile (32) and

619

620

28  Sustainable Drug Substance Processes Enabled by Catalysis

Scheme 28.5  Discovery Chemistry synthesis of 2. Reproduced with permission from Ref [23a].

employed silver fluoride for the key cycloaddition reaction to construct the racemic pyrrolidine core (Scheme 28.5) [22]. This process was not suitable for larger scale preparation due to several reasons: i) the use of stoichiometric amounts of AgF (1.2 g Ag were used to create 1 g of final product 2); ii) the use of large quantities of chlorinated solvents (76 g to make 1 g of final product 2); and iii) the need for two chromatographic purifications and an additional chiral supercritical fluid chromatography (SFC) for enantiomer separation. Intense process chemistry efforts toward a more efficient and scalable process to support clinical studies resulted in the elaboration of a catalytic asymmetric [3+2]-cycloaddition reaction of stilbene 34 and imine 41 using a catalytic amount AgOAc with (R)-MeOBIPHEP (42) as a chiral ligand. In the presence of 1.0 mol% of AgOAc and 1.1 mol% of 42, the reaction of 34 and 41 in 2-MeTHF (0 °C, 15 hours) afforded a mixture of diastereoisomers 43a-c (75 area-% by HPLC in total) and the ring-opened Michael addition product 44 (11 area-% by HPLC) (Scheme 28.6) [23]. Upon treatment of this reaction mixture directly with micronized anhydrous LiOH, the pyrrolidine diastereoisomers 43a and 43b, as well as the Michael addition product 44, were converted to the thermodynamically most stable desired ester 43c (84:16 er), which was isolated by crystallization from n-heptane/2-MeTHF in 97% yield. Hydrolysis of 43c with LiOH in iPrOH followed by filtration of the insoluble racemate, furnished a filtrate comprising enantioenriched 2 (as its lithium salt). Subsequent acidification of the filtrate with AcOH and isolation by crystallization from iPrOH/water afforded 2 with >99:1 er and 44% overall yield from 34 (36% from 32). The Ag-catalyzed process previously outlined was effectively employed to produce more than 100 kg of 2 for early clinical trials. Despite the successful supply campaigns, this process nevertheless was not considered appropriate for a potential commercial manufacturing process for the following reasons: i) efficient and complete epimerization could only be achieved with finely milled

28.2  Case Studies

Scheme 28.6  Ag-catalyzed process for the manufacture of 2. Adapted from Ref [23a] and Ref [23b].

and anhydrous LiOH; ii) silver oxide and other fine particles precipitated during the hydrolysis process and this lead to serious centrifuge clogging during filtration of the racemate; iii) laborious intermediate reactor cleaning was necessary due to the formation of Ag deposits on the reactor walls, and iv) the modest enantioselectivity of the cycloaddition reaction lead to an erosion of the yield due to the necessity for enantioenrichement via racemate removal. In an effort to address the disadvantages of the Ag process and to improve the selectivity of the [3+2] cycloaddition reaction, the use of Cu catalysts was investigated. After intense screening, CuOAc as the pre-catalyst and readily available (R)-BINAP (46) as a chiral ligand were found to be the most suitable combination for the asymmetric [3+2] cycloaddition reaction. The commercial manufacturing process ultimately developed for 2 is depicted in Scheme 28.7. The process is a convergent, 4+1 step synthesis starting from readily accessible non-complex starting materials. Imine 41 was prepared by condensation of 4-(2-amino-acetylamino)-3-methoxy-benzoic acid ethyl ester hydrochloride (45) with 3,3-dimethylbutyraldehyde (35) in the presence of Et3N in TBME at room temperature. After separation of precipitated Et3N·HCl by filtration, 41 was obtained by crystallization from n-heptane in 94% yield and excellent purity (99.9 area-%). Z-stilbene 34 was prepared by a Knoevenagel condensation of 3-chloro-2-fluoro-benzaldehyde (33) and 2-(4-chloro2-fluoro-phenyl) acetonitrile (32) in EtOH/water in the presence of catalytic amounts of NaOMe. The product 34 precipitated from the reaction mixture and was isolated by filtration. The only observed side product detected in the reaction mixture was the corresponding E-stilbene (up to 6 area-% by HPLC) which did not precipitate from the reaction mixture and was completely removed into the mother liquor during product filtration. This optimized manufacturing process delivered 34 in 91% yield and 99.9 area-% purity containing 99.8 w/w-%, >99.9 area-%, >99.9:0.1 er).

28.2  Case Studies

Finally, a recrystallization from MeCN/water furnished 2 in 55% overall yield (from nitrile 32) as the desired polymorph (form III) required for the tablet manufacturing process. The main characteristics of the developed manufacturing process for 2 include: i) the short and high-yielding convergent route using readily accessible starting materials; ii) the highly selective cycloaddition/isomerization/hydrolysis sequence using a cheap chiral Cu(I)/BINAP catalyst system; and iii) the application of the principles of Green Chemistry such as atom economy, minimization of input materials (PMI = 137), minimization of waste and energy consumption, and the avoidance of undesirable solvents. This process has already been successfully applied to produce more than 1.5 metric tons (MT) of 2.

28.2.3  Danoprevir Danoprevir (RG7227) (3) is an orally administered 15-membered macrocyclic peptidomimetic inhibitor of HCV protease and was originally developed for the treatment of hepatitis C [24]. The compound was discovered by InterMune then licensed to Roche for development and commercialization. In 2013, 3 was licensed to Ascletis Pharma by Roche for co-development in China. Five years further on, 3 was approved for commercialization in China under the trade name Ganovo® for the treatment of hepatitis C. During clinical development, several production campaigns were conducted to manufacture 3 using different synthetic routes and process variants. The initial, highly telescoped four-step manufacturing route is shown in Scheme 28.8. The 15-membered macrocyclic compound features five stereogenic centers and one double bond with syn configuration giving rise to 64 potential stereoisomers for 3. The synthetic strategy was to introduce the stereogenic centers by commencing with the enantiopure vinylcyclopropane 47, Boc-L-hydroxyproline 48 and the nonenoic acid 51. The tripeptide 52 was formed by three highly selective and efficient peptide coupling steps. Subsequently the 4-fluoroisoindoline building block 53 was introduced to create the tripeptide carbamate 54 which was then cyclized by ring-closing metathesis (RCM) employing Zhan’s pentacoordinated ruthenium catalyst 55 [25] to give the macrocycle 56. The two final stages to conclude the synthesis sequence involved introducing the cyclopropane sulfonamide building block 58 and isolation of 3 as the sodium salt. Using the synthetic pathway shown in Scheme 28.8, more than 200 kg of 3 were manufactured to supply the first clinical studies. However, several major downsides needed to be overcome for the manufacture of a projected market demand of several metric tons (MT). First, the peptide coupling sequence was conducted using potentially hazardous, environmentally unsustainable, and relatively expensive reagents (EDC, HOBT, and TBTU). Furthermore, substrate 54 used for the RCM was not crystallized but employed as a solution in toluene. Impurities and reaction byproducts originating from the carbamate formation had to be removed by extractive work-up and on production scale. Additionally, 54 was not consistently obtained in the desired quality, which had an impact on the performance of the subsequent key RCM step. In particular, impurity 60, formed from 4-fluoroisoindoline (53) and CDI (Figure 28.2), had to be well controlled as, in the RCM step, it triggered epimerization at the β-position to the ester group of 54, generating impurities of type A and B shown in Figure 28.2 and thus affecting the purity and yield of 56 [26]. The most significant limitation for a commercial process, however, was the high dilution under which the RCM had to be conducted. To attain a reasonable yield and purity, the substrate concentration in this reaction was limited to a maximum of 1 w/w-%. At higher concentration, the amount of various dimeric impurities (Figure 28.3), as well as unspecified oligomers, significantly increased, with marked negative impact on yield and product purity. These dimeric impurities had

623

624

28  Sustainable Drug Substance Processes Enabled by Catalysis

Scheme 28.8  Initial manufacturing route for 3.

28.2  Case Studies

Figure 28.2  Side product 60 (originated from 58 with CDI) furnishing type A and B by-products.

Figure 28.3  Dimers formed in ring-closing metathesis (RCM).

also the potential to be carried over to 3 if not appropriately controlled. Another downside was the rather high catalyst load (S/C 135). A reduction was desirable in order to reduce cost, to facilitate the control of elemental impurities in 3 as well as to reduce transition metal waste. As a first step to address these flaws to the existing manufacturing process, the peptide coupling sequence was rearranged and further developed (Scheme 28.10). By introducing the 4-fluoroisoindoline (53) one synthetic step earlier, the dipeptide carbamate 62 could be obtained as a new isolated intermediate in high quality (98 w/w-%) by crystallization [27]. Furthermore, the undesired coupling reagents EDC/HOBT and TBTU could be eliminated by using pivaloyl chloride to activate 48 and 51 by forming the mixed anhydrides. As such, the same coupling methodology could be applied for the synthesis of both dipeptide 49 and tripeptide carbamate 54 [27]. The latter was still processed further as a solution, but the quality of this substrate no longer contained impurities that induced epimerization in the RCM. Having an improved quality of the RCM substrate 54 in hand, the next stage was to optimize the RCM reaction. The high catalyst load in this step could be reduced by slowly dosing the dissolved catalyst 55 to the reaction mixture at reaction temperature. In this way, the catalyst loading could be improved from S/C 135 to 750. The most pressing problem remained the low concentration in the RCM step, which could be overcome by introducing the structurally modified RCM substrate 63 (Scheme 28.10). Scientists at Boehringer Ingelheim Pharmaceuticals had demonstrated that the concentration in similar macrocyclic RCM reactions could be substantially increased by selectively introducing a protecting group at the vinylcyclopropane amide nitrogen of the diene substrate [28]. Based on these findings, a screening of various protecting groups was conducted and the benzoyl group was selected as the protecting group of choice. The benzoyl group could be selectively introduced by deprotonaton of

625

626

28  Sustainable Drug Substance Processes Enabled by Catalysis

Scheme 28.9  Synthesis of ring-closing metathesis (RCM) catalyst 64.

tripeptide carbamate 54 with lithium t-butoxide in the presence of benzoyl chloride and later easily removed during the saponification of ester 65. The modified substrate for the RCM allowed this reaction to be conducted at a concentration of 10 vs 1 w/w-%, significantly improving the overall efficiency of the manufacturing process [29]. In addition to these improvements, to improve the IP-position, the Roche proprietary hexacoordinated ruthenium RCM catalyst 64 (Scheme 28.9) was developed and manufactured in kg quantities [30]. These individual improvements were implemented into a telescoped process starting from dipeptide carbamate 62 and furnishing the macrocyclic carboxylic acid 57 in a sequence of six chemical transformations (Scheme 28.10). The development of the final sequence was straightforward. A telescoped process was developed to avoid the isolation of 59, thus further improving manufacturing efficiency [31]. In addition, of note was the development of a flow process for the preparation of cyclopropane sulfonamide building block 58 which facilitated the manufacture of 0.5 MT in excellent quality (100 w/w-%) [32]. The economic and ecologic rewards derived from the described improvements were realized already during the development phase. The new process was implemented for the manufacture of 1 MT of 3 destined for late stage clinical development. Compared to the original process (Scheme 28.8) the overall yield was increased from 27–29% to 38–45%. Hazardous and not inherently biodegradable reagents EDC, TBTU, and HOBt were replaced by employing pivaloyl chloride instead and the consumption of a total of 2.4 MT of these reagents was avoided. For the RCM step, the manufacturing productivity was increased by a factor of 10. As a consequence, the consumption of 273 MT of toluene was averted. Overall, process improvements resulted in a decrease of the PMI factor from 645 to 341. The process was later transferred to Ascletis Pharma to enable commercial manufacturing of 3.

28.2.4  Ipatasertib Ipatasertib (RG7440) (4) is a potent small molecule Akt kinase inhibitor currently being tested in phase III clinical trials for the treatment of metastatic castration-resistant prostate cancer and triple negative metastatic breast cancer [33]. Ipatasertib (4) is a complex molecule with three stereocenters which were assembled in a convergent, 7+1 step synthesis from four starting materials: acid 68, rac-71, formamidine acetate (77), and N-Boc-piperazine (78) (Scheme 28.11, [34, 35]). The three stereocenters of 4 are constructed by enzyme kinetic resolution using a nitrilase, a diastereoselective ketoreductase (KRED) reduction, and metal catalyzed asymmetric hydrogenation (AH). The Discovery Chemistry route utilized (R)-pulegone (chiral pool approach) to establish the stereocenter in 74 (eleven steps) [36]. To shorten the sequence, a scalable process relying on enzymatic ester hydrolysis was introduced, generating 74 in nine steps [37]. Thereby, the lipase applied at 100 kg

28.2  Case Studies

Scheme 28.10  Optimized manufacturing route for 3.

627

628

28  Sustainable Drug Substance Processes Enabled by Catalysis

reaction scale required a high loading of 12 w/w-% enzyme powder (AYS Amano 30G), reflecting the low enzyme activity. The high enzyme loading led to a tedious and time consuming isolation. The final commercial resolution process embarked on a nitrilase at low enzyme loading to deliver (R)-71 from its racemate (Scheme 28.11). The undesired (S)-enantiomer was hydrolyzed to the corresponding carboxylic acid 80 (Scheme 28.12) and depleted in the aqueous extractions upon work-up. This commercial synthesis enabled the production of 74 in only four steps. On a laboratory scale, two equally efficient processes (Scheme 28.12) were optimized that relied on either lipase CRL III from Candida rugosa or a mutant nitrilase from Acidovorax facilis. These enzymes were discovered by screening more than 250 hydrolases (lipases, esterases, and proteases), and close to 100 nitrilases, respectively. The lipase process formed a mixture of acids as side products, and required pH control via base addition, yet displayed a very high enzyme selectivity (E >100) despite only a moderate activity. The nitrilase forms the (S)-acid and ammonia as ­products and therefore the reaction is pH neutral. The nitrilase displays a good selectivity (E ~50) and a good activity. The remaining (R)-nitrile 71 is the resolution target and as the undesired enantiomer is hydrolyzed. Therewith, the chiral purity of 71 can be controlled precisely by the conversion degree

Scheme 28.11  Optimized manufacturing route for 4. Adapted from Refs [34a] and [35].

28.2  Case Studies

to ensure the depletion of the undesired enantiomer to the specified level. Both enzymes successfully resolved racemic 71 at 20 w/w-% substrate loading with yields in the range of 40–47%. The process selection was finally based on the straightforward heterologous expression of the nitrilase in E. coli, whereas the lipase necessitates a yeast, Pichia, as the production host. The preferred enzyme production in E. coli significantly shortens the fermentation development timelines and facilitates its optimization via enzyme engineering, especially for the applied rational evolution strategy. In addition to the obvious targets (such as increased activity, stability, and selectivity), a stable liquid formulation of the nitrilase was targeted to simplify the process and reduce the sensitation risk of handling enzymes as powders. The chosen liquid formulation was based on glycerol which had an inhibitory effect on the activity rate of nitrilase I under the process conditions (see Table 28.1), and added a further challenge to the enzyme engineering. To select the starting template for enzyme evolution, an existing mutant library of the nitrilase I was screened with regard to initial activity and process stability. Subsequently, the evolvability and positive acceptance of amino acid exchanges was investigated by the saturation mutagenesis of

Scheme 28.12  Alternative resolution processes of rac-71. Adapted from Refs [34a] and [35]. Table 28.1  Nitrilase engineering: comparison of first lead with optimized enzyme. Process Details

Nitrilase I

Nitrilase VI

Volumetric activity [Ubenzonitrile /ml] 1

2.5

5.2

2

FIOP (use-test activity)

1

5

er (71) use-test (43 hours)

99:1

99.4:0.6

er (71) stability test (120 hours)

99.3:0.7

99.6:0.4

E

~60

>100

Inhibition effect of glycerol 3

−6.20%

−0.60%

Thermostability [Tm50%]

55 °C

52 °C

Expression level factor

1

1

−1

1)  U: unit = 1 µmol min . 2)  FIOP: fold improvement over parent. 3)  % of reduced activity in the presence of 2.5 v/w-% glycerol final concentration in use-test.

629

630

28  Sustainable Drug Substance Processes Enabled by Catalysis

seven amino acid positions which were selected on the basis of substrate docking and modeling. One mutant, nitrilase II, possessed increased process stability and all the activity hits were variants thereof. Thus, nitrilase II was chosen as the template for the following mutagenesis studies. In total, 2,647 variants from five mutant libraries were screened and 54 hit-variants were characterized in detail with respect to their stability under the process conditions (lower enzyme loadings at extended reaction time), thermostability, initial activity with benzonitrile as substrate, process related activity, enzyme selectivity, inhibitionary effect of glycerol, and protein expression. The final tailor-made nitrilase VI exhibits a higher activity, an increased selectivity, minor glycerol inhibition at 2.5 w/w-% glycerol concentration in the ultimate process, a slightly inferior thermostability, a high process stability, and the processability as a stable liquid formulation (Table 28.1). The enzyme engineering was performed in parallel to process development, which increased significantly the complexity of both tasks. Additional key achievements on this step were the depletion and control of residual enzyme. For this purpose, the nitrilase resolution, conducted in an aqueous buffer (pH 8.7 to 9.4), and affording crude 71 with 99.4:0.6 er, was acidified with 30% sulfuric acid (pH 1.6 to 2.2) to denature the enzyme. Subsequent TBME addition enabled the precipitated enzyme to be filtered efficiently as protein flakes. Product extraction with TBME and azeotropic water removal (250) was screened to identify a more efficient enzyme. The best hit with regard to selectivity and activity was clearly still the implemented KRED-NADP-101, a KRED not tolerating i-PrOH as hydride source. In a subsequent process optimization, a slurry to slurry biphasic process was introduced, enabling at least a four-fold increased space-time-yield and a three-fold reduced KRED loading, all at an increased substrate concentration of 10 w/w-%. Surprisingly, the GDH and NADP loadings could not be reduced. In addition, the color formation could be completely precluded by changing the cosolvent from DMSO or PEG to TMBE. This significantly improved the homogeneity of the stirred reaction mixture and reduced the solubility of 74, which was key to avoid the formation of color. The process optimization is summarized in Table 28.2. In summary, the KRED process was successfully optimized with regard to efficiency, robustness, and coloration of intermediate 69, but the cofactor recycling aspect remained unsatisfactory. To enable a coupled enzyme approach, the screening efforts were extended. A commercial panel of engineered KREDs (>350) included one highly active enzyme with good diastereoselectivity (~95:5% dr) that accepted high concentrations of i-PrOH. After initial optimization experiments,

Table 28.2  Summary table of ketoreductase (KRED) processes including enzyme engineering.

Ketoreductase

KRED-NADP-101

1st round P1B02

2nd round P1B06

3rd round P1F01

4th round P1F01

5th (Final) round CDX-040

Diastereoselectivity [dr]

>99.9:0.1

>99.9:0.1

98:2

99.6:0.4

>99.9:0.1

>99.9:0.1

>99.9:0.1

Substrate loading ≤5 [w/w-%]

10

20

20

20

10

10

Time

≥2 d

1d

1d

1d

1d

2.5 d

98 area-% by HPLC) otherwise would have been acceptable. Despite the excellent enzyme process performance, the color formation required the reduction of ketone 69 to be performed as a slurry to slurry process. This was realized at 20 °C with 8 v/w-% i-PrOH compromising the process performance and requiring a reduced substrate concentration (10%), an increased enzyme loading (S/E 33), and an extended reaction time (60 hours) to achieve the high specified conversion (99% conv. after 12 hours, 60 °C, 18 bar H2), delivering crude 67 with constant high er (Table 28.3, entry 3). Under otherwise identical conditions but in the absence of NaBr, the conversion stalled at 29% conversion (Table 28.3, entry 4). To render the process more robust on technical scales, a somewhat higher catalyst loading of S/C 4,000 at 15 bar H2 was employed. Residual amounts of ruthenium were efficiently depleted in the subsequent crystallization unit operation (99.9 area-%). The product isolation process was optimized next. In particular, the low bulk density (~0.1 kg/L) of 67 caused a low throughput in the drying step and demanded large storage capacities. An adapted crystallization procedure was thus developed whereby seed crystals were generated in situ by the addition of 0.2 eq of NaOEt (sufficient to induce nucleation), followed by aging of the formed crystals for two hours, and then the addition of 1 eq of NaOEt to complete the crystallization. With this procedure, a total of 5.5 MT of 67 with a 3.5-fold increased bulk density were manufactured in 93% yield, >99.9 w/w-% assay, >99.9 area-% purity, and >99.9:0.1 er [35]. The final three step ­synthesis of 67 (Scheme 28.15) (~74% overall yield, PMI 100) compares favorably with the first generation hydrogenation variant (PMI 150), and even more noticeably with the initial, chiral. auxiliary based process (PMI 575), leading overall to significantly reduced waste formation and increased throughput [39]. The penultimate step required the coupling of the acid salt 67 with piperazine 69 (Scheme 28.11). Activation of 67 was accomplished with pivaloyl chloride to form a mixed anhydride. In a separate reactor, the Boc-protecting group of intermediate 69 was removed with HCl in n-PrOH. The excess of HCl was neutralized with Et3N and the reaction mixture containing the mixed anhydride of 67 was added. The work-up after reaction completion was carried out by exchanging the solvent to toluene by distillation followed by extractions with aqueous HCl, aqueous NaOH and water to

Scheme 28.15  Improved synthesis of 67.

28.3  Conclusions

remove acidic and basic impurities such as unreacted 67 and 75. Product 76 was crystallized by the addition of n-heptane to the toluene solution. In the final chemical step, the Boc-protecting group of 76 was removed with HCl in n-PrOH. The pH was adjusted to 5.5 with aqueous NaOH to acquire the desired mono-HCl salt. After concentration and removal of water by distillation, precipitated NaCl was removed by filtration and the product was isolated by distillative exchange of n-PrOH with EtOAc, from which the product precipitates. Lastly, the solid state properties (e.g. particle size distribution, amorphous form) were adjusted by spray drying of 4 as an aqueous solution. In the late stage development phase, prior to process validation, the overall process was significantly improved regarding robustness, efficiency, and sustainability. In the last three manufacturing campaigns up to validation, the PMI had been reduced by a factor of five (from 1,393 to 269), mainly driven by significant solvent reductions and a greater than three-fold improved overall yield (from 3 to 10%).

28.3  Conclusions The reported case studies here demonstrate strikingly what impact catalysis can provide in constructing complex DS molecules by enabling new routes or shortcutting existing ones with markedly improved efficiencies. Catalysis has played thereby a key role in connecting building blocks, introducing functional groups and chirality in 1–4, and their associated intermediates with excellent selectivities, excellent throughputs, and high catalyst turnover frequencies. To achieve this, in-depth catalyst research and process development work was fundamental to determine tailor-made catalyst solutions for each one of the four case study molecules. Conspicuously, the presented show cases also provided evidence that a most productive catalytic step per se will fail its application in DS manufacturing if or when the substrate and catalyst accessibility, the downstream chemistry, and/or the catalyst depletion is too cumbersome. This is particularly the case when the entire process is inefficiently long or requires tedious (chromatographic) purifications and/or unstainable reagents, solvents, and reaction conditions. Clearly, the key to be able to offer spot-on catalytic solutions is to have ready access to well-diversified enzyme / metal catalyst libraries, state-of-the-art catalyst prediction tools, enzyme engineering, metal catalyst synthesis / enzyme fermentation, optimization and scale-up expertise, and adequate facilities on hand. Obviously, the delivery of a high performing and most sustainable DS process requires dedicated process research and development work early on and the steady drive to search and implement process optimizations along all development phases. These preliminary efforts result in improved PMI values, reduced energy consumptions, and the elimination of unsustainable solvents and reagents together with DS cost and lead time reductions and a sustainable use of natural resources. Accordingly, patients and likewise the entire society benefits from the outcome. Two cases, idasanutlin 2 [23b] and ipatasertib 4 [35], were recognized by the Swiss Chemical Society, which awarded the associated process development teams with the Sandmeyer Award in 2017 and 2020 for the development of economical, scalable, and sustainable processes. In a larger context made transparent via the Dow Jones Sustainability Indices (DJSI), F. Hoffmann-La Roche Ltd. was ranked for the thirteenth year running as one of the most sustainable pharmaceutical companies in terms of economic, social, and environmental performance worldwide. To maintain this standard, Roche stays fully committed to further fortify the implementation of catalysis, particularly biocatalysis in water and non-noble metal catalysis, especially with the use of recyclable catalysts

635

636

28  Sustainable Drug Substance Processes Enabled by Catalysis

and other resource saving methodologies and technologies. As further promotion of this, Roche has also created a fund to support research projects with academia and industry targeted to invent products such as new catalysts and catalyst systems with the potential for break-through performance, or to promote so far unprecedented transformations all aimed to enhance the sustainability of our current and future DS processes even further.

References 1 (a) Sheldon, R.A. (2018). ACS Sustain. Chem. Eng. 6: 32–34. (b) Jimenez-Gonzalez, C. et al. (2011). Org. Process Res. Dev. 15: 912-917. 2 Anastas, P.T. and Warner, J.C. (1998). Green Chemistry: Theory and Practice. New York: Oxford University Press. 3 The Roche solvent selection guide (based on https://www.acs.org/greenchemistry/researchinnovation/tools-for-green-chemistry/solvent-selection-tool.html from ACS), classifies solvents in four different groups as follows based on a composite score including categories such as safety, health, and environment: recommended, usable, avoid, and high uncertain. 4 Wuitschik, G. et al. (2022). Curr. Res. Green Sustain. Chem. 5: 100293. 5 Mohr, P. et al. (2009). Bioorg. Med. Chem. Lett. 19: 2468–2473. 6 Fürnsinn, C. et al. (1999). Br. J. Pharmacol. 128: 1141–1148. 7 For clarity, herein a chemical transformation or a cascade thereof (a telescoped process) counts as a step from one isolated and characterized intermediate to the next, eventually reaching the final product. Intermediates that are not isolated, either as a crude or in a purified form, but are kept in solution for further processing, are depicted in brackets. Purification methods are reported for intermediates or final products that have been isolated by crystallization (cryst), silica gel chromatography (chrom), distillation (dist), or supercritical fluid chromatography (SFC). Otherwise, the step products were further converted from their crude form. 8 (a) Goehring, W., Hoffmann, U., Scalone, M. et al. (2006). PCT International Publication. Patent application WO2005000844. (b) Junghans, B., Scalone, M. and Zeibig, T.A. (2002). US Patent US6482958. 9 Hidai, M. et al. (1989). Tetrahedron Lett. 30: 95–98. 10 As alternatives, the asymmetric reduction of 2-oxo-propanoic acid derivatives to furnish a chiral 2-hydroxy-propanoic acid derivative as well as asymmetric epoxidation approaches were not considered further, as selective etherifications were reported by our Discovery Chemistry colleagues to fail for heavily decorated, benzothiophene type molecules such as 1. 11 Ratovelomanana-Vidal, V. et al. (2010). Org. Lett. 12: 3788–3791. 12 Takaya, H., Noyori, R. et al. (1987). J. Org. Chem. 52: 3176–3178. 13 (a) Ohata, T. et al. (1996). J. Org. Chem. 61: 5510–5516. (b) Sannicolo, F. et al. (2000). J. Org. Chem. 65: 2043–2047. (c) Chan, A.S.C. et al. (2000). J. Am. Chem. Soc. 122: 11513–115144. (d) Yamagishi, T. (1996). Tetrahedron: Asymmetry 7: 3339–3342. 14 Püntener, K., and Scalone, M. (2005). PCT International Publication WO2005030764. 15 (a) Heiser, B. et al. (1991). Tetrahedron Asymmetry 2: 51–62. (b) Chan, A.S.C. et al. (1994). Inorg. Chim. Acta 223: 165–167. 16 (a) Schmid, R. et al. (1988). Helv. Chim. Acta 71: 897–929. (b) Schmid, R. et al. (1991). Helv. Chim. Acta 74: 370–389. (c) Schmid, R. et al. (1996). Pure Appl. Chem. 68: 131–138. (d) Schmid, R. and

References

Scalone M. (2015). (R)- and (S)- 2,2ʹ-Bis(diphenylphosphino)-6,6ʹ-dimethoxy–1,1ʹ-biphenyl. In: e-EROS Encyclopedia of Reagents for Organic Synthesis (eds. L. Paquette, P. Fuchs, D. Crich, and P. Wipf). Hoboken: John Wiley and Sons. 17 (a) Maligres, P.E., Krska, S.W. et al. (2004). Org. Lett. 6: 3147–3150. (b) Zhou, Q.-L. et al. (2005). Angew. Chem. Int. Ed. 44: 1118–1121. (c) Zhou, Q.-L. et al. (2006). Adv. Synth. Catal. 348: 1271–1276. (d) Chan, A.S.C. et al. (2007). Adv. Synth. Catal. 349: 517–520. (e) Houpis, I.N. et al. (2005). Org. Lett. 7: 1947– 1950. (f) Kim, T.-J. et al. (2005). Organometallics 24: 4824–4831. (g) Minnaard, A.J. et al. (2005). Angew. Chem. Int. Ed. 44: 4209–4212. (h) Chen, W. et al. (2007). Angew. Chem. Int. Ed. 46: 4141–4144. (i) Scrivanti A. et al. (2006). Tetrahedron Lett. 47: 9261–9265. 18 Zhou, Q.-L. et al. (2008). J. Am. Chem. Soc. 130: 8584–8585. 19 Püntener K. and Scalone M. (2010). PCT International Publication WO2010108861. 20 (a) Vu, B. et al. (2013). ACS Med. Chem. Lett. 4: 466–469. (b) Ray-Coquard, I. et al. (2012). Lancet Oncol. 13: 1133−1140. (c) Ray-Coquard, I. et al. (2015). Curr. Med. Chem. 22: 618−626. 21 (a) Wang, C.-J. et al. (2020). Acc. Chem. Res. 53: 1084–1100. (b) Adrio, J. and Carretero, J.C. (2019). Chem. Commun. 55: 11979−11991. (c) Zhang, J. et al. (2016). Angew. Chem. Int. Ed. 55: 6324. (d) Hashimoto, T. and Maruoka, V. (2015). Chem. Rev. 115: 5366–5412. (e) Adrio, J. and Carretero, J. C. (2014). Chem. Commun. 50: 12434–12446. (f) Garner, P. et al. (2007). Tetrahedron Lett. 48: 3867– 3870. (g) Pandey, G. et al. (2006). Chem. Rev. 106: 4484–4517. (h) Savic V. et al. (2005). Tetrahedron: Asymmetry 16(12): 2047–2061. 22 (a) Shu, L. et al. (2016). Org. Process Res. Dev. 20: 2050–2056. (b) Fishlock, D., Gu C., Shu, L. et al. (2014). PCT International Publication WO2014128094. 23 (a) Pankaj, R. et al. (2016). Org. Process Res. Dev. 20: 2057–2066. (b) Fishlock, D., Diodone, R., Hildbrand, S. et al. (2018). Chimia 72: 492–500. 24 Jiang, Y. et al. (2014). J. Med Chem. 57: 1753–1769. 25 Zhan, Z.-Y.J. (2007). PCT International Publication Patent WO2007003135. 26 Wei, X., Farina, V. et al. (2006). J. Org. Chem. 71: 8864–8875. 27 Gantz, F. and Stahr, H. (2008). PCT International Publication WO2008128921. 28 (a) Shu, C. et al. (2008). Org. Lett. 10: 1303. (b) Farina, V., Senanayake, C.H. et al. (2009). Org. Process Res. Dev. 13: 250–254. 29 Scalone, M. and Stahr, H. (2010). PCT International Publication WO2010015545. 30 (a) Püntener, K. and Scalone, M. (2009). PCT International Publication WO2009124853. (b) Doppiu, A., Karch, R., Püntener, K. et al. (2010). PCT International Publication WO2010127964. 31 Hildbrand, S., Püntener, K. and Scalone, M. (2009). PCT International Publication WO2009080542. 32 Hildbrand, S. (2009). PCT International Publication WO2009053281. 33 (a) Manning, B.D. et al. (2007). Cell 129: 1261–1274. (b) Blake J. F. et al. (2012). J. Med. Chem. 55: 8110–8127. 34 (a) Gosselin, F. et al. (2017). Org. Lett. 19: 4806–4809. (b) Gosselin, F., Iding, H., Reents, R., Scalone, M. (2015). PCT International Publication W2015073739. 35 Schuster, A. et al. (2021). Chimia 75: 605–613. 36 Lane, J.W., Remarchuk, T. et al. (2014). Org. Process Res. Dev. 18: 1641–1651. 37 Remarchuk, T., Spencer, K.L. et al. (2014). Org. Process Res. Dev. 18: 1652–1666. 38 Huisman, G.W. et al. (2007). Nat Biotechnol. 25: 338–344. 39 Remarchuk, T. et al. (2014). Org. Process Res. Dev. 18: 135–141. 40 Bachmann, S. et al. (2013). Org. Process Res. Dev. 17: 1451–1457. and references cited therein.

637

639

29 Supported Chiral Organocatalysts for Accessing Fine Chemicals Ana C. Amorim1 and Anthony J. Burke2,3,4 1 Centro de Química de Coimbra, Institute of Molecular Sciences, Departamento de Química, Faculdade de Ciências e Tecnologia, Universidade de Coimbra, Coimbra, Portugal 2 Faculty of Pharmacy, Universidade de Coimbra, Pólo das Ciências da Saúde, Azinhaga de Santa Comba, Coimbra, Portugal 3 Chemistry and Biochemistry Department, School of Science and Technology, University of Évora, Rua Romão Ramalho 59, Évora, Portugal 4 LAQV-REQUIMTE, University of Évora, Institute for Research and Advanced Studies, Rua Romão Ramalho, 59, Évora, Portugal

29.1  Introduction Organocatalysis’ position as the third pillar of catalysis, along with metal based catalysis and biocatalysis was cemented with the award of the Nobel Prize in chemistry on the 6th of October 2021 to Banjamin List and David MacMillan for their work in the area of asymmetric organocatalysis [1]. Over the last 20 years since the discovery of the proline based catalysts of List and the imidazolidinone catalyst of Macmillan (as well as a host of other types of organocatalysts), organocatalysts have been exploited with tremendous success in the preparation of a myriad of important, and challenging target compounds such as strychnine (a poison), paroxetine (for treating depression), oseltamivir (for respiratory infections), (−)-tashiromine (a naturally occurring indolizidine alkaloid), and aliskiren (a renin inhibitor). Prior to the seminal work of List and MacMillan in 2000, there was very little activity in the area. However, the pioneering work from two teams (Hajos and Parrish; and Eder, Sauer, and Wiechert), who independently showed that a key triketone substrate could be transformed to the Wieland-Miescher ketone using L-proline (Scheme 29.1) and used to access several natural products as well as steroids, should not be undervalued as this seminal work reported in the early 70s remains a watershed development that eventually led to the discoveries that received the Nobel Prize award in 2021 [2]. Despite the fact that many types of organocatalysts are known at the current time [3], and despite the fact that organocatalysis brings so many benefits to the synthesis lab (such as stability in air, lack of toxicity, and renewability), they have one major downside in that they require greater loading of the catalyst compared to those reactions catalyzed by metals. This limitation becomes all the more inconvenient when the organocatalyst is more expensive or very difficult to make, and recovery via chromatography or other purification techniques becomes difficult, or even when the reaction is run at a large scale. As a solution to this problem, many types of organocatalysts have now been immobilized to various types of support that allows their easy recovery and recycling. Immobilization has also been employed to improve catalyst stability. In this chapter we will look Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 3, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

640

29  Supported Chiral Organocatalysts for Accessing Fine Chemicals

O

O

O L-Proline (0.3 mol%), DMF, 20 h

O N H

CO2H

O Wieland-Miescher ketone 87% 94.8% ee

Natural Products and steroids

L-Proline

Scheme 29.1  The L-Proline catalyzed synthesis of the Wieland-Miescher ketone as reported by Hajos and Parrish [2] (this was found to be more efficient than with the conditions used by Wiechert et al.).

at some of the organocatalysts that have been successfully immobilized and applied for the synthesis of key targets over the last 10 years. Of course, one of the main driving forces for the immobilization of organocatalysts to solid-supports has been their utilization as part of the catalytic reactor system in flow-chemistry set-ups, which is becoming a standard tool in many pharmaceutical labs. This review is not exhaustive but will give the reader a general view on what has been achieved during this period. The reader is also encouraged to consult a number of informative reviews on the topic [3, 4]. In the very insightful review by Cozzi in 2006, he stated that only catalysts that can promote multi-reactions, and are highly active, should be considered for immobilization [4–7]. Some of the issues that we will address, through relevant examples, will be (i) the type of support; (ii) the type of linker; (iii) catalyst loading on the support; and, of course, (iv) the reaction conditions. As a final point, something that was not stated in the interviews and articles that appeared after the awarding of the Nobel Prize in October 2021, and relevant in the context of this review, is the fact that organocatalysis is ideal for immobilization to various supports, and more importantly for application in continuous flow systems.

29.2  Organocatalyst Immobilizations As mentioned previously, the ability to conduct organocatalytic reactions using immobilized catalysts brings several key advantages, like easy reaction recovery and reusage, greater activity and stability in some cases compared to their homogenous equivalents, diminished product contamination by the catalyst, and easy adaptation to continuous flow systems. To date many types of supports have been used, that range from polymers to silica and aluminosilicate supports to magnetic nanoparticles. In this section we will look at the application and effectiveness of several of the most active organocatalysts that were used in the last 20 years.

29.2.1  Proline Immobilizations Considering proline’s prowess as one of the most privileged organocatalysts available, naturally it was one of the first catalysts to be immobilized to solid supports, and several examples of their use in organocatalytic reactions are known [3–6]. Some examples of their use in the last 10 years are given in this section. In 2012, Pericàs et al. reported asymmetric aldol reactions using a polystyrene (PS) supported proline catalyst that afforded interesting cyclic aldol products in high yields and with excellent

29.2  Organocatalyst Immobilizations

stereoseletivities (Scheme 29.2) [8]. The catalyst was derived from hydroxyproline, and incorporated with a 1,2,3-triazole linker. 10 mol% loading was found to work well. After establishing the best reaction conditions, the team proceeded with a study of the reaction scope. Best results were obtained using the more electron poor aldehydes. These workers also showed that for the prototype reaction with benzaldehyde and cyclohexanone, the catalyst could be recycled up to seven times with little or no drop in the yields, or the stereoselectivities. Due to the excellent results obtained, the prototype reaction was tested with this catalyst (0.54 mmol g−1) in a continuous flow system, using a vertically oriented packed-bed reactor at a flow rate of 25 μL min−1 (residence time = approximately 26 minutes). The conversion dropped off after 30 minutes but the diasteroselectivity remained constant up to 45 minutes. Gratifyingly the set-up afforded 4.87 g of the aldol product with a diasteroselectivity of 92% de, and an ee of 97% in a 30 hour run. An interesting approach was reported by Pothanagandhi and Vijayakrishna, who developed both imidazolium based cross-linked poly(ionic liquids) and polyeletrolytic resins (PILs) bearing proline (as its conjugate base, which was non-covalently attached to the resin) and used them in a variety of reactions that included both Michael additions and Bayliss-Hillman reactions [9]. The imidazolium based achiral PIL resins were prepared by copolymerizing 1-vinyl-3-ethylimidazolium bromide with a cross-linker using 2-ethoxythiocarbonyl-sulfanylpropanoic acid ethyl ester using the reversible-addition-fragmentation-chain-transfer (RAFT) technique. The results were very satisfactory, in the case of the Michael addition between dimethyl malonate and chalcone when the reaction was conducted under neat conditions with 10 mol% catalyst (as poly(ViEIm)(dvb)-pro, where ViEIm = 1-vinyl-3-ethylimidazolium and dvb = divinylbenzene) at room temperature for 24 hours gave the product with a yield of 95% (there was no mention of the enantioselectivity, although both D and L-proline gave the same yield); and in the case of the Baylis Hillman reaction between methyl vinyl ketone and p-nitrobenzaldehyde under neat conditions with 10 mol% of the same catalyst at room temperature gave both entantiopodes with a yield of 94%, but unfortunately with ees of 9% (L-Pr) and 7%, respectively. There was no mention of catalyst reuse.

O

CHO

O

Catalyst (10 mol%), DMF/water, r.t., 5-24 h

+

OH

X

n

N

N N

O

N H

n

n = 1,2

OH

Catalyst

OH

O

OH

O

OH

O

OH

91%, 96:4 (anti), 98% ee OH

O

MeO 96%, 97:3 (anti), 98% ee

O

MeO

O2N 74%, 96:4 (anti), 98% ee

CO2H

O

41%, 98:2 (anti), >99% ee OH

28%, 98:2 (anti), >99% ee

O

MeO 96%, 97:3 (anti), 98% ee

74%, 96:4 (anti), 98% ee

Scheme 29.2  The polystyrene supported L-proline catalyzed aldol reaction as reported by Pericàs et al. [8].

641

642

29  Supported Chiral Organocatalysts for Accessing Fine Chemicals

H N N H

(a)

Br N

N

Si O O O

MWCNT xxxxxxxxxxxxxxx

O

O O HN

O

H N

O O

= Fe3O4 SiO2 surface

(b)

O

N H2

CONHSO2Me

CF3COO

Figure 29.1  (a) The magnetic nanoparticle supported L-proline catalyst developed by Kong et al [11]. (b) The multi-walled carbon nanotube (MWNT) supported proline catalysts developed by Chronopoulos et al. [12].

Proline has also been successfully immobilized to magnetic nanoparticles. Some good examples are included in the 2015 review by Angamuthu and Tai [10]. One interesting example was the report by Kong et al. on the application of an L-proline supported ionic liquid (IL)-modified magnetic nanoparticles for recyclable use in direct aldol reactions in water (Figure 29.1a) [11]. The catalyst consisted of the proline unit grafted to a magnetic ferrite (Fe3O4) core via an imidazolium linker. The catalyst was active for the reactions in water without the need for organic solvents. It was suggested that the imidazolium unit facilitated the entry of hydrophobic reactants to active sites in water and stabilized the enamine intermediate. Cyclohexanone, cyclopentanone, and benzaldehydes were used, with yields of up to 96%, diasterometric excessees of up to 99:1, and ees of up to 89%, with just 10 mol% of catalyst. Furthermore, the catalyst could be recycled four times with marginal drops in the yield and no deceases in either the diastereo- or the enantioselectivity, using cyclohexanone and 2-nitrobenzaldehyde as the test system (12 hours, 30 °C). In 2015 Chronopoulos et al. described the application of proline tethered units onto multi-walled carbon nanotubes (MWCNTs) and their application in the standard aldol reaction with acetone and a benzaldehyde derivative [12]. The catalyst was prepared using standard amide coupling procedures (Figure 29.1b), and then neutralized with aqueous NaHCO3 before being screened at a loading of 20 mol% in the bench-mark aldol reaction between acetone and p-nitrobenzaldehyde (water at room temperature, 72 hours), to give the aldol in a yield of 94%. It must be noted that the yields were very low in organic solvents. The catalyst was recycled four times and there was a steady drop off in the yield to 64% after the fifth cycle was completed. The enantioselectivity was low (i.e. 16% ee in the first cycle dropping off to 13% ee for the fifth cycle). The authors did not discuss the reasons for these low enantioselectivities, but this may have been due to back-ground catalysis from a non-proline sidechain on the MWCNT surface. Interestingly, the JørgensenHayashi diphenylprolinol unit (see the following section) was also attached to the nano-material, and also gave a high yield (84% after 96 hours), with a low enantioselectivity (14% ee). In 2013, Pericàs et al. reported a fluorous tagged proline catalyst that was used in fluorous phase catalytic aldol reactions, and as mooted by the authors to possess purported aldolase behaviour [13]. The concept of fluorous phase catalysis was pioneered by Horváth, Cornilis, and Curran in the mid-90s as an alternative approach to recycling expensive catalysts, including organocatalysts [13], but in the last decade there appears to have been a decline in its application. The proline was tagged with a C10F21 ponytail using a 1,2,3-triazole linker and initial assessment of the reaction conditions showed that the reaction between acetone and 4-nitrobenzaldehyde in perfluorohexane (C6H14) at 0 oC at a pressure of 1 bar for 16 hours gave the aldol product with 70% yield and 76% ee (Scheme 29.3). The scope of the reaction on the aldehyde component was then examined and gave highly satisfactory results over a large range of electron poor and rich benzaldehyde derivatives, with very good yields and enantioselectivities in the range 68–96% ee. Gratifyingly the catalyst could be recycled five times with only a noticeable change in the yield on the fifth cycle. The enantioselectivity remained constant at 77% ee.

29.2  Organocatalyst Immobilizations R

O

CHO

O

cat (8 mol%)

+

OH

R

C6F14, 16 h, rt 54–81%; 68–96% ee C10F21

N

N N

O Catalyst

N H

CO2H

Scheme 29.3  Fluorous phase aldol catalyzed reactions with an L-Proline tagged catalyst as reported by Pericàs et al [13].

29.2.2  Diphenylprolinol Silyl Ether (Jørgensen-Hayashi Organocatalyst) Immobilizations In 2012, Pericàs et al. reported the application of polystyrene-immobilized diphenylprolinol silyl ether catalysts for the α-amination of aliphatic aldehydes (Scheme 29.4) [14]. A variety of immobilized diphenylprolinol silyl ether catalysts were screened for the α-amination of propanal with dibenzyl azodicarboxylate (DBAD) using 10 mol% catalyst and carboxylic acid co-catalysts and it was found that the tert-butyldimethylsilyl (TBS) catalyst gave the best results (92% yield, 93% ee, with 10 mol% AcOH in CH2Cl2 (DCM) at room temperature for 45 minutes) for the final aminoalcohol product, obtained via a simple reduction of the aminoaldehyde adduct with NaBH4. The scope of

O BnO2C

+

R

N N

CO2Bn

1. cat (2 mol%), AcOH (10 mol%), CH2Cl2 (37-180 min), rt 2. NaBH4, EtOH, 10 min

Et CbzHN

OH

N Cbz

(92%, 93%ee)

CbzHN

N Cbz

N Cbz

OH

CbzHN

(94%, 94%ee)

(90%, 95%ee)

N Cbz

N Cbz

CbzHN

(91%, 90%ee)

N Cbz

N

N O

Catalyst

OH

(89%, 95%ee)

N

OH

(92%, 98%ee)

OH

Bn OH

Cl CbzHN

N Cbz

Pr OH

(CH2)5 CbzHN

R CbzHN

Ph N H

Ph

OTBS

TBS = tert-butyldimethylsilyl

Scheme 29.4  Polystyrene supported prolinol catalysts for the α-amination of aldehydes as reported by Pericàs et al [14].

643

644

29  Supported Chiral Organocatalysts for Accessing Fine Chemicals

the reaction was also investigated, and a screening study of a variety of aldehyde products showed that the reaction was tolerant of a wide range of substrates, as both the yields and ees were high. The catalyst could also be recycled up to six times with little alteration in the yield or the enantioselectivity, demonstrating the potential of this catalyst. Finally, the authors performed the reaction under continuous flow conditions, using a packed bed reactor to house the catalyst, and a flow system comprising 2 pumps. One pump was responsible for pumping the aldehyde as a 1.25 M solution in DCM to a T-mixer at a flow rate of 0.075 mLmin−1, and the other the DBAD (0.25 M, at the same flow rate) also to the T-mixer. The mixture was then passed into the catalytic reactor (containing 300 mg of the immobilized catalyst) at a flow rate of 0.15 mL min−1 (residence time = 6 min) and the effluent was collected at 0 oC, reduced to the aminoalcohol and analysed. The conversion was quantitative for the first six hours, then dropped off to 90% and 87% after the seventh and eighth hours, respectively. Gratifyingly the ee remained constant at approximately 85%. This was interesting work, but it’s funny that these workers did not access the useful aminoalcohol products by breaking the N–N bond. In 2019, Szcześniak et al. reported the application of a key organocatalyzed Michael addition using a prolinol catalyst supported on Wang resin as the key step for the synthesis of the antidepressant active pharmaceutical ingredients (APIs) of (−)-paroxetine (Paxil®, Seroxat®) and the selective serotonin reuptake inhibitor (+)-femoxetine (developed by Ferrosan pharmaceuticals in the 70s) (Scheme 29.5) [15]. After preparing the immobilized catalyst, the reactions with two aldehydes and two nitroolefins were investigated under batch conditions.

O +

R

NO2

cat (15 mol%), CDCl3, rt

R

R

O

R1 syn

R1

NO2

O anti

NO2

R1

F

N O

O

NO2

NO2

O

O

O Catalyst

O

O

N

N O

Ph N H

Ph

OTMS

OMe

(100% conversion; (100% conversion; syn/anti 4.6:1, 93% ee (syn)) synlanti 3.6:1, 89% ee (syn))

H N

N O

O

O

O F (+)-Paroxetine

OMe

(+)-Femoxetine

Scheme 29.5  The key asymmetric step in the synthesis of the antidepressant agents (−)-paroxetine and (+)-femoxetine as reported by Szcześniak et al [15].

29.2  Organocatalyst Immobilizations

The reactions afforded both products in very good yields and with high enantioselectivities for the syn-isomer (Scheme 29.5). The reaction was also conducted under continuous flow conditions, using a packed-bed reactor to immobilize the catalyst. Full conversion and a diasterometric ratio of 3/1 was achieved when the reaction was run for 11 hours with 20 mol% benzoic acid in CHCl3 at a flow rate of 0.005 mL min−1 (operation time = approximately 17 hours). Once the Michael adducts were obtained in sufficient quantities, they were transformed into the anti-depression compounds (−)-paroxetine (35%, 4-steps from the Michael adduct, and 80% ee) and (+)-femoxetine (27%, 5-steps, 75% ee) via a multi-sequential pathway (Scheme 29.5). This also represents a general strategy for the asymmetric synthesis of piperidine scaffolds with trans-stereochemistry at C3 and C4.

29.2.3  Organocatalysts Based on Immobilized Pyrrolidines In 2013 Pericàs et al. (pioneers in the exploitation of continuous flow for organocatalysis, as we see in this chapter) reported the use of a PS supported 1,2,3-triazole linked 3-trifluoromethylsulfonylaminopyrrolidine catalyst for asymmetric anti-Mannich reactions in continuous flow (Scheme 29.6) [16]. After preliminary optimization studies with isovaleraldehyde and N-PMP ethyl glyoxylate,

O

Cat (2 mol%aldehyde, 4 mol%ketone), DMF, 4h, 0 oC (in flow 8-10 oC) O

PMP

R 1

+

CO2Et

R

R1 CO2Et R syn

PMP = p-Methoxyphenyl iPr O

nPent

CO2Et

O

(83%; 92/8 (syn), 97% ee (syn))

O

CO2Et

(75%; 95/5 (syn), 97% ee)

Bn O

CO2Et

(90%; 92/8 (syn), 97% ee)

7

O

CO2Et

(66%; 95/5 (syn), 96% ee)

CO2Et

(82%; 92/8 (syn), 97% ee)

(67% conversion8h; (95% conversion4h; (93% conversion6h; (97% conversion4h; >95/5 (syn), 937 (syn), 92/8 (syn), 89/11 (syn), 95% ee (syn); TON = 20) 96% ee (syn); TON = 72) 97% ee (syn); TON = 24) 95% ee (syn); TON = 59) O

NHPMP

O

NHPMP

O

NHPMP

CO2Et

CO2Et

CO2Et

O

NHPMP CO2Et

N Me (82%; >95/5 (syn), 98% ee (syn)) (92% conversion5h; >95/5 (syn), 96% ee (syn); TON = 30)

(83%; 93/7 (syn), (73%5h; >95/5 (syn), 98% ee (syn)) 98% ee (syn)) (97% conversion53h; >94/6 (syn), 96% ee (syn); TON = 286)

N

N

N O

Catalyst

(90%rt, 20 h ; 86/14 (syn), 96% ee (syn))

NHTf N H

Tf = CF3SO2

Scheme 29.6  The asymmetric anti-Mannich reaction with a supported 3-aminopyrrolidine catalyst as

645

646

29  Supported Chiral Organocatalysts for Accessing Fine Chemicals

the best conditions were 2 mol% of the immobilized catalyst, in THF at 0 °C for four hours (95% conversion, 91/9 [syn], 96% ee). Polymer swelling is an important issue as regards mass transfer effects, and both THF and DMF (93% conversion at 4 mol% catalyst loading, at 0 °C for four hours, giving 92/8 (syn) and 97% ee) showed excellent properties in this regard. The reaction scope on the keto unit was then evaluated using both aldehyde and ketone substrates and N-PMP ethyl glyoxylate (unfortunately this component was not varied). Moreover, it was observed that the polymer supported catalyst gave better results than the homogenous counterpart. Recycling and reuse of the catalyst was assayed using cyclohexanone (in DMF) and the reaction could be recycled up to five times, with no drop in the conversion nor in the stereoselectivity which also remained constant. In the case of the continuous flow experiments, the catalyst (0.23 mmol on 500 mg resin), was charged to a packed-bed reactor and the carbonyl (0.52 M) and the N-PMP ethyl glyoxylate (0.26 M) reagents introduced to the system at a total flow rate of 0.2 mL min-1 (residence time = 12 minutes) (Scheme 29.6). The products were obtained with essentially the same yields and stereoselectivities as the batch conditions. The turn-over-numbers (TONS) were in the range 22 to 72. However, to demonstrate the robustness of this system, some experiments were performed for more than two days (see Scheme 29.6) and these proved to be exceptionally efficient, achieving TONs of >285 (see Scheme 29.6) and yields of products in the order of 68.15 mmol. The 5-(pyrrolidine-2-yl)tetrazole catalyst was reported independently by Arvidsson, Ley and Yamamoto in 2004, and immediately was found to be more effective than proline in a number of organocatalyzed reactions [17]. It became clear soon afterwards that it was an ideal catalyst for continuous flow reactions as demonstrated by Odedra and Seeberger in 2009, who demonstrated its application in Aldol and Mannich reactions [18]. It was also of interest to immobilize this efficient catalyst to solids so that it could be integrated into catalytic reactors. As a good example, in 2016, Massi et al. integrated this catalyst into a monolithic flow microreactor (in a monolithic reactor the catalyst occupies the space of the reactor in the form of a monolith; which is a structured material possessing a regular or irregular network of channels, which is generally formed within the reactor) by attaching it to a polystyrene resin [19]. (These workers have also previously worked with silicasupported pyrrolidinyl-tetrazole catalysts, which were accessed in both batch and continuous-flow conditions [20]). After attaching the tetrazole to the pyrrolidine unit by standard chemistry to form a styryl-functionalized pyrrolidinyl-tetrazole, it was polymerized in a glass tube to form the polystyrene monolithic catalyst using standard Fréchet-Svec conditions (Scheme 29.7a). The immobilized catalyst was then investigated in a simple bench-mark reaction with cyclohexanone and p-nitrobenzaldehyde, using 10 mol% of the catalyst and a variety of solvents and conditions. When the reaction was performed in water for four hours at room temperature with a catalyst loading of 10 mol% the reaction gave a yield of 95%, a diastereoselectivity of 10/1 (syn) and an ee for the major anti-isomer of 95%. Subsequently, for the key reactions in flow, the polymerization to form the monolithic microreactor was conducted in a stainless-steel column (the Boc deprotection was carried out by sequentially flowing TFA/THF and Et3N/THF solutions through the column). The scope of the reaction was then investigated, and the best conditions were conducting the reaction at 25 °C at a flow rate of 10–20 mL min−1 for five hours to give the aldol in quantitative yield, with good diaseteroselectivities and ees (Scheme 29.7b). It should also be pointed out that the reaction could be maintained five days on stream without any degradation of the catalyst, which was probably due to the resistance of the tetrazole unit toward any known side-reactions.

29.2  Organocatalyst Immobilizations

Scheme 29.7  (a) Preparation of the monolithic 5-(pyrrolidine-2-yl)tetrazole catalyst for an asymmetric aldol reaction reported by Massi et al (b) Organization of the flow microreactor used by Massi et al in an aldol reaction [19].

29.2.4  Organocatalysts Based on Immobilized Imidazolidinones These robust and easy to prepare organocatalysts, also known as MacMillan catalysts, are one of the most effective catalysts for a variety of asymmetric organocatalytic transformations and of course were the grounds for awarding the Nobel prize to David MacMillan [1]. This category of organocatalyst was first synthesized by MacMillan et al. at the beginning of the millennium, been originally employed in the enantioselective Diels-Alder reaction [21]. Its application increased significantly over time and, consequently, modifications and immobilizations to different supports were explored [22]. Some recent examples of immobilized chiral imidazolidinone catalysts using continuous-flow approaches are described. Back in 2014, Mandoli et al. reported the first example of a chiral imidazolidinone polymerbased monolithic reactor for a cycloaddition reaction leading to isooxazoline compounds which are present in diverse pharmacologically active compounds [23, 24]. After the MacMillan monomer preparation and copolymerization inside a high-performance liquid chromatography (HPLC) column (loading = 0.51 mmolg−1), the catalyst was used in the form of tetraflouroborate salt to synthesize the isoxazoline target compounds, through a 1,3-dipolar addition of N-benzyl-C-phenyl and crotonaldehyde using wet CH3NO2, under continuous flow conditions (at a flow rate of 2 µLmin−1 at room temperature). This reaction was a success, affording the product in 71% yield with an ee of 90% (endo) and a dr of 91:9 (endo/exo) (Scheme 29.8a). Additionally, two other reactions were tested exhibiting the reactional scope of this monolithic reactor. A Diels-Alder

647

648

29  Supported Chiral Organocatalysts for Accessing Fine Chemicals

Scheme 29.8  Enantioselective catalytic reactions using a imidazolidinone polystyrene supported monolithic reactor under continuous flow conditions. (a) 1,3-Dipolar cycloaddition of N-benzyl-C-phenyl nitrone and crotonaldehyde; (b) Diels-Alder cycloaddition between cinnamaldehyde and cyclopentadiene; (c) Friedel-Crafts alkylation of N-methyl pyrrole with cinnamaldehyde.

cycloaddition between cyclopentadiene and cinnamaldehyde (Scheme 29.8b) gave an impressive 99% yield, 90% ee (endo) and 47/53 dr, and the Friedel-Crafts alkylation between N-methyl pyrrole and cinnamaldehyde (Scheme 29.8c) provided the product in 55% yield and 15% ee (endo). This latter reaction gave less satisfactory results suggesting the need for optimization studies. Experiments under batch conditions were also performed indicating a loss in catalytic activity in a reasonably short period of time when compared with this continuous flow set-up. Also, a decrease in the yield could be verified after one cycle. Three years later, a polymer supported chiral N-picolimidazolidinone catalyst was developed by Benaglia et al. [25] for enantioselective imine reduction reactions. The products of these reactions (chiral amines) are important pharmacophores that are present in many pharmaceutical and agrochemical compounds. The previously prepared catalyst was recovered and reused during six reaction cycles, under batch conditions, affording the desire amine with good yields and ees. After the sixth cycle, a slight decrease in its catalytic activity occurred (76% yield and 87% ee). A considerable number of imines was then tested, giving the corresponding amines in good yields (62–99% yield) and ees (55–97%) (Scheme 29.9a). For the continuous flow approach (Scheme 29.9b), the polystyrene supported organocatalyst (at a loading of 0.57 mmol g−1) was packed into a HPLC column. Then, a mixture of the imine and HSiCl3 in dry DCM was pumped inside the reactor at 0.4 mLh−1 at room temperature. After treating the effluent from the reactor with a 10% solution of NaOH the resultant amine product was obtained with excellent yield and enantioselectivity (up to 91% ee).

29.2.5  Other Amino Acid and Peptide Type Catalysts Beyond the already mentioned catalysts, other successful amino acids and peptide type organocatalysts have been immobilized and tested in asymmetric flow processes. In this section we present some examples, mainly focused on supported primary amino acids and peptide derivatives. A more thorough overview can be found in the literature [26].

29.2  Organocatalyst Immobilizations

Scheme 29.9  Imine reduction with HSiCl3 using a PS immobilized picolimidazolidinone organocatalyst under (a) batch and (b) continuous flow conditions.

29.2.5.1  Supported-primary Amino Acid Catalysts

Primary amino acids have been a very successful type of organocatalyst for several bench-mark asymmetric transformations [27], catching the attention of many research groups. In 2014 Pericás et al. reported the application of a PS-supported threonine derivative as an immobilized organocatalyst in three-component anti-Mannich reactions [28]. It was found in preliminary tests that a mixture of DMF and CH2Cl2 (1:1) was the best solvent system for the bench mark reaction as DMF gave higher stereoselectivity and DCM afforded better results regarding sweeling ability and reaction rate. The reactional scope was explored by using multiple aniline and aldehyde derivatives. The reactions were completed after 3-8h providing the antiadducts, which are advanced synthons for the preparation of biologically active molecules, in moderate to very good yields (62 to 96%), enantio- (up to 95%) and diastereoselectivities (up to 8:92) (Scheme 29.10). The recyclability of this immobilized catalyst was also verified, providing conversions up to 99% and very good enantio- and diastereoselectivities during the first three cycles. However, a decrease in all three parameters was perceptible in the fourth cycle. A small number of reactions were additionally performed under continuous flow system, using a vertical

649

650

29  Supported Chiral Organocatalysts for Accessing Fine Chemicals

Scheme 29.10  Three-component anti-Mannich reaction catalyzed by a polystyrene-supported threonine derivative under continuous flow conditions as described by Pericàs et al

mounted packed-bed reactor, with a loading of 1.02 mmol g−1 of catalyst, and a flow rate of 30 µL min-1 at room temperature with a mixture of DMF and CH2Cl2 as solvent system. Five target antiMannich adducts were successfully obtained in good yields and with high diastereo- and enantiomeric purities (Scheme 29.10). 29.2.5.2  Supported-peptide Derivative Catalysts

Due to their modular nature, readily designable structure and inherent chirality, synthetic peptides are of great interest for applications in organocatalytic syntheses. These compounds have a wider structural diversity and are easier to immobilize, as compared to proline. A curious study using such compounds was made by Fülöp et al., who reported the catalytic activity of N-terminal prolyl-peptides in the asymmetric α-amination of aldehydes [29]. Some prolyl-peptides were prepared, but H-Pro-Asp-NH showed the best results in batch reactions, affording a quantitative conversion (100%), a good enantioselectivity of 79% ee and a productivity of 0.5 mmol product×mmol catalyst−1 × h−1. With all the optimizations in hand, the researchers looked at the continuous flow approach. The dipeptide was first immobilized in a polystyrene-polyethylene glycol grafted copolymer TentaGel – that is an ideal robust and pressure stable resin for flow-based purposes, with a loading of 0.48 mmol g−1. Then, 200 mg of the catalyst were housed in a packed-bed column and the reaction components consisting of the aldehyde and DBAD in CHCl3 were passed as a continuous stream through the column at a flow rate of 0.1 mL min−1 at room temperature. The α-hydrazino aldehyde formed was immediately treated with NaBH4 to form the corresponding alcohol, in order to avoid racemization (very similar to the strategy described by Pericàs in Scheme 29.4). The product, which is a precursor to a variety of biological active compounds, was obtained with a conversion of 86% and a remarkable ee of 90%. The scope of the reaction was studied using different aldehyde derivatives (Scheme 29.11). The reactions proceeded with conversions in the range 86–100%, and ees between 90 and 99%.

29.2  Organocatalyst Immobilizations

Scheme 29.11  α-Amination of aldehydes using a H-Pro-Asp-NH-TentaGel catalyst in a continuous-flow reactor.

29.2.6  Immobilized Amino-Cinchona Based Organocatalysts Cinchona alkaloids and their derivatives have a prominent position in organocatalysis, being some of the most active and stereoselective catalysts of the last decade. 9-Amino-9-deoxy-epi-cinchona catalysts have been used to catalyze several chemical transformations [30]. Their excellent versatility and enantioselectivity, plus the fact that their synthesis is not straight-forward turn the immobilization of these compounds into a very attractive option from an industrial point of view. In 2015, Puglisi et al. described an efficient immobilized 9-amino-9-deoxy-epi-quinine catalysts for a Michael addition reaction [31]. The general strategy to prepare the catalysts involved the introduction of a linker and then the attachment to a polystyrene support or simply the direct attachment to the solid support with no linker. Preliminary studies using the bench-mark Michael addition reaction of isobutyralde and trans-β-nitrostyrene, suggested that the use of the linker and a loading of 0.7 mmol g−1 of catalyst gave the best results (>99% yield, 95% ee). Recycling experiments of the catalyst showed that besides the excellent ees up until the sixth cycle, the yield started to decrease after the third cycle. The general scope of these catalysts was demonstrated using distinct α-branched aldehydes or ketones and nitroolefins as reagents, affording the products in moderate to good yields and very high enantioselectivities (Scheme 29.12a). Continuous flow conditions were also applied to these immobilized organocatalysts. For this, a stainless-steel column was packed with the polymer supported catalyst (250 mg) and a mixture of β-nitrostyrene, isobutyraldehyde and benzoic acid in toluene was pumped at room temperature into the reactor with a flow rate of 0.1 mL h−1. The Michael adducts were obtained with very good yields (70–99%) and very high enantioselectivities (85–93% ee) (Scheme 29.12b). The general applicability of this immobilized organocatalyst was demonstrated using two other reactions with different activation pathways. 29.2.6.1  Cinchona Picolinamide Derivatives

Another example of the successful application of cinchona-based organocatalysts, is that of the cinchona picolinamide types, that are often used effectively in hydrosilylation reactions [32, 33]. The cinchona picolinamide catalyst developed by Burke and Benaglia is one example to highlight [33]. A chiral picolinamide derived from 9-amino-epi-cinchonine immobilized on a polystyrene

651

652

29  Supported Chiral Organocatalysts for Accessing Fine Chemicals

Scheme 29.12  Enantioselective Michael addition reaction catalyzed by a polystyrene-immobilized amino cinchona derivative under (a) batch and (b) continuous-flow conditions as described by Puglisi et al

support was the best organocatalyst developed by this group, allowing the reduction of an imine with HSiCl3 and affording the product with a remarkable yield of 98% and an ee of 91% with a loading of only 10 mol%. The reaction scope was studied using this immobilized organocatalyst with various imines as starting material. In general, the yields were excellent (93–98%) and very good enantioselectivities were achieved (84–87% ee) (Scheme 29.13a). Furthermore, the first continuous flow set-up for the stereoselective synthesis of chiral amines using trichlorosilane was demonstrated, using the immobilized cinchona-based organocatalyst shown previously. Therefore, a packed-bed reactor was prepared with 170 mg of the catalyst. Then, a solution of the imine and HSiCl3 in CH2Cl2 was injected into the reactor at a flow rate of 0.02 mL min−1 at room temperature. The effluent was treated with an aqueous 10% NaOH solution. The product was first obtained with 98% yield and 47% ee, but unfortunately the ee dropped drastically to 20%, after 150 hours of continuous production (Scheme 29.13b). 29.2.6.2  Cinchona Squaramide Derivatives

These compounds were first studied as organocatalysts by Rawal [34] in 2008, who discovered that they have high catalytic activity for asymmetric Michael reactions. Since then, they have been applied frequently in this type of reaction and, in some cases, they have been evaluated and tested under continuous-flow conditions.

29.2  Organocatalyst Immobilizations

Scheme 29.13  Stereoselective reduction of imine derivatives with HSiCl3 using solid-supported chiral picolinamides developed by Burke and Benaglia in (a) batch and in (b) continuous flow set-up.

Szekely et al. [35] reported in 2019 the synthesis and application of cinchona squaramidemodified permethyl-β-cyclodextrin catalyst for the asymmetric Michael addition. Cyclodextrins (CD) are perfect platforms for the development of immobilized organocatalysts as they are obtained from renewable resources, are robust and stable. With that in mind, the catalysts were prepared by the attachment of the cinchona amine unit to the CD using the squaramide moiety as a linker. The activity of these catalysts was evaluated using 1,3-diketones and trans-βnitrostyrenes for a Michael addition under batch conditions. As can be seen from Scheme 29.14a, most products were obtained with high yields (up to 95%) and high enantioselectivities (up to 99% ee). For the continuous flow mode, the best reaction from the batch approach was chosen. The continuous flow set-up was performed in a coiled reactor coupled with a membrane separation unit, used to purify the product, and recycle the catalyst (Scheme 29.14b). A fresh solution of the reactants was pumped into the reactor (at room temperature) and then into the membrane (at 50 °C), where it met the retentate stream, which allowed the full recyclability of the catalyst (100%) and 50% of the solvent (2-MeTHF). The product was successfully obtained with 100% conversion, 98% purity, and 99% ee.

29.2.7  Other Organocatalysts In this final section, we will briefly look at two other families of organocatalysts that have also been used successfully in supported organocatalytic reactions.

653

654

29  Supported Chiral Organocatalysts for Accessing Fine Chemicals

Scheme 29.14  Michael addition of a 1,3-diketone with nitrostyrene catalyzed by cinchona squaramidemodified permethyl-β-cycodextrins under (a) batch and (b) continuous-flow modes.

29.2.7.1  Phosphoric Acid Catalysts

1,1´-Bi-2-naphthol (BINOL)-derived phosphoric acid (PA) organocatalysts have been applied with great success in asymmetric reactions over the last number of years [36]. These catalysts are ideal for bifunctional catalysis, due to the presence of a Brønsted acidic (P–OH) and Lewis basic site (P=O). They are air stable and easily stored. In 2014, Pericàs et al. reported on an enantioselective continuous-flow production of 3-indolylmethanamines catalyzed by a PS-immobilized PA [37]. The catalyst was immobilized at a loading of 0.26–0.37 mmol g−1 in four steps starting from (R)-6hydroxymethyl-2,2ʹ-bis(methoxymethyloxy)-1,1ʹ-binaphthalene. The catalyst was then used in an aza-Friedel-Crafts (loading = 10 mol%) with a tosylimine and indole presenting very good results; the aza-Friedel-Crafts adduct was obtained with a yield of 77% and an ee of 94% (the absolute configuration was assigned using single-crystal x-ray crystallography) after six hours at room temperature (Scheme 29.15). The catalyst could be recycled 14 times and reactivated by a simple acidic wash. The scope was then evaluated, and again good results were observed; except that sterically hindered and electron rich tosyl imines, like those derived from o-methylbenzaldehyde and p-methoxybenzaldehyde gave lower yields. When the cyclohexyl substituted imine was used the results were poor: 44% yield and 14% ee. The scope of the reaction on the indole component was also

29.2  Organocatalyst Immobilizations

Scheme 29.15  Enantioselective production of 3-indolylmethanamines using a PS-immobilized chiral phosphoric acid catalyst as described by Pericàs et al

evaluated and also afforded some good results (Scheme 29.15). The reaction was also conducted in continuous-flow mode between indole and the tosyl imine derived from p-tolualdehyde to give the reaction product with a conversion of ≥97%, and an ee of ≥91%, with an operation time of six hours using a system of two syringe pumps and a catalytic packed-bed reactor containing the catalyst (360 mg; 0.25 mmol g−1) at a combined flux rate of 0.2 mL min−1 to produce the indole product with an isolated yield of 80% (3.6 g) and a calculated TON of 102. Moreover, the catalyst loading for the global process was only 0.8 mol% (at a residence time of 9.3 minutes), which was 12 times lower than the loading for the batch process. This catalytic reactor was also used for a small amino-indole library synthesis that is not discussed here. 29.2.7.2  Isothiourea Catalysts

The final immobilized organocatalyst system that we will consider is the benzotetramisole (BTM) isothiourea class that was pioneered by Birmann [38] and championed over the last decade by

655

656

29  Supported Chiral Organocatalysts for Accessing Fine Chemicals

Smith [39]. Once again, Pericàs et al. have been instrumental in their application of this unique efficient and somewhat underrated catalyst system in heterogenous catalytic reactions, when they applied a polystyrene BTM-isothiourea to an interesting [8+2]-cycloaddition reaction that gave rise to cycloheptatriene-fused pyrrolidine derivatives [40]. This is a very useful method synthetically and was first introduced by von Doering and Wiley in 1960, but, unfortunately, there have been few asymmetric approaches reported in the intervening years. For this formal [8+2] annulation an azaheptafulvene was reacted with phenylacetic acid in the presence of a PS-supported-triazole tethered BTM-isothiourea (at 10 mol% loading) with pivaloyl chloride and 1,8-diazabicyclo[5,4,0] undec-7-ene (DBU) at 0 oC in DCM giving the target product with >95% conversion, a dr of 96/4, and an enantioselectivity of 91% ee (Scheme 29.16). The scope of the reaction was then investigated, for both the acetic acid and the azaheptafulvene components. When a broad range of phenylacetic acid derivatives were evaluated the cycloadducts were obtained in yields of 65–80%, with high enantioselectivities in the range of 90–97% ee and excellent diastereoselectivity (>20:1). The absolute configuration of the cycloadduct was assigned through recourse to single crystal x-ray crystallography. When the structure of the azaheptafulvene component was varied, it was obvious that the annulation was insensitive to the electronic nature of the aromatic N-substituent, as the cycloadducts were obtained in yield of 70–85%, diastereoselectivities of >20:1 and ees in the range 90–98%. Recycling experiments using the standard reaction (see previously), demonstrated that there was no significant drop in the stereoselectivity from the first to the seventh cycle, but there was a drop off in the yield becoming more extreme on the sixth and seventh cycles (approximately 50 and 40%, respectively). Finally, the accumulated TON for these recycling experiments was 44.7. A putative mechanism was presented by the authors which was essentially a non-concerted polar type [8+2] cycloaddition process.

Scheme 29.16  Catalytic asymmetric [8+2] annulation reactions using a PS-immobilized chiral benzotetramisole catalyst as described by Pericàs et al.

References

29.3  Conclusions In the autumn of 2022, organocatalysis was instated as the third pillar of catalysis with the award of the Nobel prize in chemistry to Benajamin List and David MacMillan for pioneering, conceptual, and developmental work in this new field. In this chapter, we have considered some of the major advances that have been undertaken in the last 10 years on the immobilization of these catalysts, with the principal goals of activation/stabilization, recycling, and reuse. As we have seen in this monograph, the ability to immobilize organocatalysts not only brings benefits in terms of catalyst recycling and reuse (reducing waste and costs) that is particularly beneficial for industry, but, more interestingly, immobilized organocatalysts have been the main staple of many continuous flow procedures, which has now become a standard industrial tool for accessing important APIs. In the coming years, we expect that immobilized organocatalysts will have greater impact on industrial processes, particularly from a drug development and production point of view.

References 1 (a) The Nobel Prize in Chemistry 2021. https://www.nobelprize.org/prizes/chemistry/2021/ summary; (b) Boerner, L.K. (2021). Pioneers of asymmetric organocatalysis win 2021 Nobel Prize in Chemistry. Chem. Eng. News. https://cen.acs.org/people/nobel-prize/Asymmetricorganocatalysis-List-MacMillan-Nobel-Prize-Chemistry-organic-synthesis/99/web/2021/10; (c) Rouhi, A.M. (2004.). A renaissance in organocatalysis. Chem. Eng. News 6: 41–45; (d) Durrani, J. (2021). How organocatalysis won the Nobel prize. Chem. World 18: 22–27. 2 (a) Hajos, Z.G. and Parrish, D.R. (1974). J. Org. Chem. 39: 1615–1621; (b) Eder, U., Sauer, G., and Weichert, R. (1971). Angew. Chem. Int. Ed. 10: 496–497. 3 Benaglia, M. (2021). Organocatalysis – Stereoselective Reactions and Applications in Organic Synthesis. Berlin/Boston: De Gruyter. 4 Cozzi, F. (2006). Adv. Synth. Catal. 348: 1367–1390. 5 Kristensen, T.E. and Hansen, T. (2010). Eur. J. Org. Chem. 3179–3204. 6 Krištofílová, D., Modrocká, V., Mečiarová, M., and Šebesta, R. (2020). ChemSusChem. 13: 2828–2858. 7 (a) Atodiresei, I., Vila, C., and Rueping, M. (2015). ACS Catal. 5: 1972–1985; (b) Rodríguez-Escrich, C. and Pericas, M.A. (2019). Chem. Rec. 19: 1872–1890. 8 Ayats, C., Henseler, A.H., and Pericàs, M.A. (2012). ChemSusChem. 5: 320–325. 9 Pothanagandhi, N. and Vijayakrishna, K. (2017). Eur. Poly. J. 95: 785–794. 10 Angamuthu, V. and Tai, D.-F. (2015). App. Cat. A. Gen. 506: 254–260. 11 Kong, Y., Tan, R., Zhao, L., and Yin, D. (2013). Green Chem. 15: 2422–2433. 12 Chronopoulos, D.D., Kokotos, C.G., Tsakos, M. et al. (2015). Mat. Lett. 157: 212–214. 13 Miranda, P.O., Llanes, P., Torkian, L., and Pericàs, M.A. (2013). Eur. J. Org. Chem. 6254–6258. 14 Fan, X., Sayalero, S., and Pericas, M.A. (2012). Adv. Synth. Catal. 354: 2971–2976. 15 Szcześniak, P., Buda, S., Lefevre, L. et al. (2019). Eur. J. Org. Chem. 6973–6982. 16 Martin-Rapún, R., Sayalero, S., and Pericàs, M.A. (2013). Green Chem. 15: 3295–3301. 17 (a) Hartikka, A. and Arvidsson, P.I. (2004). Tetrahedron Asymm. 15: 1831–1834; (b) Cobb, A.J.A., Shaw, D.M., and Ley, S.V. (2004). Synlett 558–560; (c) Torri, M., Nakadai, M., Ishihara, K. et al. (2004). Angew. Chem. Int. Ed. 43: 1983–1986. 18 Odedra, A. and Seeberger, P.H. (2009). Angew. Chem. Int. Ed. 48: 2699–2702. 19 Greco, R., Caciolli, L., Zaghi, A. et al. (2016). React. Chem. Eng. 1: 183–193. 20 Bortolini, O., Caciolli, L., Cavazzini, A. et al. (2012). Green Chem. 14: 992–1000.

657

658

29  Supported Chiral Organocatalysts for Accessing Fine Chemicals

21 Ahrendt, K., Borths, C., and MacMillan, D. (2000). J. Am. Soc. 122: 4243–4244. 22 (a) Raimondi, L., Faverio, C., and Boselli, M. (2021). Chapter 4: Chiral imidazolidinones: a class of privileged organocatalysts in stereoselective organic synthesis. In: Organocatalysis Stereoselective Reactions and Applications in Organic Synthesis (ed. M. Benaglia), 177–196. Berlin/Boston: Walter de Gruyter GmbH; (b) Deepa and Singh, S. (2021). Adv. Synth. Catal. 363: 629–656. 23 Chiroli, V., Benaglia, M., Puglisi, A. et al. (2014). Green Chem. 16: 2798–2806. 24 Kumar, G. and Shankar, R. (2020). ChemMedChem. 16: 430–447. 25 Porta, R., Benaglia, M., Annunziata, R. et al. (2017). Avd. Synth. Catal. 359: 2375–2382. 26 Triandafillidi, I., Voutyritsa, E., and Kokotos, C. (2021). Chapter 2: Recent advances in reactions promoted by amino acids and oligopeptides. In: Organocatalysis Stereoselective Reactions and Applications in Organic Synthesis (ed. M. Benaglia), 29–84. Berlin/Boston: Walter de Gruyter GmbH. 27 Coeffard, V., Greck, C., Moreau, X., and Thomassigny, C. (2015). Chapter 12: Other amino acids as asymmetric organocatalysts. In: Sustainable Catalysis: Without Metals or Other Endangered Elements, Part 1 (ed. M. North), 297–308. Cambridge: Royal Society of Chemistry. 28 Ayats, C., Hensele, A., Dibello, E., and Pericàs, M. (2014). ACS Catal. 4: 3027–3033. 29 Ötvös, S., Szloszár, A., Mándity, I., and Fülöp, F. (2015). Adv. Synth. Catal. 257: 3671–3680. 30 Burke, A. and Hermann, G. (2021). Chapter 3: Amino-cinchona derivatives. In: Organocatalysis Stereoselective Reactions and Applications in Organic Synthesis (ed. M. Benaglia), 85–175. Berlin/ Boston: Walter de Gruyter GmbH. 31 Porta, R., Benaglia, M., Coccia, F. et al. (2015). Adv. Synth. Catal. 357: 377–383. 32 Barrulas, P., Genoni, A., Benaglia, M., and Burke, A.J. (2014). Eur. J. Org. Chem. 7339–7342. 33 Fernandes, S., Porta, R., Barrulas, P. et al. (2016). Molecules 21: 1182–1190. 34 Malerich, J., Hagihara, K., and Rawal, V. (2008). J. Am. Chem. Soc. 130 (44): 14416–14417. 35 Kisszekelyi, P., Alammar, A., Kupai, J. et al. (2019). J. Catal. 371: 255–261. 36 Orlandi, M. (2021). Basic Principles of Substrate Activation through Non-covalent Bond Interactions in Organocatalysis: Stereoselective Reactions and Applications in Organic Synthesis (ed. M. Benaglia). Berlin/Boston: De Gruyter. 37 Osorio-Planes, L., Rodríguez-Escrich, C., and Pericàs, M.A. (2014). Chem. Eur. J. 20: 2367–2372. 38 Birman, V.B., Uffman, E.W., Jiang, H. et al. (2004). J. Am. Chem. Soc. 126: 12226–12227. 39 McLaughlin, C. and Smith, A.D. (2021). Chem. Eur. J. 27: 1533–1555. 40 Wang, S., Rodríguez-Escrich, C., and Pericàs, M.A. (2017). Angew. Chem. Int. Ed. 56: 15068–15072.

659

30 Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis Kotohiro Nomura1 and Nor Wahida Binti Awang2 1 2

Department of Chemistry, Tokyo Metropolitan University, Hachioji, Tokyo, 1920397, Japan Faculty of Applied Sciences, Universiti Teknologi MARA Sarawak Branch, 94300 Kota Samarahan, Sarawak, Malaysia

30.1  Introduction Development of sustainable functional polymers from renewable feedstocks has been an attractive subject in the fields of polymer chemistry, synthetic chemistry, and green sustainable chemistry [1–7]. There are many reports of chemicals derived from oxygen-rich molecular biomass (e.g. carboxylic acids, polyols, furans) by using ring-opening polymerization and condensation polymerization [5, 6, 8]. Hydrocarbon-rich molecular biomass (linseed, sunflower, soybean, castor, palm, and olive oils; vegetable oils and fatty acids) have been the most abundant and low-cost molecular biomass, obtained as fatty acids or as fatty acid methyl esters (FAME) obtained from vegetable oils by chemical modifications. Studies of bio-based advanced polyesters (exhibiting tunable mechanical properties and biodegradability) [9], in particular long chain aliphatic polyesters, have been promising semicrystalline materials as alternatives to linear polyethylene. It has been recognized that the precise polymerization technique provides new strategy and methodology for design of the macromolecular architectures [5–20]. Two major routes, condensation polymerization and acyclic diene metathesis (ADMET) polymerization and subsequent hydrogenation (Scheme 30.1), have been considered for this purpose [4, 5, 12, 14–16]. The approach of adopting (living) ring-opening polymerization (ROP) of cyclic monomers has also been considered, but is not introduced in this chapter due to limited monomer scope compared to the two methods previously mentioned. In this chapter, recent results in the synthesis of bio-based aliphatic polyesters by adopting two approaches are summarized in detail. In particular, we wish to introduce important points that must be considered for the goal of adopting these approaches.

Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 3, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

660

30  Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis

Scheme 30.1  Two approaches for synthesis of bio-based polyesters: (a) isomerization carbonylation and condensation polymerization, (b) olefin metathesis and hydrogenation.

30.2  Synthesis of Bio-Based Aliphatic Polyesters by Condensation Polymerization Low melting temperatures (Tm values) in entirely aliphatic polyesters with short and middle chain lengths (except for short chain aliphatic polyesters such as polylactic acid and polybutylene succinate) present a problem in thermoplastic processing [15, 21, 22]: undesired softening at elevated ambient temperatures [15, 23]. This problem could be overcome by the introduction of longer, crystallizable methylene segments [15, 24, 25]. It has been known that Tm values in the polyesters in the formula, [-O-(CH2)x-O-C(O)-(CH2)y-C(O)-]n, are strongly affected by methylene length (y) and whether there is an even or odd number in the diacid monomer unit [15, 23, 26]. This trend is due to the opposite/identical directions of dipoles between ester groups in close layers in the crystalline phase [15, 21, 27]; this also affects the crystalline structures. The extent of the odd−even effect, however, decreases with increasing number of methylene units and the Tm values become dependent on the total number of methylene groups and their distribution within the polymer chain [21]. Tm values higher than 100 °C (eventually close to linear polyethylene) could be achieved by increasing the percentage of long [21, 28], crystallizable methylene segments [15, 21, 25]. Polyesters are prepared by polycondensation through the esterification of dicarboxylic acids with diols or transesterification of the diesters with diols (AA, BB type monomers) or ω-hydroxy acids (AB type monomers), as shown in Scheme 30.2; the use of activated monomers (acid chlorides, anhydrides) is also considered. Synthesis of the polymers by ADMET polymerization [29–32] and subsequent hydrogenation is an alternative route [4, 9, 12, 15, 16, 33–39]. In contrast, synthesis by ring-opening polymerization (ROP), considered as an alternative route, could not be applied to large ring lactones due to their lower availability and low ring strain (considered to be the driving force) [40]. Many examples for catalyzed ROP of various lactones including large membered rings, exemplified by synthesis of high molecular weight poly(pentadecalactone) (Mn > 150,000; Mn = number average molecular weight) using Al-salen catalyst [41, 42], have been reported [41–45] and progress will provide possibilities in the future.

30.2  Synthesis of Bio-Based Aliphatic Polyesters by Condensation Polymerization

Scheme 30.2  Typical methods for the synthesis of aliphatic polyesters.

It is important to note that the precise control of molar ratios with very high conversions should be required for the synthesis of high molecular weight polymers (or, for certain applications, several 104 g mol−1) through condensation polymerization [46, 47]. Therefore, these polymerizations are generally conducted at a high temperature for long reaction times with efficient removals of byproducts (small molecules such as water) to reach a high degree of polymerization (DPn) to obtain high molecular weight polymers. For example, in the synthesis of poly(ethylene terephthalate) (PET), condensation polymerization of purified terephthalic acid with ethylene glycol (excess amount) has been conducted under reduced pressure at elevated temperatures up to 290 °C [47]. Due to the severe difficulty of removing diols owing to their high boiling points (e.g. 1,12-dodecanediol, 189 °C/12 mmHg; 1,16-hexadecane diol 197–199 °C/3 mmHg), this method cannot be considered for the synthesis of long chain aliphatic polyesters. A precise stoichiometric balance (the diol and the diacid or the diacid derivative) should be required to obtain of high molecular weight polymers. Precise control (purity) of monomers with methylene repeat units should be equally important, because we consider the odd-even effect toward the Tm values in the resultant polymers with α,ω-diacid repeat units with various chain lengths as mentioned previously [48–50]. For instance, precise stoichiometry of diols (algae oil) and diesters (C17 and C19) to obtain polyesters via polycondensation requires high molecular weight (Mn = 4.0 × 104) and melting temperatures (Tm = 99 °C) [51].

30.2.1  Synthesis of Bio-Based Aliphatic Polyesters by Condensation Polymerization and Dehydrogenative Condensation Pd-catalyzed isomerization methoxycarbonylation, as depicted in Scheme 30.3, yielded not only dimethyl 1,19-nonadecanedioate exclusively from methyl oleate (MO) [24], but also dimethyl 1,ω-carboxylate from unsaturated fatty acid esters (methyl oleate, methyl erucate, methyl linoleate) [24, 52]. Reduction of the resultant linear terminal C19 or C23 diacid esters with LiAlH4 yielded corresponding diols. Long-chain linear polyesters prepared by the polycondensation in the presence of Ti(OBu)4 possessed high molecular weights and high Tm values (102, 107 °C) [25]. The approach thus clearly demonstrates a possibility of synthesizing semicrystalline polyesters with high Tm values [24, 25]. However, the stochiometric reduction of diesters by LiAlH4 was employed to obtain diols because the catalytic hydrogenation of the esters generally requires harsh conditions (high temperature, high H2 pressure). Direct synthesis from linear 1,ω-diols by Ru catalyst containing a pincer ligand (called Milstein catalyst) was demonstrated by Robertson et al. (Scheme 30.3) [53]. The polycondensation could be

661

662

30  Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis

Scheme 30.3  Synthesis of aliphatic polyesters by palladium catalyzed isomerization methoxycarbonylation and polycondensation [25, 51], dehydrogenation polycondensation [53], and carbonylation polycondensation [55].

achieved by dehydrogenation of 1,ω-diols to afford hydroxy monoaldehyde, which was converted to esters by subsequent dehydrogenation from the hemiacetal formed by the addition of alcohol. The ruthenium catalyst has been known to exhibit highly efficient conversion of alcohols to esters [54]. The synthesis, however, requires rather severe conditions (150 °C for five days) for the completion and by-produced cyclic ester. The resultant semicrystalline polyesters from 1,10-decane diol or 1,9-nonane diol possessed high molecular weights (Mn = 138,000, 91,800), but rather low Tm values [53]. The direct synthesis from undecen-1-ol was demonstrated by carbonylation polymerization using a palladium catalyst containing diphosphine ligand (Scheme 30.3) [55]. The reaction in the presence of Pd-xantphos catalyst seemed to be preferred for the synthesis of high molecular weight linear polyesters (Mn = 10,300–17,400) without the formation of aldehyde and/or isomerized olefin byproducts; the phosphine ligands play the key role. Self-metathesis of erucic acid by G2 followed by catalytic hydrogenation gave 1,26-hexacosanedioate, and subsequent treatment with LiAlH4 yielded the diol. Polycondensation of 1,26-hexacosanedioate with the diol in the presence of Ti(OBu)4 gave the polyester (PE-26.26), which possesses a Tm value of 114 °C (Scheme 30.4) [56]. The Tm value is rather high compared to that reported previously for a similar polycondensation by Meier (104 °C) [57], in which PE-26.26, PE-12.26, and PE-4.26 were prepared by polycondensation to explore the effect of a alkoxy methylene spacer on thermal properties [57]. Self metathesis of undecanoic acid followed by hydrogenation by Pd/C to

30.2  Synthesis of Bio-Based Aliphatic Polyesters by Condensation Polymerization

Scheme 30.4  Synthesis of aliphatic long chain polyesters [56, 58].

give 1,20-eicosanedioic acid, and reduction with LiAlH4 yielded eicosane-1,20-diol. The subsequent polycondensation afforded the polyester (PE-20.20) possessing a Tm value of 108 °C [58]. Synthesis of linear polyesters based on 1,18‑(Z)‑octadec-9-enedioic acid, a natural unsaturated fatty diacid present in its esterified form, was reported by Roumanet et al. (Scheme 30.4) [59]. In particular, poly(1,18-octadecylene 1,18-octadecanedioate) prepared from 1,18-octadecanedioic acid and 1,18-octadecane diol possessed high melting temperature (Tm = 100 °C) and the possibility of using these polyesters as an alternative to petroleum-based synthetic polymers was suggested [59]. A biodegradation process was also highlighted for the resultant polyesters, such as poly(1,18-(Z)-octadec-9-enylene 1,18-(Z)-octadec-9-enedioate) and poly(1,18-octadecylene 1,18-octadecanedioate) [59]. More recently, Mecking et al. demonstrated closed-loop chemical recycling of polyester (Scheme 30.5) [60]. PE18,18 was prepared by the condensation polymerization of 1,18-octadeca dicarboxylic acid (derived from methyl oleate by biorefining) and 1,18-octadecane diol (prepared from the corresponding diacid ester, according to Scheme 30.3) in the presence of T(OnBu)4. The condensation polymerization of the diol with diethyl carbonate (DEC) in the presence of LiH gave polycarbonate (PC18, Mn = 90,000, Mw/Mn = 2.7). The resultant PC18 was treated with 10 wt% KOH in ethanol (at 120 °C for 24 hours) to yield 1,18-octadecanediol (purity 99% after recrystallization from MeOH) exclusively (yield 98%); the reaction of PC18 with MeOH at 150 °C for 24 hours also recovered the diol (Scheme 30.5) [60]. The recovered diol was treated with DEC to give the high molecular weight PC18 (recycled PC18), which possessed similar properties to the virgin polymer (Mn = 70,000, Mw/Mn = 3.4) [60]. Moreover, the resultant reaction mixture from PE18,18 treated in MeOH (150 °C, 12 hours) was polymerized in the presence of Ti(OnBu)4 to afford high molecular weight PE18,18 (Mn = 79,000, Mw/Mn = 1.9) [60]. These results clearly demonstrate a possibility of closed-loop chemical recycling

30.2.2  Synthesis of BioBasd Aliphatic Polyesters by Acyclic Diene Metathesis (ADMET) Polymerization and Subsequent Hydrogenation The ADMET polymerization approach [29–32] has also been considered, and three commercially available Ru-carbene catalysts, RuCl2(PCy3)2(CHPh) (G1; Cy = cyclohexyl) RuCl2(PCy3)(IMesH2) (CHPh) [G2; IMesH2 = 1,3-bis(2,4,6-trimethylphenyl)imidazolin-2-ylidene] and RuCl2(IMesH2)

663

664

30  Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis

Scheme 30.5  Closed-loop chemical recycling of long chain aliphatic polyesters by polycondensation and depolymerization [60].

(CH-2-OiPr-C6H4) (HG2), shown in Scheme 30.6, have been employed (probably due to their better functional group tolerance and insensitivity to moisture) [61–66]. The polymerization of 10-undecenoic acid and 10-undecenol (1, derived from castor oil) in the presence of G2 or HG2 gave high molecular weight polymers, especially with G2 (Mn = 22,000, 26,500). The Mn value was controlled by the addition of a mono-olefin (methyl 10-undecenoate, stearyl acrylate) used as a chain transfer agent for the end modification; polymerization in the presence of acrylate containing an oligomeric ethylene glycol segment was also studied for the synthesis of triblock copolymers [33]. The polymerization of dianhydro-D-glucityl bis(undec-10-enoate) (2) derived from castor oil (10-undecenoic acid) and glucose (isosorbide) was also demonstrated in the presence of G1, G2 (Scheme 30.6) [34]. The Mn values in the resultant polymers, poly(2), were low, and the value improved by conducting the polymerization under a nitrogen purge for the removal of the ethylene byproduct (run 2 vs. run 3, run 6 vs. run 7, Table 30.1). Due to the decomposition of the catalyst, there was accompanying olefin isomerization (probably by ruthenium-hydride) [67] generated in situ; the degree was measured by gas chromatography (GC) after transesterification of poly(2) by treating it with MeOH/ H2SO4 (depolymerization by the ester cleavage). The degree of isomerization was affected by the Ru catalyst rather than the polymerization temperature or nitrogen purging during the reaction [35]. The results also demonstrated the possibility of depolymerization of poly(2) by the treatment of acidified methanol. The degree of the isomerization was extensively reduced by the addition of benzoquinone during the polymerization [37]. The synthesis of aliphatic polyamides by adopting the ADMET polymerization of monomers derived from 10-undecenoic acid and 1,ω-diaminoalkane was also demonstrated [35]. Copolymerization of the linear α,ω-diene monomer (3) and the n-hexyl branched α,ω-diene monomer (4), derived from castor oil and vernonia oil, was conducted by G1 at 85 °C and subsequent hydrogenation in the presence of Pd/C (10%, H2 50 bar) yielded the corresponding long chain aliphatic polyesters containing a certain percentage of branching (Scheme 30.6) [39]. The resultant polymers were considered as a mimic of linear low-density polyethylene (LLDPE) and very-low-density polyethylene (VLDPE), but the composition was not uniform confirmed by differential scanning calorimetry (DSC) thermograms (multiple Tm values) [39].

30.2  Synthesis of Bio-Based Aliphatic Polyesters by Condensation Polymerization

Scheme 30.6  Acyclic diene metathesis (ADMET) polymerization of monomers derived from plant oil (castor oil) and sugar [34, 39, 68, 69, 71, 73, 74].

The polymerization of α,ω-diene monomers containing phenyl units derived from vanillin (5) [68] and eugenol (6) [69] in the presence of ruthenium catalysts was also reported. The resultant polymer, denoted as poly(5), possessed a rather high molecular weight (Mn = 10,000, Mw/Mn = 1.6) and a Tg value of 4 °C due to its C12 aliphatic chain [68]. The resultant polymers from 4-allyl-2-methoxyphenyl

665

666

30  Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis

Table 30.1  Synthesis of aliphatic polyester, poly(2), by acyclic diene metathesis (ADMET) polymerization using Ru-carbene catalysts1 [34]. N2 purge2

Mn3

Mw/Mn4

70

No

4,400

1.57

3

G1

80

No

4,750

1.56

4

G1

80

Yes

6,600

1.77

3

4

G1

100

No

5,000

1.61

425

5

G2

70

No

6,000

1.71

49

6

G2

80

No

6,100

1.61

69

7

G2

80

Yes

8,400

1.75

76

8

G2

90

No

6,200

1.65

66

run

cat.1

1

G1

2 3

temp. / °C

degree of isomerization5/ %

1)  Conditions: 1.0 mol% Ru cat, five hours. 2)  Continuous nitrogen purging during polymerization. 3)  Gel permeation chromatography (GPC) in tetrahydrofuran (THF) vs. polystyrene standards. 4)  Percentage of isomerized diesters observed with gas chromatography-mass spectrometry (GC-MS) after transesterification. 5)  Unidentified side-products.

10-undecenoate (6) by G2 possessed high molecular weight with unimodal molecular weight distribution [Mn = 10,300–12,700, Mw/Mn = 1.64–1.97] as well as uniform composition as amorphous materials based on DSC thermogram (Tg −9.6 °C) [69]. The reaction conditions (e.g. type of catalyst, amount, time) were optimized to obtain high molecular weight polymers, and conducting the polymerization using G2 (1.5–2.0 mol%) at 50 °C seemed to be preferred [69]. The polymerization of 6 in the presence of 5-formylbenzene-1,2,3-triyl tris(undec-10-enoate) as a crosslinker gave rather high molecular weight polymers (Mn = 13,600, Mw/Mn = 2.3) and formation of soluble network polymers was suggested from the 1H NMR spectra based on the disappearance of resonances assigned to terminal olefins [69]. These polymerizations were conducted at 50 °C to avoid rapid catalyst decomposition [4, 29, 70–72]. The polymerization of trehalose diundecenoate (7), prepared from trehalose by selective enzymatic esterification (the primary hydroxyl groups in the 6 and 6ʹ-positions) by treatment with vinyl undecenoate, by HG2 (4.0 mol%, 45 °C) gave high molecular weight semicrystalline material (Mn = 13,200) with a high Tm value (Tm = 156 °C, Scheme 30.6) [71, 73]. In the copolymerization with undec-10-en-1-yl undec-10-enoate (1), both the Mn and the Tm values decreased upon increasing the percentage of 1 in the resultant copolymer [71, 73]. The ADMET polymerization of 1,ω-dienes (8,9), prepared from D-xylose and D-mannose and a castor oil, by G2 under bulk conditions (no solvent) gave high molecular weight polymers [poly(8), Ru 0.1 mol%, Mn = 7.14‒7.16 × 104, Mw/Mn = 2.2‒2.3; poly(9), Mn = 3.24 × 104, Mw/Mn = 2.4, Scheme 30.6] [74]. The resultant polymers were amorphous materials and the Mn value was affected by the reaction temperature and the monomer/Ru molar ratios; conducting the polymerization at 90 °C with low Ru loading (0.1 mol%) seemed preferable to obtain high molecular weight polymers [74]. The polymerizations of symmetrical 1,ω-dienes (2,10) esters with isosorbide and glucarodilactone, respectively, by G2 (1.0 mol%), were conducted (80 °C, 16 hours) in the presence of methyl10-undecenoate as the end-capping (chain transfer) reagent [75]. Synthesis of the copolymers, poly(2-co-10)s, with various molar ratios were also explored [75]. Poly(2) is thermally stable and exhibits a lower glass transition temperature (Td = 369 °C, Tg = -10 °C) than poly(10) (Td = 206 °C, Tg = 32 °C) or poly(2-co-10)s. The copolymers featured a rubbery material that possessed a low

30.2  Synthesis of Bio-Based Aliphatic Polyesters by Condensation Polymerization

Young’s modulus with an average elongation at break of 480% and 640%, respectively, and shape memory properties [75]. The polymerization of 2,5-bis(hydroxymethylfuran) undecenoate (11) and the copolymerizations with 2 and 10 were also explored (Scheme 30.7) [76]. Replacement of 2 with 11 led to a significant loss of the elasticity and shape memory in the copolymers; these suggest that 2 (isosorbide as the middle) plays a role in mechanical performance [76]. It has been known that Tm value in the aliphatic polyesters [-(CH2)n-CO-O-]m is affected by number of the methylene unit (n). Copolymerization of undec-10-en-1-yl undec-10-enoate (1) and undeca-1,10-diene (UDD) with various molar ratios in the presence of Ru catalyst (G1) gave polyesters (Mn = 7,000–10,300). Subsequent olefin hydrogenation by another Ru catalyst (H2 40 bar, 110 °C, two days) afforded various long chain aliphatic polyesters (Scheme 30.8) and placed ester groups in different methylene spacing units [38]. As shown in Figure 30.1, a linear relationship between the Tm values and number of ester groups per methylene units differed from those in polyesters with precisely controlled methylene units. The results are due to a random distribution of the ester group in the polyethylene chain; the Tm values decreased linearly with increases in the mole fractions of ester groups [38]. A similar relationship was observed in the hydrogenated copolymers of dianhydro-D-glucityl bis(undec-10-enoate) (2) with 1,9-decadiene, prepared by tandem ADMET copolymerization and subsequent hydrogenation (Scheme 30.8) [72]. The polymerizations of di(icos-19-en-1-yl)tricosanedioate (12a) and di(tricos-22-en-1-yl)tricosanedioate (12b) by G1 and following hydrogenation using another Ru catalyst gave the corresponding polyesters (PE-38.23, and PE-44.23 (Scheme 30.9) [56]. As shown in Figure 30.2, a linear

Scheme 30.7  Synthesis of end-modified polyesters by acyclic diene metathesis (ADMET) polymerizations of monomers derived from isosorbide (2), glucarodilactone (10) and 2,5-bis(hydroxymethyl)furan (11) [75, 76].

667

668

30  Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis

Figure 30.1  Plots of Tm value vs. number of ester groups per methylene units [38].

Scheme 30.8  Synthesis of aliphatic polyesters by acyclic diene metathesis (ADMET) copolymerization undec-10-en-1-yl undec-10-enoate (1) with undeca-1,10-diene (UDD) or dianhydro-D-glucityl bis(undec-10enoate) (2) with 1,9-decadiene (DCD) and subsequent hydrogenation [38, 72].

relationship between the Tm value and the number of ester groups per methylene unit in [-O-(CH2)xO-C(O)-(CH2)21-C(O)-]n (PE-23.x), prepared by ADMET polymerization and the subsequent hydrogenation or condensation polymerization, was demonstrated [56, 58, 77]. Moreover, the Tm value in (PE-23.x) reached a constant value upon increasing number of methylene units (x) [56]. In contrast, the resultant polymer (PE-20.20) from undec-10-en-1-yl undec-10-enoate (9) by ADMET polymerization and subsequent hydrogenation possessed a Tm value of 103 °C (Mn = 28,000, Mw/Mn = 1.9) [58], which was rather low compared to the polyester (PE-20.20) prepared by polycondensation

30.2  Synthesis of Bio-Based Aliphatic Polyesters by Condensation Polymerization

Scheme 30.9  Synthesis of aliphatic long chain polyester prepared by acyclic diene metathesis (ADMET) polymerization and subsequent hydrogenation [38, 39, 56, 58].

Figure 30.2  Plots of melting temperature (Tm) vs. number of ester groups per methylene unit in [-O-(CH2)xO-C(O)-(CH2)21-C(O)-]n (PE-23.x) [56].

of 1,20-eicosanedioic acid and eicosane-1,20-diol (Tm = 108 °C). The results thus suggest that, as described [15], precise control of the microstructure plays the role. As described previously, internal olefinic double bonds in the resultant polymers prepared by ADMET polymerization must be hydrogenated for synthesis of saturated aliphatic bio-based polyesters, possessing higher Tm value (semicrystalline materials) as well as better stability [14–16, 24, 25, 28, 33–39, 68–70, 75, 77–80]. However, as exemplified in Schemes 30.8 and 30.9 [38, 39, 56, 58], the hydrogenation generally requires harsh conditions [81–88]. The catalytic hydrogenations (after isolation of unsaturated polymers) were carried out in the presence of RuCl2(PCy3)2(CHOEt) (4.0 MPa, 110 °C, 48 hours) [38, 56] or 10 wt% of Pd/C (H2 5.0 MPa, 80 °C, 20–72 hours) [39, 58]. Moreover, the hydrogenations using RhCl(PPh3)3 [84, 85], RuHCl(CO)(PCy3)2 [86], or Pd/C [87] under high pressure (3.1– 13.8 MPa, 80–90 °C, two to five days), or Ir(COD)(PCy3)(pyridine) (COD = 1,5-cyclooctadiene, called Crabtree’s catalyst) under rather mild conditions (c. Ir 2 mol%, c. H2 2 MPa at room temperature,

669

670

30  Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis

20 hours) [88] were also reported for synthesis of saturated polymers after the ADMET polymerization. p-Toluenesulfonyl hydrazide has been widely used in the laboratory level for this purpose [81–83], even with polymers prepared by ring opening metathesis polymerization [89, 90]. No reports concerning the tandem systems (hydrogenation without isolation of unsaturated polymers especially after ADMET polymerization) were studied for the synthesis of the saturated polyesters until recently.

30.2.3  One Pot Synthesis of Bio-Based Long Chain Aliphatic Polyesters by Tandem ADMET Polymerization and Hydrogenation. Depolymerization by Reaction with Ethylene Recently, one pot synthesis of rather high molecular weight polymers by ADMET polymerization of monomers (2, 13–15) derived from castor oil and sugars using G2 or HG2 and the subsequent tandem hydrogenation was demonstrated (Scheme 30.10). The polymerization temperature (50 °C) was lower than those conducted previously (80–100 °C) [70] to prevent the rapid decomposition of the catalyst [91–94]. Instead, the ethylene byproduct produced during the polymerization was removed from the reaction medium [32, 70]. The resultant polymers, poly(2), possessed higher Mn values (Mn = 11,900–14,000) [70] than those reported previously (Mn = 4,400–8,400) [34]; the polymerization by HG2 gave higher molecular weight polymers compared to those prepared by G2 (Table 30.2). No significant differences in the Mn values were observed irrespective of the reaction scale [poly(2): Mn = 15,900 (300 mg scale) vs. 15,800 (1.0 g scale)] [70]. Repetitive removal of ethylene should thus be a prerequisite for obtaining high molecular weight polymers; the removal of ethylene by replacement of solvent was also effective for this purpose (with higher Mn values) [72]. The subsequent olefin hydrogenation was conducted under mild conditions (H2 1.0 MPa, 50 °C) with the addition of a small amount of Al2O3 into the reaction mixture after the polymerization. Addition of Al2O3 in a small amount (1.0–1.7 wt %) is necessary for the one pot synthesis, and no significant differences in the Mn and the Mw/Mn values were seen before/after hydrogenation (Table 30.2) [70]. Moreover, the hydrogenation should be checked by DSC thermograms, not by 1H NMR spectra, because confirmation of complete hydrogenation seemed difficult due to insufficient accuracy in the 1H NMR spectra. Only one (rather sharp) Tm value was observed after completion of the hydrogenation, because (as previously described), melting temperatures of these polyesters are sensitive toward microstructure.

Scheme 30.10  Synthesis of bio-based polyesters by Ru-catalyzed acyclic diene metathesis (ADMET) polymerization and tandem hydrogenation, and reaction with ethylene after the polymerization [70].

30.3  Concluding Remarks and Outlook

Table 30.2  Synthesis of bio-based polyesters by tandem acrylic diene metathesis (ADMET) polymerization and hydrogenation [70].1 monomer (mmol)

2 (0.65)

cat.

Before H21 Mn3

Mw/Mn3

after H24 Mn3

Mw/Mn5

“”yield5 /%

HG2

14,000

1.42

15900

1.44

87

2 (0.65)

G2

11,900

1.38

13100

1.28

85

2 (2.08)

HG2

13,700

1.48

15800

1.53

93

13 (0.65)

HG2

10,200

1.38

11100

1.39

92

13 (2.08)

HG2

10,900

1.46

11,000

1.53

92

14 (2.09)

HG2

8,600

1.48

9,200

1.51

92

15 (0.71)

HG2

12,800

1.41

13,800

1.45

88

15 (2.36)

HG2

16,400

1.51

16,600

1.48

90

1)  Polymerization conditions: Ru cat. 2.0 mol %, monomer 300 mg (in CHCl3 0.14 mL) or 1000 mg (in CHCl3 0.34 or 0.38 mL), 50 °C, 24 hours. 2)  Sample before hydrogenation. 3)  Gel permeation chromatography (GPC) data in tetrahydrofuran (THF) vs. polystyrene standards. 4)  Sample after hydrogenation (H2 1.0 M pa, 50 °C, three hours, Al2O3 1.0–1.7 wt %). 5)  Isolated polymer yield.

It was revealed that treatment of the reaction solution, after the polymerization, with ethylene (0.8 MPa, 50 °C, one hour) gave oligomers [70]. This means that reaction of unsaturated olefinic double bonds in the polymers with ethylene led to the depolymerization, because the polycondensation proceeds by the removal of ethylene (or propylene) [29–32]. Resonances ascribed to protons/carbons in the terminal olefins were observed (in the 1H and 13C NMR spectra) for the sample after the reaction, whereas the other resonances were retained. The major product after the reaction would be dimers (including trimers and others) on the basis of the integration ratio of internal/ terminal olefins [70].

30.3  Concluding Remarks and Outlook This chapter introduces recent results focused on the synthesis of long chain aliphatic polyesters, focused on condensation polymerization (of dicarboxylic acids with diols) and ADMET polymerization. For monomer synthesis, the catalyst developments for key technologies such as isomerization alkoxycarbonylation, hydrogenation of esters, and others [95] play an essential role for success. Catalyst development for ADMET polymerization, in terms of activity, for synthesis of high molecular weight polymers should also be important. The monomer/polymer design, including end modification for purposes such as making blocks, grafting, and controlled networks, through new methodologies, should be important. Moreover, closed-loop chemical recycling (and upcycling) has been demonstrated recently (Scheme 30.11) [4, 60, 96]; we thus highly that believe extensive progress in these and related projects will be introduced in the near future.

671

672

30  Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis

Scheme 30.11  Mechanical recycling and chemical recycling of (bio-based) polyesters [4, 60, 95].

References 1 Gandini, A. (2008). Macromolecules 41: 9491–9504. 2 Coates, G.W. and Hillmyer, M.A. (2009). Macromolecules 42: 7987–7989. 3 Mülhaupt, R. (2013). Macromol. Chem. Phys. 214: 159–174. 4 Nomura, K. and Awang, N.W. (2020). ACS Sustainable Chem. Eng. 9: 5486–5505. 5 Gandini, A. and Lacerda, T.M. (2019). Monomers and polymers from chemically modified plant oils and their fatty acids. In: Polymers from Plant Oils, 2e. (eds. A. Gandini and T.M. Lacerda), 33–82. Beverly: Scrivener Publishing, LLC. 6 Yao, K. and Tang, C. (2013). Macromolecules 46: 1689–1712. 7 Wang, Z., Yuan, L., and Tang, C. (2017). Acc. Chem. Res. 50: 1762–1773. 8 Coulembier, O., Degée, P., Hedrick, J.L., and Dubois, P. (2006). Prog. Polym. Sci. 31: 723–747. 9 Meier, M.A.R., Metzger, J.O., and Schubert, U.S. (2007). Chem. Soc. Rev. 36: 1788–1802. 10 Hillmyer, M.A. and Tolman, W.B. (2014). Acc. Chem. Res. 47: 2390–2396. 11 Fortman, D.J., Brutman, J.P., De Hoe, G.X. et al. (2018). ACS Sustainable Chem. Eng. 6: 11145–11159. 12 Mutlu, H. and Meier, M.A.R. (2010). Eur. J. Lipid Sci. Technol. 112: 10–30. 13 Xia, Y. and Larock, R.C. (2010). Green Chem. 12: 1893–1909. 14 Biermann, U., Bornscheuer, U., Meier, M.A.R. et al. (2011). Angew. Chem. Int. Ed. 50: 3854–3871. 15 Stempfle, F., Ortmann, P., and Mecking, S. (2016). Chem. Rev. 116: 4597–4641. 16 Gandini, A. and Lacerda, T.M. (2019). Metathesis reactions applied to plant oils and polymers derived from the ensuing products. In: Polymers from Plant Oils, 2e. (eds. A. Gandini and T.M. Lacerda), 83–108. Beverly: Scrivener Publishing, LLC. 17 Zakzeski, J., Bruijnincx, P.C.A., Jongerius, A.L., and Weckhuysen, B.M. (2010). Chem. Rev. 110: 3552–3599. 18 Fenouillot, F., Rousseau, A., Colomines, G. et al. (2010). Prog. Polym. Sci. 35: 578–622. 19 Wilbon, P.A., Chu, F., and Tang, C. (2013). Macromol. Rapid Comm. 34: 8–37. 20 Thomsett, M.R., Storr, T.E., Monaphan, O.R. et al. (2016). Green Mater. 4: 115–134. 21 Korshak, V.V. and Vinogradova, S.V. (1965). Polyesters. Oxford, U.K: Pergamon Press. 22 Mandelkern, L. and Alamo, R.G. (2007). Thermodynamic quantities governing melting. In: Physical Properties of Polymers Handbook (ed. J.E. Mark), 165–186. New York: Springer. 23 Ishioka, R., Kitakuni, E., and Ichikawa, Y. (2002). Aliphatic polyesters: “Bionolle”. In: Biopolymers, 4. (eds. A. Steinbüchel and Y. Doi), 275–297. Weinheim, Germany: Wiley-VCH.

References

24 Quinzler, D. and Mecking, S. (2010). Angew. Chem. Int. Ed. 49: 4306–4308. 25 Stempfle, F., Ritter, B.S., Mülhaupt, R., and Mecking, S. (2014). Green Chem. 16: 2008–2014. 26 Korshak, W.V. and Vinogradova, S.V. (1953). Bull. Acad. Sci. USSR, Div. Chem. Sci. 2: 995–998. 27 Bunn, C.W. (1955). J. Polym. Sci. 16: 323–343. 28 Moser, B.R., Vermillion, K.E., Banks, B.N., and Doll, K.M. (2020). J. Am. Oil Chem. Soc. 97: 517–530. 29 Selected reviews for acyclic diene metathesis (ADMET) polymerization, see refs. 29-32 Schwendeman, J.E., Church, A.C., and Wagener, K.B. (2002). Adv. Synth. Catal. 344: 597–613. 30 Atallah, P., Wagener, K.B., and Schulz, M.D. (2013). Macromolecules 46: 4735–4741. 31 Pribyl, J., Wagener, K.B., and Rojas, G. (2021). Mater. Chem. Front. 5: 14–43. 32 Chen, Y., Abdellatif, M.M., and Nomura, K. (2018). Tetrahedron 74: 619–692. 33 Rybak, A. and Meier, M.A.R. (2008). ChemSusChem 1: 542–547. 34 Fokou, P.A. and Meier, M.A.R. (2009). J. Am. Chem. Soc. 131: 1664–1665. 35 Mutlu, H. and Meier, M.A.R. (2009). Macromol. Chem. Phys. 210: 1019–1025. 36 De Espinosa, L.M., Ronda, J.C., Galià, M. et al. (2009). J. Polym. Sci.: Part A: Polym. Chem. 47: 5760–5771. 37 Fokou, P.A. and Meier, M.A.R. (2010). Macromol. Rapid Commun. 31: 368–373. 38 Ortmann, P. and Mecking, S. (2013). Macromolecules 46: 7213–7218. 39 Lebarbé, T., Neqal, M., Grau, E. et al. (2014). Green Chem. 16: 1755–1758. 40 Dubois, P., Coulembier, O., and Raquez, J.-M. (2009). Handbook of Ring- Opening Polymerization. Weinheim, Germany: Wiley-VCH. 41 Van der Meulen, I., Gubbels, E., Huijser, S. et al. (2011). Macromolecules 44: 4301–4305. 42 Pepels, M.P.F., Koeken, R.A.C., van der Linden, S.J.J. et al. (2015). Macromolecules 48: 4779–4792. 43 Pepels, M.P.F., Bouyahyi, M., Heise, A., and Duchateau, R. (2013). Macromolecules 46: 4324–4334. 44 Bouyahyi, M. and Duchateau, R. (2014). Macromolecules 47: 517–524. 45 Witt, T. and Mecking, S. (2013). Green Chem. 15: 2361–2364. 46 Köpnick, H., Schmidt, M., Brügging, W. et al. (2000). Polyesters. In: Ullmann’s Encyclopedia of Industrial Chemistry (eds. W. Gerhartz and B. Elvers). Weinheim, Germany: Wiley-VCH. 47 Rogers, M.E. and Long, T.E. (2003). Synthetic Methods in Step-growth Polymers. Hoboken: Wiley-Interscience. 48 Pepels, M.P.F., Hansen, M.R., Goossens, H., and Duchateau, R. (2013). Macromolecules 46: 7668–7677. 49 Le Fevere de Ten Hove, C., Penelle, J., Ivanov, D.A., and Jonas, A.M. (2004). Nat. Mater. 3: 33–37. 50 Menges, M.G., Penelle, J., Le Fevere de Ten Hove, C. et al. (2007). Macromolecules 40: 8714–8725. 51 Roesle, P., Stempfle, F., Hess, S.K. et al. (2014). Angew. Chem. Int. Ed. 53: 6800–6804. 52 Jiménez-Rodriguez, C., Eastham, G.R., and Cole-Hamilton, D.J. (2005). Inorg. Chem. Commun. 8: 878–881. 53 Hunsicker, D.M., Dauphinais, B.C., Mc Ilrath, S.P., and Robertson, N.J. (2012). Macromol. Rapid Commun. 33: 232–236. 54 Gunanathan, C. and Milstein, D. (2011). Acc. Chem. Res. 44: 588–602. 55 Liu, Y. and Mecking, S. (2019). Angew. Chem. Int. Ed. 58: 3346–3350. 56 Stempfle, F., Ortmann, P., and Mecking, S. (2013). Macromol. Rapid Commun. 34 (1): 47–50. 57 Vilela, C., Silvestre, A.J.D., and Meier, M.A.R. (2012). Macromol. Chem. Phys. 213: 2220–2227. 58 Trzaskowski, J., Quinzler, D., Bährle, C., and Mecking, S. (2011). Macromol. Rapid Commun. 32: 1352–1356. 59 Roumanet, P.-J., Jarroux, N., Goujard, L. et al. (2020). ACS Sustainable Chem. Eng. 8: 16853–16860. 60 Häußler, M., Eck, M., Rothauer, D., and Mecking, S. (2021). Nature 590: 423–427. 61 Grubbs, R.H. (ed.) (2003). Books 61–63. In: Handbook of Metathesis, 1e Weinheim: Wiley-VCH.

673

674

30  Synthesis of Bio-based Aliphatic Polyesters from Plant Oils by Efficient Molecular Catalysis

62 Grela, K. (ed.) (2014). Olefin Metathesis: Theory and Practice. Hoboken, New Jersey, USA: John Wiley & Sons, Inc. 63 Grubbs, R.H., Wenzel, A.G., O’Leary, D.J., and Khosravi, E. (eds.) (2015). Handbook of Metathesis, 2e. Weinheim: Wiley-VCH. 64 Trnka, T.M. and Grubbs, R.H. (2001). Acc. Chem. Res. 34: 18–29. 65 Samojzowicz, C., Bieniek, M., and Grela, K. (2009). Chem. Rev. 109: 3708–3742. 66 Vougioukalakis, G. and Grubbs, R.H. (2010). Chem. Rev. 110: 1746–1787. 67 Schmidt, B. (2004). Eur. J. Org. Chem. 2009: 1865–1880. 68 Llevot, A., Grau, E., Carlotti, S. et al. (2015). Polym. Chem. 6: 7693–7700. 69 Le, D., Samart, C., Kongparakul, S., and Nomura, K. (2019). RSC Adv. 9: 10245–10252. 70 Nomura, K., Chaijaroen, P., and Abdellatif, M.M. (2020). ACS Omega 5: 18301–18312. 71 Piccini, M., Leak, D.J., Chuck, C.J., and Buchard, A. (2020). Polym. Chem. 11: 2681–2691. 72 Kojima, M., Abdellatif, M.M., and Nomura, K. (2021). Catalysts 11: 1098–1106. 73 Piccini, M., Lightfoot, J., Castro, D.B., and Buchard, A. (2021). ACS Appl. Polym. Mater. 3: 5870–5881. 74 Hibert, G., Grau, E., Pintori, D. et al. (2017). Polym. Chem. 8: 3731–3739. 75 Shearouse, W.C., Lillie, L.M., Reineke, T.M., and Tolman, W.B. (2015). ACS Macro Lett. 4: 284–288. 76 Lillie, L.M., Tolman, W.B., and Reineke, T.M. (2017). Polym. Chem. 8: 3746–3754. 77 Stempfle, F., Quinzler, D., Heckler, I., and Mecking, S. (2011). Macromolecules 44: 4159–4166. 78 Barbara, I., Flourat, A.L., and Allais, F. (2015). Eur. Polym. J. 62: 236–243. 79 Dannecker, P., Biermann, U., Sink, A. et al. (2019). Macromol. Chem. Phys. 220: 1800400. 80 Barbiroli, G., Lorenzetti, C., Berti, C. et al. (2003). Eur. Polym. J. 39: 655–661. 81 Boz, E., Nemeth, A.J., Alamo, R.G., and Wagener, K.B. (2007). Adv. Synth. Catal. 349: 137–141. 82 Rojas, G., Inci, B., Wei, Y., and Wagener, K.B. (2009). J. Am. Chem. Soc. 131: 17376–17386. 83 Boz, E., Nemeth, A.J., Ghiviriga, I. et al. (2007). Macromolecules 40: 6545–6551. 84 Hydrogenation of ADMET polymer using RhCl(PPh3)3, see refs. 85,86. Inci, B., Lieberwirth, I., Steffen, W. et al. (2012). Macromolecules 45: 3367–3376. 85 Baughman, T.W., Chan, C.D., Winey, K.I., and Wagener, K.B. (2007). Macromolecules 40: 6564–6571. 86 Sworen, J.C., Smith, J.A., Berg, J.M., and Wagener, K.B. (2004). J. Am. Chem. Soc. 126: 11238–11246. 87 Li, H., Rojas, G., and Wagener, K.B. (2015). ACS Macro Lett. 4: 1225–1228. 88 Pesko, D.M., Webb, M.A., Jung, Y. et al. (2016). Macromolecules 49: 5244–5255. 89 Gibson, V.C. and Okada, T. (2000). Macromolecules 33: 655–656. 90 Hou, X. and Nomura, K. (2016). J. Am. Chem. Soc. 138: 11840–11849. 91 Lehman, S.E. and Wagener, K.B. (2001). Macromolecules 35: 48–53. 92 Courchay, F.C., Sworen, J.C., and Wagener, K.B. (2003). Macromolecules 36: 8231–8239. 93 Hong, S.H., Wenzel, A.G., Salguero, T.T. et al. (2007). J. Am. Chem. Soc. 129: 7961–7968. 94 Jawiczuk, M., Marczyk, A., and Trzaskowsk, B. (2020). Catalysts 10: 887. 95 For example Witt, T., Haußler, M., Kulpa, S., and Mecking, S. (2017). Angew. Chem. Int. Ed. 56: 7589–7594. 96 Collias, D.I., James, M., and Layman, J.M. (eds.) (2021). Circular Economy of Polymers: Topics in Recycling Technologies. ACS Symposium Series, 1391. Washington, DC: American Chemical Society.

675

31 Modern Strategies for Electron Injection by Means of Organic Photocatalysts: Beyond Metallic Reagents Takashi Koike Department of Applied Chemistry, Faculty of Fundamental Engineering, Nippon Institute of Technology, 4-1 Gakuendai, Miyashiro-machi Saitama, Japan

31.1  Introduction The transfer of electrons is directly associated with the occurrence of redox reactions. More specifically, compounds that inject electrons into systems are known as reductants, or electron donors. The electron injection process by such compounds is pivotal and fundamental in the activation of inorganic metal salts and the transformation of organic compounds, as exemplified by a tremendous number of organic reactions, such as the Birch reduction, the deprotection of sulfonyl groups, and the pinacol coupling reaction. Classically, metallic reagents (e.g. lithium and samarium iodide) and electrolytic techniques have been employed to inject electrons. In addition, the tetrathiafulvalenes (TTFs) and tetraaminoethylenes (TAEs) have been recognized as strong organic electron donors in the ground state. Recently, the Murphy group has developed exquisitely designed organic electron donors, such as 1 and 2 (Figure 31.1) [1–6]. Indeed, the replacement of highly reactive and hard-to-handle alkali metal reagents and rare metals with organic molecular reagents is a significant topic in synthetic chemistry in the context of operational simplicity, versatility, and sustainability. More recently, organic photoredox catalysts (OPCs) exhibiting highly reducing properties have attracted great interest as a game-changing strategy for single-electron injection. In this chapter, recent advancements in organic photocatalytic systems for the reductive transformations of organic substrates are discussed.

31.2  Basic and Advanced Concepts for 1e− Injection by Organic Photoredox Catalysis In modern organic synthesis, it is essential to control the reaction selectivity. Similarly, controlling the number of electrons to be transferred is a major premise for constructing selective electron injection systems. In this context, metal-free organic photocatalytic systems have received growing attention. More specifically, over the past several years, organic photoredox catalysis has been Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 3, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

676

31  Modern Strategies for Electron Injection by Means of Organic Photocatalysts: Beyond Metallic Reagents

Figure 31.1  Conventional electron-injection strategies for organic synthesis.

regarded as a well-controlled single-electron transfer (SET) system (i.e. a 1e− injection system) [7–10]. In addition, photoredox systems are usually performed under operationally facile reaction conditions, such as under visible-light irradiation below room temperature. Thus, Figure 31.2 outlines two concepts for the 1e injection processes taking place during photoredox catalysis. More specifically, one involves the basic concepts of oxidative quenching and reductive quenching (Figure 31.2a), whereas the other involves advanced concepts based on consecutive photoinduced electron transfer (conPET) (Figure 31.2b). As shown in Figure 31.2, irradiation of the OPC (organic photoredox catalyst) with light (hν1) produces the excited state of photocatalyst (*OPC), which can initiate either a 1e− oxidation or a 1e− reduction process. Single-electron transfer (SET) from *OPC to an electron-accepting (EA) substrate in the first SET event generates a reduced substrate (EA·−) together with the oxidized catalyst (OPC·+), which serves as a highly oxidizing agent. The second SET event involves SET from an electron-donating (ED) substrate to OPC·+, which regenerates the ground state OPC, and produces the oxidized substrate (ED·+). This process is known as the oxidative quenching pathway (*OPC→OPC·+→OPC). In contrast, the process triggered by the reduction of the excited *OPC species in the first SET event, followed by the second SET event starting from the reduced OPC·– is also viable, and is known as the reductive quenching pathway (*OPC→OPC·–→OPC). The advanced strategy known as conPET follows the formation of the reduced species (OPC·–) generated not only by the reductive quenching pathway (i.e. the blue pathway in Figure 31.2b), but also by means of electrochemical methods (i.e. the purple pathway in Figure 31.2b). Further photoexcitation (hν 2 ) generates the excited reduced species (*[OPC·–]), which serves as a super 1e− reductant. Finally, following these electron transfer processes, OPC is regenerated (i.e. the brown pathway in Figure 31.2b). Thus, if the reaction system is designed appropriately, *OPC, OPC·–, and *[OPC·–] can be employed as electron-injection reagents. Although transition metal complexes such as [Ru(bpy)3]2+ and fac-[Ir(ppy)3] (bpy: 2,2′-bipyridine, ppy: 2-pyridylphenyl) are still commonly used as photoredox catalysts [11], the development of new OPCs is necessary from the perspective of synthetic electron-injection strategies for sustainable development goals (SDGs). In the following sections, we describe the recent advances in organic photoredox catalysis. In particular, the molecular design of highly reducing OPCs and conceptual designs of conPET using OPCs are discussed. More recently, an electron-injection system combining organic photoredox catalysis and electrolytic techniques has emerged as a more sustainable strategy, and so this will be also presented.

31.3  Triarylamine-based Highly Reducing Organic Photocatalysts B

R

R

R

O

T

T A

S

R R

Figure 31.2  (a) Basic and (b) advanced concepts for 1e− injection by organic photoredox catalysis.

31.3  Triarylamine-based Highly Reducing Organic Photocatalysts As mentioned in the previous section, the oxidative quenching pathway is a basic electron-injection strategy that uses OPCs. In particular, by appropriately designing OPCs, electron transfer from excited species (*OPC) to less electrophilic substrates is viable. In other words, OPCs exhibit a sufficiently high reducing power even in the absence of sacrificial EDs. In recent years, an increasing number of molecular scaffolds have been reported for organic photocatalysis. Thus, in this section, we discuss several representative examples of recently reported triarylamine-based OPCs with high reducing powers. The electron-rich triarylamine unit is a substructure that is frequently found in organic dyes, and many triarylamines exhibit reversible redox behaviors. These facts have therefore encouraged researchers to develop triarylamine-based OPCs. It should be noted that in the

677

678

31  Modern Strategies for Electron Injection by Means of Organic Photocatalysts: Beyond Metallic Reagents

following systems, the reported overall photocatalytic performance is not always consistent with the reducing power of the photoexcited OPC (*Eox(OPC)) because the performance is dependent on the balance of the electron injection and back-electron transfer processes (Figure 31.2). Furthermore, the durability of the catalyst, especially in the case of OPCs, is often more important than the reduction power (*Eox[OPC]). In 2015, the Hawker and Alaniz group reported that 10-phenylphenothiazine (PTH) serves as a highly reducing OPC (*Eox = −2.1 V vs. SCE) for the dehalogenative hydrogen transfer reactions of carbon–halogen bonds such as in the case of alkyl and aryl halides (i.e. 3 and 4) (Figure 31.3) [12]. Aryl bromides, aryl chlorides, and aryl iodides have been employed in the reaction, which is assumed to be triggered by the highly-reducing excited *PTH species, leading to the corresponding aliphatic and aromatic products 5 and 6, respectively. The Hawker and Fors group has also demonstrated the application of PTH in the atom transfer radical polymerization (ATRP) of alkyl methacrylate 7 in the presence of a traditional ATRP initiator, α-bromocarbonyl 3b (Figure 31.4) [13]. More recently, Hawker et al. and Miyake et al. extensively studied the catalytic structure–activity relationships for the ATRP process. In particular, the Miyake group conducted diverse modifications of phenothiazine (PTH), phenoxazine (POX), dihydroacridine (ACR), and dihydrophenazine (PAZ) (Figure 31.5) [14–19]. These scaffolds generally exhibit high reducing powers in their excited states (i.e. *Eox ≈ −2.3 V), and it was determined that their photochemical properties are derived from the connections between the individual scaffolds, the auxiliaries present on the aromatic rings, and the

S

Figure 31.3  Dehalogenative hydrogen transfer of organic halides by 10-phenylphenothiazine (PTH) catalysis.

Figure 31.4  Atom transfer radical polymerization (ATRP) by means of phenylphenothiazine (PTH) catalysis.

31.3  Triarylamine-based Highly Reducing Organic Photocatalysts

I

R

A L

Figure 31.5  Phenothiazine derivatives for use as organic photoredox catalysts (OPCs).

incorporation of heteroatoms. More specifically, a change from a phenyl to a 1-naphthyl moiety promotes a charge transfer (CT)-type excitation. In addition, the heterocyclic components are strongly associated with determining the oxidation potential, absorption properties, and emission properties of these compounds, which can be readily tuned by variation in their substituents. Furthermore, in 2016 and 2017, Miyake et al. reported that naph-PAZ (*Eox = −2.12 V) served as a good OPC for the ATRP reaction involving α-bromocarbonyl 3b and methyl methacrylate 7, in addition to action in the trifluoromethylation of electron-rich heteroarene 8 with CF3I 9, leading to synthesis of CF3-heteroarene 10 [14, 16]. It was deduced that the key process in these reactions involves generation of the corresponding alkyl radicals via 1e− injection from the excited OPC (*OPC) to the alkyl halides (Figure 31.6). Over the past several decades, the catalytic fluoromethylation reaction has developed rapidly due to the importance of the fluoromethyl group as a useful structural motif, especially in the area of medicinal chemistry [20–23]. Although the radical fluoromethylation process has become one of the reliable methods for the preparation of organofluorine compounds, the radical monofluoromethylation process has yet to be reported in detail. As some limited examples, in 2019 and 2021, Koike et al. reported that 1,4-bis(diarylamino)naphthalenes (BDNs) serve as highly reducing OPCs (*Eox ≈ −2.0 V) for the radical monofluoromethylation process (Figure 31.7) [24–26]. In particular, BDN itself is effective in the generation of monofluoromethyl radicals via 1e− injection into the much less electrophilic neutral CH2F-reagent N-tosyl-S-monofluoromethyl-S-phenylsulfoximine 11. Furthermore, the hydroxy-monofluoromethylation of aromatic alkenes 12 has also been developed. This is a synthetic method for γ-fluoroalcohols 13 from simple alkenes 12. In general, the efficiency of photochemical reaction depends on the longevity of the excited state and the amount of photoactivated species. In the case of BDNs, despite the short lifetime of the excited state species (τ = 8–10 ns), high fluorescence quantum yields (φ = 0.94–0.99) were observed. The efficient formation of the excited state may therefore compensate for the short lifetime of the excited species. In addition, after the first SET event from the excited species (*BDNs) to the electron-deficient substrate (11), the formation of a 1e−-oxidized species (i.e. BDNs·+) stabilized by the partial delocalization of a hole over the two diarylamino units is a feature of these catalytic systems. These results may therefore act to guide future catalyst design. Although the OPCs highlighted in the present section exhibit a high reducing power (*Eox) via the oxidative quenching pathway, they do not reach those obtained by alkali metals. However, the use of OPCs provides reaction systems with a high tolerance toward various functional groups, as well as air and water. Thus, the development of super-electron-injecting OPC systems remains an ongoing aim.

679

680

31  Modern Strategies for Electron Injection by Means of Organic Photocatalysts: Beyond Metallic Reagents

S

Figure 31.6  (a) Atom transfer radical polymerization (ATRP) and (b) the radical trifluoromethylation of electron-rich aromatics by means of naph- dihydrophenazine (PAZ) catalysis.

A

S

Figure 31.7  Photocatalytic monofluoromethylation by 1,4-bis(diarylamino)naphthalene (BDN) (Ar = Ph).

31.4  Consecutive Photoinduced Electron Transfer (conPET) by Organic Photoredox Catalysis

31.4  Consecutive Photoinduced Electron Transfer (conPET) by Organic Photoredox Catalysis As shown in Figure 31.2b, advanced concepts, such as conPET, provide stronger reducing catalysts than more basic concepts [27]. As an example of a representative conPET system, in 2010, König and Ghosh et al. [28] investigated perylene diimides, a class of fluorescent dyes that have been employed as pigments, colorants, photoreceptors, and electronic materials. They found that N,N-bis(2,6diisopropylphenyl)perylene-3,4,9,10-bis(dicarboximide) (PDI) was effective for the dehalogenative hydrogen transfer of aryl halides and the homolytic aromatic substitution (HAS)-type aryl–aryl crosscoupling between electron-deficient aryl halides and electron-rich heteroaromatics (Figure 31.8). During this process, photoexcitation (hν1 ) results in the excited *PDI being reductively quenched by Et3N to give PDI·– and the radical cation of triethylamine (Et3N·+). The subsequent excitation of a 1e−-reduced species (PDI·–) is a key process here, and the ground state of PDI·–, which possesses a relatively long lifetime, can be excited again (hν 2 ) to generate *[PDI·–]. This excited species then undergoes SET to aryl halide 4, giving the corresponding aryl radical (·Ar) and regenerating ground state PDI. In addition, it was found that the second excited species (*[PDI·–]) exhibited a higher reducing power than the 1e−-reduced species (PDI·–). It should be noted that when the aryl radical abstracts a hydrogen atom from either Et3N·+ or the solvent (S–H), the dehalogenative hydrogen transfer product (i.e. compound 6) is obtained. In contrast, when the aryl radical reacts with the heteroaromatic moiety 14 and subsequent deprotonation takes place, coupled biaryl product 15 is obtained via its corresponding radical and cationic intermediates (i.e. 16 and 17). As shown in Figure 31.9, it has also been reported that rhodamine-6G (Rh-6G) [29], 9,10-dicyanoanthrathene (DCA), and 1,8-dihydroxyanthraquinone (Aq-OH) can act as OPCs for the conPET process [30, 31]. In these systems, the OPCs play the roles of 1e− oxidants and 1e− reductants, and so they are not always, effective despite their high reducing powers. Although the short lifetime of the *OPC species in its singlet excited state may be considered a detrimental point, it is not necessary for the photoexcited species in these conPET systems to be particularly long-lived species if a sacrificial electron donor is present in the system at a concentration sufficient to quench the short-lived *OPC. Although the above-mentioned systems are useful for the reaction of electron-deficient aryl halides, they are not appropriate for systems involving electron-rich aryl halides. In this context, in 2020, Nicewicz et al. overturned the concept of the established photocatalytic system based on 9-mesityl acridinium salts (Mes-Acr+), which are well-known to serve as highly oxidizing catalysts (oxidizing power of *[Mes-Acr+]: *Ered ≈ +2.2 V) [32, 33]. They found that the photoexcited *[MesAcr·] species generated by means of the conPET process could serve as a super-reducing catalyst (*Eox = −3.36 V) to reduce tosyl amides 18 and electron-rich aryl halides 20 [34]. Importantly, the reducing power of this species was determined to be comparable to that of lithium (i.e. −3.29 V). As shown in Figure 31.10, this system is useful for the de-tosylation into N–H amines 19, which is usually performed using alkali metals (Figure 31.10a), and the dehalogenative hydrogen transfer of 20 (Figure 31.10b). More specifically, in the case of the latter reaction, Mes-Acr+ is initially excited by photoirradiation (hν1 ) prior to its reductive quenching by Et3N to give Mes-Acr· along with the trialkylamine radical cation (R3N·+). Subsequently, Mes-Acr· is excited again (hν 2 ) to generate *[Mes-Acr·], which serves as a super-electron injector. As a result, even electron-rich aryl halide 20 can be reduced by *[Mes-Acr·] to give ·Ar and regenerate the ground state Mes-Acr+. Finally, the aryl radical abstracts a hydrogen atom from R3N·+ to produce the dehalogenative hydrogen transfer product 21.

681

682

31  Modern Strategies for Electron Injection by Means of Organic Photocatalysts: Beyond Metallic Reagents

S

P

Figure 31.8  (a) Dehalogenative hydrogen transfer and (b) Homolytic aromatic substitution (HAS)-type aryl–aryl cross-coupling by consecutive photoinduced electron transfer (conPET) of N,N-bis(2,6diisopropylphenyl)perylene-3,4,9,10-bis(dicarboximide) (PDI).

Figure 31.9  Examples of organic photoredox catalysts (OPCs) for use in the consecutive photoinduced electron transfer (conPET) process.

31.5  Consecutive Photoinduced Electron Transfer (conPET) by the Combination of Organic Photocatalysis and Electrolysis

Figure 31.10  (a) Reductive de-tosylation and (b) dehalogenative hydrogen transfer by the consecutive photoinduced electron transfer (conPET) of Mes-Acr+.

31.5  Consecutive Photoinduced Electron Transfer (conPET) by the Combination of Organic Photocatalysis and Electrolysis As highlighted above (Section 31.4), Mes-Acr+ serves as a super-electron donor, where in the presence of sacrificial electron donors (NR3) results in its reducing power being comparable to those obtained by alkali metal reagents. However, the use of sacrificial electron donors often limits the scope of the reaction because the oxidized species (R3N·+) serves either as an oxidant or as a hydrogen donor. Thus, the development of super-electron-injecting catalytic systems that do not require sacrificial terminal reductants remains of key importance in the context of further sustainability. To address this issue, cathodic reduction has been regarded as the most appropriate

683

684

31  Modern Strategies for Electron Injection by Means of Organic Photocatalysts: Beyond Metallic Reagents

solution. Indeed, the recent merging of photocatalysis with electrochemistry has opened up unexplored areas of redox chemistry, with several groups reported the use of oxidative electrophotocatalysis [35–38]. However, in the present section, we describe some recent studies into the reductive organic electrophotocatalysis process [39–42]. More specifically, in 2020, the groups of Lin, Lambert, and Wickens obtained super-electroninjecting systems by merging cathodic reduction with organic photoredox catalysis to reach a potential of approx. −3.3 V [43, 44]. This concept is illustrated in Figure 31.11a, wherein OPCs, DCA and naphthalene monoimide (NpMI), are the main components. In this system, the ground state OPC initially undergoes SET from the cathode to form a 1e−-reduced species (OPC·–); this process is an alternative to 1e− transfer from a sacrificial tertiary alkyl amine via a reductive quenching pathway. Subsequently, the obtained OPC·– is excited by visible light (hν) to form a super-reducing excited species (*[OPC·–]), which promotes the 1e− reduction of aryl halides (20), including aryl chlorides. It should be noted that this system was found to be applicable to various

Figure 31.11  A consecutive photoinduced electron transfer (conPET)-based arylation process by the combination of organic photocatalysis and electrolysis.

References

arylation reactions, including a HAS-type aryl–aryl coupling (22), and borylation (23), stannylation (24), and phosphorylation (25) reactions. More diverse functionalization of the aryl structure was also reported, presumably because the terminal reductant was changed from an alkyl amine to provide clean and sustainable electricity.

31.6  Summary and Outlook Alkali metals have been traditionally employed as electron sources for organic synthesis. However, they are highly reactive and dangerous, which is undesirable for a number of reasons. In particular, the water sensitivities of such species and the resulting deterioration in their functionalities are major issues in the contexts of functional group tolerance and sustainability. Thus, an ongoing challenge in the area of synthetic organic chemistry is the development of operationally more facile strategies to produce electron sources. In this context, noble metal-free organic super-reducing photocatalysts have been investigated as a potential alternative. More specifically, the appropriate design of organic photoredox catalyst systems has the potential to yield super-electron injectors that act through single-electron transfer. Although this review provided only a brief overview of such systems, we wish to emphasize the present 1e− injection systems can be applied to a range of synthetically valuable reactions, which are often challenging under alkali metal conditions due to damage to the reactant functional groups. A rational approach to catalyst design therefore involves the evolution of electron-injection strategies in terms of the operational simplicity and green chemistry. Moreover, the studies discussed herein demonstrate that the combination of organic photocatalysis with electrolysis has taken both the photo- and electrochemical synthetic methods to new levels. These modern electron-injection strategies are therefore expected to lead to further important innovations in area of organic synthesis.

References 1 Murphy, J.A. (2014). J. Org. Chem. 79: 3731–3746. 2 Doni, E. and Murphy, J.A. (2014). Chem. Commun. 50: 6073–6087. 3 Broggi, J., Terme, T., and Vanelle, P. (2014). Angew. Chem. Int. Ed. Engl. 53: 384–413. 4 Murphy, J.A., Khan, T.A., Zhou, S.Z. et al. (2005). Angew. Chem. Int. Ed. Engl. 44: 1356–1360. 5 Murphy, J.A., Zhou, S.Z., Thomson, D.W. et al. (2007). Angew. Chem. Int. Ed. Engl. 46: 5178–5183. 6 Koike, T. and Akita, M. (2021). Trends Chem. 3: 416–427. 7 Romero, N.A. and Nicewicz, D.A. (2016). Chem. Rev. 116: 10075–10166. 8 Silvi, M. and Melchiorre, P. (2018). Nature 554: 41–49. 9 Lee, Y. and Kwon, M.S. (2020). Eur. J. Org. Chem. 2020: 6028–6043. 10 Vega-Penaloza, A., Mateos, J., Companyo, X. et al. (2021). Angew. Chem. Int. Ed. Engl. 60: 1082–1097. 11 Marzo, L., Pagire, S.K., Reiser, O., and König, B. (2018). Angew. Chem. Int. Ed. Engl. 57: 10034–10072. 12 Discekici, E.H., Treat, N.J., Poelma, S.O. et al. (2015). Chem. Commun. 51: 11705–11708. 13 Treat, N.J., Sprafke, H., Kramer, J.W. et al. (2014). J. Am. Chem. Soc. 136: 16096–16101. 14 Theriot, J.C., Lim, C.-H., Yang, H. et al. (2016). Science 352: 1082–1086. 15 Pearson, R.M., Lim, C.H., McCarthy, B.G. et al. (2016). J. Am. Chem. Soc. 138: 11399–11407. 16 Du, Y., Pearson, R.M., Lim, C.H. et al. (2017). Chem. Eur. J. 23: 10962–10968.

685

686

31  Modern Strategies for Electron Injection by Means of Organic Photocatalysts: Beyond Metallic Reagents

17 McCarthy, B.G., Pearson, R.M., Lim, C.H. et al. (2018). J. Am. Chem. Soc. 140: 5088–5101. 18 Sartor, S.M., McCarthy, B.G., Pearson, R.M. et al. (2018). J. Am. Chem. Soc. 140: 4778–4781. 19 Buss, B.L., Lim, C.H., and Miyake, G.M. (2020). Angew. Chem. Int. Ed. Engl. 59: 3209–3217. 20 Hagmann, W.K. (2008). J. Med. Chem. 51: 4359–4369. 21 Zhou, Y., Wang, J., Gu, Z. et al. (2016). Chem. Rev. 116: 422–518. 22 Bezencon, O., Heidmann, B., Siegrist, R. et al. (2017). J. Med. Chem. 60: 9769–9789. 23 Meanwell, N.A. (2011). J. Med. Chem. 54: 2529–2591. 24 Noto, N., Koike, T., and Akita, M. (2019). ACS Catal. 9: 4382–4387. 25 Taniguchi, R., Noto, N., Tanaka, S. et al. (2021). Chem. Commun. 57: 2609–2612. 26 Noto, N., Takahashi, K., Goryo, S. et al. (2020). J. Org. Chem. 85: 13220–13227. 27 Glaser, F., Kerzig, C., and Wenger, O.S. (2020). Angew. Chem. Int. Ed. Engl. 59: 10266–10284. 28 Indrajit Ghosh, T.G., Bardagi, J.I., and König, B. (2014). Science 346: 725–728. 29 Ghosh, I. and Konig, B. (2016). Angew. Chem. Int. Ed. Engl. 55: 7676–7679. 30 Neumeier, M., Sampedro, D., Majek, M. et al. (2018). Chem. Eur. J. 24: 105–108. 31 Bardagi, J.I., Ghosh, I., Schmalzbauer, M. et al. (2018). Eur. J. Org. Chem. 2018: 34–40. 32 Fukuzumi, S., Ohkubo, K., Ogo, S. et al. (2004). J. Am. Chem. Soc. 126: 1600–1601. 33 Fukuzumi, S. and Ohkubo, K. (2013). Chem. Sci. 4: 561–574. 34 MacKenzie, I.A., Wang, L., Onuska, N.P.R. et al. (2020). Nature 580: 76–80. 35 Huang, H., Strater, Z.M., Rauch, M. et al. (2019). Angew. Chem. Int. Ed. Engl. 58: 13318–13322. 36 Zhang, W., Carpenter, K.L., and Lin, S. (2020). Angew. Chem. Int. Ed. Engl. 59: 409–417. 37 Capaldo, L., Quadri, L.L., and Ravelli, D. (2019). Angew. Chem. Int. Ed. Engl. 58: 17508–17510. 38 Hong Yan, Z.-W.H. and Xu, H.-C (2019). Angew. Chem. Int. Ed. Engl. 58: 4592–4595. 39 Liu, J., Lu, L., Wood, D., and Lin, S. (2020). ACS Cent. Sci. 6: 1317–1340. 40 Barham, J.P. and Konig, B. (2020). Angew. Chem. Int. Ed. Engl. 59: 11732–11747. 41 Lai, X.L., Shu, X.M., Song, J., and Xu, H.C. (2020). Angew. Chem. Int. Ed. Engl. 59: 10626–10632. 42 Hossain, M.J., Ono, T., Yano, Y., and Hisaeda, Y. (2019). ChemElectroChem 6: 4199–4203. 43 Kim, H., Kim, H., Lambert, T.H., and Lin, S. (2020). J. Am. Chem. Soc. 142: 2087–2092. 44 Cowper, N.G.W., Chernowsky, C.P., Williams, O.P., and Wickens, Z.K. (2020). J. Am. Chem. Soc. 142: 2093–2099.

687

32 Visible Light as an Alternative Energy Source in Enantioselective Catalysis Ana Maria Faisca Phillips1 and Armando J.L. Pombeiro1,2,* 1 Centro de Química Estrutural, Institute of Molecular Sciences, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais 1, Lisboa, Portugal 2 Research Institute of Chemistry, Peoples’s Friendship University of Russia (RUDN University), 6 Miklukho-Maklaya Street, Moscow, Russia * Corresponding author

32.1  Introduction Present day concerns with our environment and with the preservation of natural resources have led chemists to develop more sustainable methods of synthesis (the so-called “green chemistry”) [1]. Energy savings are also important to consider, and, amongst the new developments, photoredox catalysis stands out. Contrary to the classical photochemical reactions, such as the [2+2] cycloaddition of olefins, or the well-known Paterno-Buchi reaction for the synthesis of oxetanes from aldehydes or ketones and olefins which, as a result of molecular orbital symmetry considerations, cannot proceed with thermal activation, but does so in the presence of high energy UV light (100– 380 nm) [2], photoredox catalysis works with mild visible light (400–800 nm). In many cases, the reactions proceed at room temperature, which is an important advantage [3]. The lower energy used avoids decomposition and undesirable side reactions [4]. However, many organic molecules do not absorb visible light, and another solution was needed to implement a visible light activation strategy. The way that nature uses various visible light absorbing chromophores/photocatalysts to convert solar energy to chemical energy, as in photosynthesis, undoubtedly inspired many research groups, and hence the photocatalysts were discovered. Photocatalysts have been now applied not only in organic synthesis, including in proton coupled electron transfer, but also in water splitting, solar energy storage, and photovoltaics [5–8]. Photoredox enantioselective catalysis is a fairly new branch of organic synthesis, and most developments have taken place since 2008. Seminal works in this field which sparkled a significant interest and research in this area were early reports by Bach [9], MacMillan [10], Yoon [11], and Stephenson [12], describing highly valuable single-electron transfer processes in synthesis. In 2005, Bach et al. showed for the first time that an organocatalyst could have the dual role of chirality transfer agent and photocatalyst. A chiral xanthone was utilized to induce chirality in a cyclization reaction leading to pyrrolidine ring-formation, while at the same time it transferred light energy to the substrate (by triplet energy transfer [ET]) to make the reaction possible. This process required, however, radiation from a UV source (366 nm) [9]. MacMillan et al. combined a chiral imidazolidinone with a ruthenium photocatalyst, for a highly enantioselective alkylation of enals Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 3, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

688

32  Visible Light as an Alternative Energy Source in Enantioselective Catalysis

in the presence of visible light, introducing the term metallaphotoredox catalysis [10]. These first two processes are further discussed later in this chapter. The other two reports, although not enantioselective, described new approaches with the use of visible light. Yoon et al. showed for the first time that a [2+2] cycloaddition (of enones) could be made to proceed with visible light irradiation, by using Lewis acid activation [11]. Stephenson et al. described a photoinduced tin-free reductive dehalogenation reaction, based on photoredox catalysis by Ru(bpy)3Cl2 and an amine as hydrogen source [12]. Photodehalogenation is particularly significant for industrial applications, since polyhalogenated compounds are known persistent organic pollutents, which may cause health hazards, being banned by the Stockholm convention [13]. Practical dehalogenation techniques are useful, particularly since the tin hydrides which are otherwise frequently used, have associated toxicity and purification problems. The power of the new methodologies was soon recognized, and nowadays enantioselective photoredox catalysis is a very active area of research, as several recent reviews on the subject show [14–24]. The photocatalysts are very versatile and allow different new forms of reactivity, because besides having the capacity to convert visible light into chemical energy, they can act both as oxidants and as reductants simultaneously in their excited states [15]. To bring about a photoredox reaction, a photocatalyst has to absorb visible light. This promotes an electron from its highest-occupied molecular orbital (HOMO) to its lowest unoccupied molecular orbital (LUMO). The singlet excited-state catalytic species produced undergoes intersystem crossing (ISC) to a long-lived triplet excited-state. This excited species can lose its energy in a number of ways: i) by single-electron transfer (SET); ii) by ET; iii) by atom or group transfer; and iv) via accelerated elementary organometallic steps [23]. This reactivity of the excited state brings up many possibilities for the development of new transformations with advantages over the conventional ground-state catalysis. In addition, the photocatalyst may also accept the excitation energy from a previously sensitized substrate. When a substrate accepts the excitation energy of a photocatalyst, a chemical reaction can start. Figure 32.1 shows the four reaction pathways possible, which depend on the redox potentials of photocatalyst and those of the substrates: The excited state of the photocatalyst may be quenched either oxidatively (a) or reductively (b) [3]. The excited state has a very short lifetime, and hence ΔG ≤ 0 for the process. It may also happen that the photocatalyst is quenched by a sacrificial electron donor, typically an amine or ascorbic acid (c) and the photocatalyst will then oxidize a substrate in a ground-state

Figure 32.1  Mechanistic pathways of photoredox catalysis. In each case, the activated substrate will engage in further chemical reactions [3, 14].

32.1  Introduction

reaction. Finally there is also the possibility of catalyst quenching by an electron acceptor, such as oxygen, (from air or peroxodisulfate) and the catalyst can then reduce a substrate in a ground state reaction (d). When there is subsequently a fast and irreversible reaction step, endergonic electron transfers are possible with ΔG>0 (up to 500 mV). The majority of photoredox catalysis reactions follow the first path (i), starting with single-electron transfer between the excited photocatalyst (photosensitizer) and the organic substrate or reagent. A transient redox-active intermediate is formed, as well as an ion-radical species that reacts further to give products. The intermediate can return to the ground state by another SET event. The overall photoredox process can be considered as being net-oxidative, net-reductive, or redox-neutral system, depending on the number of incoming and outgoing electrons. In the redoxneutral reactions, both oxidants and reductants to the sensitizer are incorporated into the product and there is no waste. The system is also insensitive to the sequence of redox events, since both electron-donors and acceptors co-exist in the same reaction vessel. These characteristics make these processes very versatile. In photoredox chemistry, radicals are generated from nonradical species in a much more controlled manner than in other radical reactions. Because radicals are very reactive and tend to react indiscriminately, providing low selectivity between multiple possible stereo- and regioisomers, high ees can be difficult to achieve, but this is not the case in photoredox catalysis. Low selectivities can also result from the low energy barriers that reactions of odd electron species have [3, 23, 25]. Often the real challenge in these reactions is to overcome the high rate of racemic background reactions [14]. The concentration of reactive radical intermediates remains low and the rate of the unwanted diffusion-controlled radical/radical recombination process is also slow. The photocatalysts may be inorganic complexes or organic molecules. A selection of the common ones is shown in Figure 32.2. The Ru(II)polypyridine complexes, of which the most commonly employed is the commercially available Ru(bpy)3Cl2, are quite useful, due to their ease of synthesis, stability at room temperature, and excellent photoredox properties. To render the photoredox processes enantioselective, a chiral catalyst is used. In some cases the chiral catalyst also plays the role of photocatalyst, but more frequently a dual-catalyst approach is used (a synergistic approach), and the chiral catalyst can be an organocatalyst or a transition metal complex. Since 2008, there have been a vast number of achievements in this field. The aim of this chapter is to highlight the different types of catalytic activation that are possible in combination with photoredox catalysis to bring about an enantioselective transformation. This is of relevance for the development of future reactions. Besides the description of some historical cases that demonstrate advances in the field, we concentrate on examples that highlight the advantages of using visible light as an energy source in relation (as an alternative) to other methods of synthesis. The reactions covered are activated by visible light only (wavelength 400–700 nm) and those driven by UV light are excluded (wavelength 20:1.

32.2  Dual Chiral Organocatalysis and Photoredox Catalysis

Figure 32.4  (a) Asymmetric cross-dehydrogenative coupling (CDC) of aldehydes with xanthenes [41]. (b) Enantioselective inter- and intramolecular α-alkylation of aldehydes using simple olefins [42]. DME = dimethoxyethane; Dmppy = 2,4-bis(3,5-dimethylphenyl)pyridine; Dtbbpy = (4,4ʹ-di-tert-butyl-2,2ʹ-bipyridine).

According to the reaction mechanism proposed, an enamine intermediate (A) is involved and the hydrogen transfer step involves the transfer of a hydrogen atom from B to the alkyl radical species formed after olefin addition (C). This step is by no means trivial, because the HAT catalyst has to be able to discriminate between trapping of an initial electrophilic 3πe – enaminyl radical intermediate (B) or the alkyl radical species generated following olefin coupling. In addition, since the olefin addition is a reversible reaction, the retention or loss of the enantiocontrol gained in the SOMO-addition step is dependent on the kinetic efficiency of the HAT catalyst. Iridium photocatalyst Ir(Fmppy)2(dtbbpy)PF6 (Fmppy = 2-(4-fluorophenyl)-4-(methylpyri­ dine), dtbbpy = 4,4′-di-tert-butyl-2,2′-bipyridine) was found efficient to produce an excited-state

693

694

32  Visible Light as an Alternative Energy Source in Enantioselective Catalysis

complex upon irradiation with visible light (E1/2 red [*IrIII/IrII] = +0.77 V vs SCE). This complex can mediate single-electron transfer from the electron-rich enamine A, resulting from the reaction between organocatalyst 14 and the aldehyde, which generates a reduced Ir complex and the critical 3πe–enaminyl radical species B. In the olefin addition step (C–C bond formation), which leads to secondary alkyl radical C, the stereogenic center is created. Thiophenol 15 (BDES–H =78  kcal.mol–1 [43]) works well as HAT catalyst, and after hydrolysis of the iminium ion, the alkylated aldehyde is released and the organocatalyst is also set free to enter a new catalytic cycle (Figure 32.4b). There are many known examples of iminium ion activation, particularly for the conjugate addition of soft nucleophiles to the β-carbon atom of unsaturated carbonyl compounds. However, to trap nucleophilic radicals in a stereoselective manner has been more problematic. The reason for this lies in the fact when a radical adds to a cationic iminium ion a reactive unstable α-iminyl radical cation is generated, with a high tendency to undergo β-scission and to regenerate the more stable iminium ion. Melchiorre et al. were interested in developing this type of methodology, for the creation of a chiral quaternary carbon center since, contrary to what happens in ionic reactions, the reactivity of radicals is only marginally affected by steric factors [44]. Nevertheless, examples of this type of approach for the construction of quaternary chiral centers were still very limited [16]. The radical conjugate addition (RCA) between cyclic enones 17 and a variety of benzodioxole-derived radicals or alkyl aryl amines 18 afforded products in high yields and very high ees via the cooperative catalysis of photoredox Ir catalyst Ir[dF(CF3) ppy]2(dtbbpy)PF6 and diamine 19, an organocatalyst bearing a redox-active carbazole moiety (Figure 32.5a) [45]. In this reaction the organocatalyst operates via the formation of iminium ions A and is capable of trapping the photochemically generated carbon-centered radicals stereoselectively by means of an electron-relay mechanism. The carbazole moiety can undergo a rapid intramolecular SET reduction of the unstable α-iminyl radical cation B generated initially yielding C, which prevents it from breaking down. The nascent enamine intermediate tautomerizes to the more stable imine, thus avoiding a possible competitive back-electron transfer. The carbazole radical cation then undergoes single-electron reduction from the reduced photoredox catalyst, regaining its neutrality, while yielding the quaternary product. Hence the photocatalyst not only creates the nucleophilic radical but it also promotes the final redox process, which is in fact the turnover-limiting step of the overall reaction. The decarboxylation of aliphatic acids or activated esters is a means of generating highly reactive alkyl radicals for the formation of C–C bonds. Recently the unprecedented enantioselective α-alkynylation of β-ketocarbonyl compounds 21 with propiolic acids 22 was achieved by photocatalytic decarboxylation with visible light and a chiral primary amine catalyst (Figure 32.5b). This transformation, developed by Luo et al. had, as distinctive feature, the intermediacy of an α-imino radical, which could be generated with a photoredox catalytic system involving Ru(bpy)3Cl2 and hypervalent iodine reagent (BI-OH) [46]. The oxidative reagent had been previously used for photochemical decarboxylation [47, 48] but not in enantioselective reactions of this type. In the absence of this dual reagent system no alkynylation took place. DFT calculations suggested that this N–H bond has a pKa in the case of radical cation 26, which suggests that the transfer of a proton to generate the α-imino radical is a facile process. A wide range of phenylpropiolic acids bearing either electron-donating or electron-withdrawing groups at the meta or para position of the arene moiety yielded alkynylation products bearing all-carbon quaternary centers in moderate yield and excellent ees, together with the corresponding alkylation products, which were obtained in low yields and moderate to good ees. The β-ketocarbonyl compounds could be cyclic or acyclic.

32.2  Dual Chiral Organocatalysis and Photoredox Catalysis

Figure 32.5  (a) Enantioselective 1,4-addition of α-amino radicals to enones [45]. (b) Enantioselective decarboxylative α‑alkynylation of β‑ketocarbonyl compounds [46]. (c) Asymmetric α-acylation of tertiary amines [49]. DNB = 1,3-dinitrobenzene; dF(CF3)ppy = 2-(2,4-difluorophenyl)-5-(trifluoromethyl)- pyridine.

32.2.2  N-Heterocyclic Carbenes (NHCs) as Catalysts The catalytic activation modes discussed in the previous section were all examples of covalent catalysis. Another example of this type of catalysis is that obtained when chiral NHCs are used as

695

696

32  Visible Light as an Alternative Energy Source in Enantioselective Catalysis

catalysts. The utilization of these catalysts in combination with photoredox catalysis has allowed the enantioselective acylation of tetrahydroisoquinolines (THIQs) 27 with aldehydes 28 as described by DiRocco and Rovis (Figure 32.5c) [49]. The acylated THIQs 30 were obtained in high yields and ees. The role of the photocatalyst In this case is not to generate a reactive radical, but a reactive iminium ion A through the photooxidation of the THIQ with the powerful photochemically generated oxidant [Ru(bpy)3]3+ (1.29 V vs SCE). In this way, single electron abstraction takes place followed by hydrogen abstraction in the presence of m-dinitrobenzene (m-DNB), acting as oxidative quencher. The imine is then trapped by the nucleophilic Breslow intermediate B formed by the reaction between the NHC catalyst and the aldehyde to yield C. Elimination of the NHC yields the product. This method is compatible with the presence of other functional groups on the aldehyde, namely thioethers, esters, and protected amines, without compromising the yields or ees.

32.2.3  Chiral Phosphoric Acids as Catalysts Chiral phosphoric acids (CPAs) have emerged in the last two decades as highly efficient catalysts for many enantioselective transformations. A significant number of reports involve photoredox catalysis. It may be difficult to distinguish strictly between hydrogen bonding catalysis and Brønsted acid/base catalysis when considering reactions in which these catalysts are involved. If an exchangeable proton is required for the final transformation, even if hydrogen bonding interactions are involved, the latter type of catalysis may be considered to be in operation [21, 50]. The first enantioselective example of cooperative hydrogen bonding/photoredox catalysis was described by Knowles et al. in 2013 [51]. It was an intramolecular enantioselective aza-pinacol cyclization of ketone tethered hydrazones 31, made possible by the action of 1,1′-bi-2-naphthol (BINOL)derived CPA (CPA) 32, Ir(ppy)2(dtbpy)PF6 and Hantzsch ester (HE) 33, used as a stoichiometric reductant (HE is a series of dialkyl 1,4-dihydro-2,6-dialkyl-3,5-pyridinedicarboxylate) (Figure 32.6a). Although ketones are not reduced by PC due to their high negative reduction potential, the method devised involved concerted proton-coupled electron transfer (PCET), which surpassed this problem through the generation of lower energy neutral ketyl radical intermediates (e.g. 35, which underwent C–C bond formation producing 36 followed by reduction to yield the desired products 34) (Figure 32.6a). In PCET, a single electron transfer and a proton transfer occur in a concerted manner. Excellent levels of enantio- and diastereoselectivity were obtained for a range of products. Azaarenes are important structural components of many natural products, pharmaceuticals and agrochemicals. These substances often bear substituents incorporating chiral centers. Minisci reactions are often used in these fields for the formation of carbon-carbon bonds, by joining carbon radicals to carbon centers adjacent to nitrogen in pyridine rings. However, enantioselective approaches had not been described until a 2018 report by Phipps et al. [52]. In this procedure, prochiral radicals, generated from amino acid derivatives 37, are added to nitrogen heterocycles 38, namely pyridines and quinolines (Figure 32.6b). Photoredox catalysis with Ir(dF(CF3)ppy)2(dtbbpy) PF6 and a CPA, (R)-TRIP (3,3ʹ-bis(2,4,6-triisopropylphenyl)-2,2ʹ-binaphtholate) (39) or (R)-TCYP ((11bR)-4-hydroxy-2,6-bis(2,4,6-tricyclohexylphenyl)-4-oxide-dinaphtho[2,1-d:1′,2′-f][1,3,2]dioxaphosphepin) (40) allowed the construction of α-tertiary stereocenters in a highly stereoselective manner producing compounds 41. The method was applied to the regio- and stereoselective latestage functionalization of two pharmaceuticals, namely, metyrapone, used for the diagnosis of adrenal insufficiency, and etofibrate, a fibrate [52]. One year later, Jiang et al. described a procedure for the enantioselective addition of alkyl or ketyl radicals to 2- and 4-vinylpyridines 42 also with a combination of Brønsted acid catalysis and photoredox catalysis, although in this case an organic photocatalyst, 5,6-bis(5-methoxythiophen-2-yl)

32.2  Dual Chiral Organocatalysis and Photoredox Catalysis

Figure 32.6  (a) Intramolecular enantioselective aza-pinacol cyclization of ketone-tethered hydrazones [51]. (b) Enantioselective Minisci-type addition to heteroarenes [52]. (c) Enantioselective reductive coupling of vinylpyridines with aldehydes, ketones, and imines [53, 54]. (d) Asymmetric synthesis of heterocyclic γ-amino-acid and diamine derivatives by three-component radical cascade reactions [55]. Dtbpy = 4,4ʹ-ditert-butyl bipyridine.

pyrazine-2,3-dicarbonitrile (DPZ) was used (Figure 32.6c) [53]. Chiral γ-phenyl and aminopropylpyridines 46 were obtained with very high yields and ees, in spite of the much longer distance between the α-amino radical and the nitrogen-coordinated chiral catalyst. In this reaction, the organocatalyst is axially chiral 1,1′-spirobiindane-7,7′-diol (SPINOL)-CPA 44, which activates the pyridine and the C=X bond (SPINOL refers to a series of axially chiral 1,1′-spirobiindane-7,7′diols). The α-aminoalkyl and ketyl radicals are generated in situ from the corresponding ketones, aldehydes, and imines, undergoing subsequently conjugate addition to vinylpyridine. 2-Vinyl pyridines with very electron-withdrawing ring substituents, did not react. A reductive coupling strategy was used in this synthesis, with HE 45 playing the role of stoichiometric reductant [53, 54].

697

698

32  Visible Light as an Alternative Energy Source in Enantioselective Catalysis

A further development on Minisci-type reactions was the three-component reaction between azaarenes 47, enamides 48 and α-bromo carbonyl compounds 49, to yield valuable chiral heterocyclic γ-amino acid derivatives 50 in high yields and enantioselectivities (Figure 32.6d) [55]. This Ir-CPA-catalyzed cascade reaction described by Zheng and Studer has very broad substrate scope, being compatible with quinolines, pyridines, and several quinolines with an additional fused benzene ring. Photoredox catalysis in combination with organocatalysis has also been used to bring about radical cross-coupling reactions. This is not a trivial matter when enantioselective reactions are involved, due to high reactivity of radicals and the possibility of strong racemic background reactions [50]. Chemoselectivity could be a problem, with homocoupling competing [56, 57]. However, this approach has been successfully used a number of times. Jiang et al. obtained β-amino alcohols in which the amino group is primary and the alcohol tertiary, by coupling N-aryl glycines 51 with acyclic and cyclic activated ketones (1,2-diketones (53) and N-Boc isatins (56)) (Figure 32.7a) [58]. SPINOL-CPAs 53 and 55, in combination with DPZ, performed well and the products were obtained in high yields and ees. The reactions were compatible with the formation of heteroquaternary stereocenters. This example represents the first catalytic asymmetric method ever reported for the synthesis of enantioenriched 1,2-amino tertiary alcohols. The rational for the reaction design was based on the fact that product formation could probably be achieved via radical coupling (Figure 32.7b). Ketyls A can be generated directly from 1,2-diketones with DPZ e.g. E1/2 red (benzyl) = −1.169 V and −1.251 V vs. a SCE in CH3CN and E1/2 red (DPZ) = −1.45 V; vs. SCE in CH2Cl2. In addition N-aryl glycines can be precursors of α-amino radicals that can generate radical species B via single electron oxidative decarboxylation by the visible-light-activated DPZ (N-phenyl glycine, carboxylate ion: Ep = +0.52 [for the N moiety] and +1.09 V [for the COO− moiety] vs. Ag/AgCl in CH3CN). A similar radical approach was used by Jiang et al. to couple N-aryl amino acids and α-halo carbonyl compounds [59] and to couple 3-chlorooxindoles with N-aryl glycines [60]. A similar SPINOL-CPA 58/DPZ system could be used to couple N-aryl glycines to a range of α-branched 2-vinylpyridines 59 yielding, by decarboxylation, conjugate addition and enantioselective protonation, γ-chiral aryl amines 60, in high yields and ees (Figure 32.7c) [61]. The procedure could also be applied with 2-vinylquinolines with SPINOL-CPA 61 instead of 58. The versatility of the DPZ/SPINOL-CPA system was further demonstrated when it was utilized for dehalogenation-enantioselective D2O deuteration of racemic α-halogenated azaarenes (Figure 32.7d) [62]. A large range of α-chiral-α-deuterated quinolones 62 and related heterocycles (benzothiazole, phenanthidrine, pyridine) could be obtained with 80–99% ee and with up to 95% incorporation of deuterium, using this procedure, although with the pyridine-based system the ee was lower (53%). In this process an HE (64) acts as the terminal reductant. The reduction-deuteration of azaarene-substituted ketones was also feasible. 2,2-Disubstituted indolin-3-ones, structural entities found present in many alkaloids, may also be obtained via cooperative photoredox catalysis and chiral CPA catalysis. In 2014, Xiao et al. [63] reported the first example of an enantioselective approach to these substances via a cascade aerobic oxidation and semipinacol rearrangement of 2-aryl-3-alkyl-substituted indoles 66 (Figure 32.7e). Ru(bpy)3Cl2•6H2O and 3,3ʹ-anthracyl-substituted CPA 67 provided the activation needed for this transformation in the presence of dioxygen and white light emitting diodes (LEDs). In these transformations, mechanistically, the superoxide anion radical adds to a radical cation A produced from the indole via single-electron oxidation, to yield species B. Proton transfer and O–O cleavage give rise to a tertiary alcohol C which undergoes the semi-pinacol rearrangement to yield the final products 68. This chemistry was further explored by Jiang [64]. The mild reaction

32.2  Dual Chiral Organocatalysis and Photoredox Catalysis

Figure 32.7  (a) Enantioselective radical coupling of activated ketones with N-aryl glycines [58] / Royal Society of Chemistry / CC BY 3.0. (b) The mechanism proposed for the radical coupling [58] / Royal Society of Chemistry / CC BY 3.0. (c) Enantioselective conjugate addition−enantioselective protonation of N‑aryl glycines to α‑branched 2‑vinylazaarenes [61]. (d) Enantioselective α-deuteration of azaarenes with D2O [62]. (e) Enantioselective synthesis of 2,2-bisubstituted indolin-3-ones by dual photoredox/chiral phosphoric acid (CPA) catalysis [63, 64]. (f) Asymmetric α‑coupling of N‑arylaminomethanes with aldimines [65]. BArF = [3,5-(CF3)2C6H3]4B; CPME = cyclopentyl methyl ether; dmphen = 2,9-dimethyl-1,10-phenanthroline; DIPEA = diisopropyl ethyl amine; dmphen = 2,9-dimethyl-1,10-phenanthroline; TBAPNP = tetrabutylammonium 4-nitrophenoxide; tbpb = tetra-n-butylphosphonium bromide; TIPS-EBX = 1,2-benziodoxol-3(1H)-one.

699

700

32  Visible Light as an Alternative Energy Source in Enantioselective Catalysis

conditions and green oxidizing agent (dioxygen) make this a valuable method to obtain these substances, previously only attained by a stepwise route with harsher oxidizing agents and reaction conditions. Jiang et al. found that the reactions could be conveniently performed under normal atmospheric conditions (in air) if DPZ was used as catalyst with CPA 69 and obtained a large range of 2,2-disubstituted indolin-3-ones 70 in high yields and ees. Córdova et al. had previously reported an early example of the utilization of dioxygen as a reagent, in an enantioselective α-oxygenation of aldehydes co-catalyzed by a chiral secondary amine and tetraphenylporphyrin as photocatalyst [66]. An intermediate peroxide was obtained which could be reduced in situ with NaBH4 to afford the corresponding enantioenriched diols. CPAs have not been the only Brønsted acids explored for cooperative catalysis under photoredox conditions [50]. Ooi et al. developed a P-spiro chiral arylaminophosphonium barfate 71 that made possible a highly enantioselective α-coupling between aldimines 72 and N‑arylaminomethanes 73 (Figure 32.7f) [65]. The resulting chiral 1,2-diamines 74 were obtained in high yields. A reaction mechanism involving chiral cation-directed catalysis was proposed, with ion-pairing between the chiral tetraaminophosphonium ion and an α-amino anion-radical A generated photochemically by the action of iridium catalyst [Ir(ppy)2(bpy)]BarF being responsible for the enantiocontrol obtained. This fact was confirmed when a reaction was performed with tetrabutylammonium salt (Bu4N·BArF), which lacks the H-bond-donor ability, in place of the chiral catalyst 71·HBArF under identical photoredox conditions in the presence of a photocatalyst, which yielded mainly the corresponding meso product. The utilization of nonionic Brønsted acids such as the phosphoric acid diester, benzoic acid, and the 3,3´-Ph2-BINOL as H-bond donors was also ineffective and the meso product was obtained as the major product along with other unidentified side products, reinforcing the idea that the ionic H-bond donor character is a prerequisite for catalysis in this case.

32.2.4  Miscellaneous CPAs have been utilized in catalysis both in their acid forms and as Brønsted bases. The latter approach was utilized by Knowles et al. to obtain from tryptamines 75 the three-ring-fused system of pyrroloindolines 76 via the formation of indole radical cation intermediates (Figure 32.8a) [67]. Catalysis in this case takes place via PCET, with the neutral free radical intermediates existing as H-bonded adducts with a chiral Brønsted base. The 2,2,6,6,-tetramethylpiperidine 1-oxyl radical (TEMPO) worked well as the radical source. Several N´-Cbz-protected tryptamines were converted into pyrroloindolines in a highly enantioselective manner when H8-TRIP phosphate (77) was used as catalyst, by interception of a TEMPO radical in a sequential single-electron oxidation to A, radical coupling to B and addition processes. TIPS-EBX (1,2-benziodoxol-3(1H)-one) is used as the sacrificial oxidant and proton acceptor. Although TIPS-EBX is a poor one-electron oxidant (E1/2 = −1.12 V vs SCE), and it cannot react directly with the indole by electron transfer reactions, it can do so easily with the Ir(ppy)3 photocatalyst in its excited state (*E1/2 = −1.73 V vs SCE). The carboxylate base produced in this process is the stoichiometric acceptor for the proton liberated in the catalytic cycle, a step which prevents the formation of TEMPO-H. Interestingly, the products of these reactions, the TEMPO-substituted pyrroloindolines 76, could be reacted further forming carbocation intermediates via a catalytic single-electron oxidation/mesolytic cleavage under the influence of the iridium catalyst [Ir(dCF3Me-ppy)2(dtbbpy)]-PF6 and blue LEDS and undergo a substitution reaction with a variety of nucleophiles. A range of alkaloid natural products, including (–)-calycanthidine, (–)-chimonanthine and (–)-psychotriasine were obtained via this route in four steps from tryptamine 75. Brønsted base (phosphate base) catalysis in combination with photoredox catalysis and HAT also made possible the deracemization of a racemic urea (±78) (Figure 32.8b) [68]. Upon photooxidation to A and proton transfer, an achiral radical intermediate (B) is generated, which undergoes

32.2  Dual Chiral Organocatalysis and Photoredox Catalysis

Figure 32.8  (a) Enantioselective synthesis of pyrroloindolines [67]. (b) Light-driven deracemization of ureas [68]. (c) Enantioselective photoredox dehalogenative protonation [72] / Royal Society of Chemistry / CC BY 3.0. (d) Isothiourea-catalyzed enantioselective addition of 4‑nitrophenyl esters to iminium ions [73]. (e) Enantioselective aerobic oxidative C(sp3)–H olefination of amines [75]. NaBArF = sodium tetrakis[(3,5trifluoromethyl)phenyl]borate; TBADT = tetrabutylammonium decatungstate.

enantioselective HAT when a chiral thiol catalyst is used. Ees up to 88% were obtained for compounds 81 when the method was applied to several ureas [68]. Chiral tertiary α-haloketones are important building blocks for numerous biologically active substances [69]. Direct halogenation can be used, although enantioselective methods for α-chlorination

701

702

32  Visible Light as an Alternative Energy Source in Enantioselective Catalysis

and α-fluorination of aliphatic ketones have been developed relying on enamine catalysis, they are not compatible with aromatic variants, because of the slow enamine formation and the formation of approximately equimolar enamine rotational isomers [70, 71]. In 2019, an alternative was developed, based on enantioselective photoredox dehalogenative protonation (Figure 32.8c) [72]. This method allowed a range of cyclic and acyclic ketones with labile chiral secondary C–F, C–Cl and C–Br bonds at the α-position to be obtained in high yields and good to excellent ees 83 from dihalogenated compounds 82. DPZ was used as photosensitizer in combination with a chiral H-bonding catalyst, a l-tert-leucine-based squaramide amine 84 under visible light. 2,3-Dimethyl-1,2,3,4tetrahydroquinoxaline (85) acted as the terminal reductant in this case. The synergistic combination of chiral isothiourea/photoredox catalysis was utilized by Smith et al. to couple THIQ (27) with aryl esters (86) via the generation of C1-ammonium enolates (Figure 32.8d) [73, 74]. The esters obtained are converted in situ to amides by reaction with amines and the amide-functionalized THIQs 88 are obtained in high yields and excellent ers. According to the mechanism proposed, the THIQs are initially oxidized to iminium salts A under photocatalytic conditions. The Lewis basic isothiourea is acylated by the active aryl ester to B with release of a nucleophilic aryloxide. Enolization of this intermediate results in the formation of an ammonium enolate C, conformationally locked due to an nO to σ*C−S interaction between the enolate O and the catalyst 87. This interaction eventually determines the enantioselectivity, since addition takes place preferentially anti- to the phenyl stereodirecting group. Nucleophilic attack of the aryloxide on the acyl ammonium, gives the β-amino ester product D, releasing at the same time the catalyst for a new cycle. Tetrabutylammonium 4-nitrophenoxide (TBANP) increases the catalytic activity of the isothiourea Lewis base catalyst and increases the polarity of the reaction. The main challenge in this approach was to select an electrophilic reaction partner (in this case the tetrahydroisoquinolinium) to react with the catalytically-generated C(1)-ammonium enolate, which was at the same time compatible with the nucleophilic tertiary amine catalyst as well as with the aryloxide. Lewis base catalysis was also combined with PC to promote a CDC between THIQs and enals 89 (Figure 32.8e). DPZ was used as photosensitizer, cinchona alkaloid derivative β-isocupreidine (β-ICD) (90) as chiral catalyst and air (dioxygen) worked well as oxidant. High yields and ees of α-alkenylated products 91 were obtained for a large range of substrates [75, 76]. The enantioselective aerobic oxidative C(sp3)−H olefination of tetrahydro-β-carbolines was accomplished under similar conditions [75]. Another interesting example of enantioselective oxidative C–H functionalization of THIQs was their reaction with silyl ketene acetals. β-enamino esters were produced in high yields and ees when photoredox catalysis with [Ru(bpy)3Cl2] was combined with asymmetric anion-binding catalysis obtained with an amide-based thiourea [77]. Cinchona alkaloids were also utilized by Melchiorre et al. as phase transfer catalysts for the asymmetric perfluoroalkylation of β-ketoesters. Quaternary cinchonine ammonium salts 92 and 93 allowed the reaction to be performed under heterogeneous conditions [78].

32.3  Metal Catalyzed Processes 32.3.1  Dual Transition Metal/Photoredox Catalysis The dual approach utilizing a metal and a photoredox catalyst has been named metallaphotoredox catalysis by MacMillan. Nickel, copper, chromium and more rarely palladium, are the main metals utilized in a synergistic approach with photoredox catalysts [15, 18]. The chirality is provided by chiral ligands, and hence efficient ligand design plays an important role in this area. Besides the

32.3  Metal Catalyzed Processes

ligand, the oxidation state, as well as catalyst excitation, can be used to modulate the reactivity of the metal. The reactions highlighted in this section are distinct from those in which the metal plays simply the role of a Lewis acid, which are discussed later [79]. Metallaphotoredox catalysis is particularly useful for C(sp3) fragment couplings, traditionally challenging to achieve with transition metals alone [18]. Nickel catalysis works well in these cross couplings, because of the ability of nickel catalysts to undergo facile oxidative addition with alkyl electrophiles at room temperature, their favorable single-electron redox potentials, and the lower susceptibility to undergo β-hydride elimination steps [18]. Traditionally cross-coupling reactions have been achieved with two-electron transmetallation reactions, whereas the dual nickel/photocatalytic approach allows cross coupling to take place via single electron transmetallation. Pioneers in this area were Molander et al. [80] and Doyle, MacMillan et al. [81], who published works in this field in 2014. In the first example, potassium alkoxyalkyl- and benzyltrifluoroborates were crosscoupled with a range of aryl bromides at rt, in the absence of a strong base and under visible light, to yield arylated products in good yields. Only one example of an enantioselective coupling was reported, of a secondary benzyltrifluoroborate (94) with 95 and with the chirality provided by chiral ligand 96, but the ee of cross-coupled product 97 was only moderate (Figure 32.9a). Doyle et al. and MacMillan et al. described a similar cross-coupling reaction proceeding with Ni/photoredox co-catalysis, but employed carboxylic acids instead of organoboron reagents [81]. This procedure was extended to a highly enantioselective decarboxylative Csp2–Csp3 cross-coupling with N-Boc-protected amino acids 98 and compounds 99 through the use of chiral cyano-bisoxazoline ligand 100 (Figure 32.9b) [82]. This reaction has good functional group tolerance, allowing the presence of carbamate, ether, ester, alkyl chloride, carbonate, indole, and thiophene substituents in a large range of amino acids and aryl halides (or halo-substituted pyridines), with the α-arylated α-amino acids 101 being obtained with ees as high as 96%. This procedure replaces the organometallic reagents that are traditionally used as coupling partners with amino acids. According to the reaction mechanism proposed, photocatalyst-mediated oxidation and decarboxylation of a α-amino acid produces a prochiral α-amino radical A. The aryl halide is activated by oxidative addition, which results in the formation of a Ni(II)−aryl complex B, that reacts with the newly generated α-amino radical. The resulting diorganonickel(III) adduct C undergoes reductive elimination producing the desired coupled product. The enantioselective nickel/photoredox catalyzed asymmetric reductive cross-coupling of racemic α-chloro esters 102 with aryl iodides 103 was reported for the first time by Mao, Walsh et al. in 2020 (Figure 32.9c) [83]. The best photoredox catalyst for this transformation was 1,2,3,5-tetrakis(carbazol-9-yl)-4,6-dicyanobenzene (4CzIPN) (104) and an organic reductant (HE 106) was used, whereas most reductive cross-coupling reactions use stoichiometric metals. α-Aryl esters 107 were obtained in this case in good yields and high ees. The highly enantioselective cross-coupling of alkyl benzenes 108 and aryl bromides 99 by photoredox/nickel dual catalysis was made possible with chiral sterically hindered biimidazoline ligands 109 by Lu et al. (Figure 32.9d) [84]. In this C(sp3)−H bond arylation functional group compatibility was achieved under mild conditions and external redox reagents were not required. A reaction mechanism was proposed based on mechanistic studies and previous findings (Figure 32.9e). An initial oxidative addition of the aryl bromide to in-situ generated Ni(II) complex A produces B which undergoes single-electron oxidation induced the photoexcited Ir catalyst to give C and an active bromine atom. Ir is reduced to Ir(II) simultaneously. The occurrence of an electron transfer process in the initiation of the bromine-free radical was also considered a possibility. The benzylic radical was formed, by an HAT process, from the bromine free radical (BDEHBR = 366 kJ/ mol, BDEEtPh = 357 kJ/mol, BDEdioxane = 406 kJ/mol) with 4,4′-dimethoxybenzophenone (DMBP)

703

704

32  Visible Light as an Alternative Energy Source in Enantioselective Catalysis

Figure 32.9  (a) Enantioselective coupling of α-carboxyl sp3-carbons with aryl halides [81]. (b) Enantioselective decarboxylative arylation of α‑amino acids [82]. (c) Asymmetric reductive cross-coupling of racemic α-chloro esters with aryl iodides [83]. (d) Enantioselective benzylic CH arylation [84] / Springer Nature / CC BY 4.0. (e) The mechanism proposed for the arylation reaction shown in (d) [84] / Springer Nature / CC BY 4.0. (f) Synthesis of 2,2´disubstituted indoles by enantioselective nickel-photoredox catalysis [85]. COD = 1,5-cyclooctadiene; TBAI = tetrabutyl ammonium iodide; CyNMe = N,N-dimethylcyclohexylamine; 4CzlPN =1,2,3,5-Tetrakis(carbazol-9-yl)-4,6-dicyanobenzene.

32.3  Metal Catalyzed Processes

as a co-catalyst, being trapped by a aryl Ni(II) species C to afford Ni(III) complex D (BDE = bond dissociation energy). Reductive elimination gives rise to the chiral 1,1-diaryl alkanes and a Ni(I) complex E, which can undergo single-electron reduction to regenerate Ni(0) species A and the photocatalyst. Wang et al. developed a procedure for the synthesis of 3,3´-difunctionalized 2-oxindoles [85] based on an asymmetric difunctionalization of alkenes, merging nickel and PC. In this process aldehydes were used as an acyl source via tetrabutylammonium decatungstate (TBADT) catalyzed acyl C–H activation, which was combined with nickel catalysis in the presence of a monophosphine-oxazoline chiral ligand 112 (Figure 32.9f). The desired products 113 were obtained in high ees [85]. This was the first example of dual HAT photochemistry and asymmetric transition metal catalysis in alkene difunctionalization. High ees were obtained for a range of oxindoles produced containing a quaternary stereogenic center. This process, however, required light with a wavelength just under the visible region, 390 nm. The combination of photocatalysts with copper catalysts has been used to incorporate functional groups into organic molecules, but C–C bond forming reactions have also been particularly successful, and more specifically cyanation reactions. A reactive chiral copper(II) cyanide can be used to trap benzylic radicals enantioselectively, and in this way optically pure benzyl nitriles can be synthesized efficiently. In 2016, this was demonstrated by Liu et al. via single electron oxidation achieved using an electrophilic F+ reagent (NFSI = F–N(SO2Ph)2) as oxidant and as chiral catalyst a copper-bisoxazoline complex [86]. More recently Lin, Liu et al. found that N-hydroxyphthalimide (NHP) esters 114 could be used as sources of benzylic radicals in enantioselective decarboxylation reactions without employing electrophilic oxidants, if the reaction was performed under a visible light source in the presence of a photocatalyst (Figure 32.10a) [87]. The reactions ran under mild conditions (rt), with trimethylsilyl chloride (TMSCl) as cyanide source, had broad substrate scope, and high yields and high ees were obtained with a Cu-115 complex as a catalyst for products 116. The reaction could also be conducted on a 270 mmol scale, and it was applied to the synthesis of key intermediates for the synthesis of the chiral antidepressant (R)-Phenibut and weight-losing drug (R)-Lorcaserin. In the mechanistic proposal (Figure 32.10b), the excited photocatalyst transfers an electron to NHP, initiating the decarboxylation to A and the formation of L*Cu(II)CN. Radical rebound of the benzylic radical with L*Cu(II) (CN)2 and reductive elimination, yields the product 116 and regenerates the catalyst. The enantioselective radical cyanoalkylation of styrenes 117 has also been described [88]. In this copper-catalyzed reaction the alkyl group radicals are produced from NHP esters as in the previous methods, with dual copper/photoredox catalysis and TMSCN as a cyanating agent (Figure 32.10c) [88].The method, which represents the first example of photoredox and copper-catalyzed radical difunctionalization of alkenes, is compatible with primary, secondary, and tertiary alkyl substituted NHP esters. The products 119 were obtained with high to excellent ees. A further development was the reaction of N-alkoxypyridinium salts with readily available silyl reagents (TMSN3, TMSCN, TMSNCS), which led to δ-azido, δ-cyano, and δ-thiocyanato alcohols in high yields [89]. The products were obtained via a domino process involving alkoxy radical generation, 1,5-hydrogen atom transfer (1,5-HAT) and copper-catalyzed functionalization of the resulting C-centered radical. It was shown that an enantioselective approach was also possible, and δ-C(sp3)–H cyanation was achieved with TMSCN and a mixture of copper acetate, chiral ligand ent-115 and fac-Ir(ppy)3 under irradiation by blue LEDs. 4-Phenyl-2-cyanobutanol was obtained in this way in 74% yield and 86% ee from a N-alkoxypyridinium salt of 4-phenylbutanol. The Nozaki-Hiyama-Kishi (NHK) reaction is the chromium catalyzed allylation of aldehydes. NiCl2 is used as a co-catalyst. The reaction has excellent chemoselectivity for aldehydes over ketones or esters and the anti diastereomer is obtained regardless the double bond geometry of the original allylic C–C double bond. One drawback is the stoichiometric amounts generation of toxic chromium

705

706

32  Visible Light as an Alternative Energy Source in Enantioselective Catalysis

Figure 32.10  (a) Enantioselective decarboxylative cyanation [87]. (b) The mechanism proposed for the decarboxylative cyanation [87]. (c) Enantioselective radical cyanoalkylation of styrenes [87]. (d) Cyanoalkylation of unactivated alkenes [92] / Royal Society of Chemistry / CC BY 3.0. (e) The mechanism proposed for the cyanoalkylation of alkenes [92] / Royal Society of Chemistry / CC BY 3.0. (f) Dialkylation of 1,3-dienes by dual photoredox and chromium catalysis [93]. (g) Enantioselective allylic alkylation with 4‑alkyl-1,4-DHPs [94]. (h) Enantioselective [2+2] photocycloadditions of enones using visible light [95]. CFL = compact fluorescent light; Dba = dibenzylideneacetone; NMP = N-methyl-2-pyrrolidinone.

32.3  Metal Catalyzed Processes

waste. Related reactions employ excess of manganese as a terminal reductant [90]. In 2018, Glorius et al. developed a redox neutral diastereoselective procedure using dual chromium/photoredox catalysis, which expands the scope of the NHK reaction. One enantioselective example was tried but the ee was only moderate [91]. Masamune et al. subsequently developed an enantioselective version of the chromium-photoredox process for alkenes 120 using a chiral bisoxazoline ligand (ent-115) and an acridinium (acr) photocatalyst (121 or 122) (Figure 32.10d) [92]. An interesting aspect of this work is that it reports the unprecedented use of unactivated hydrocarbon alkenes (e.g. 120a) as precursors to chiral allylchromium nucleophiles for the asymmetric allylation of aldehydes. The homoallylic alcohols produced were obtained with a diastereomeric ratio >20/1 and up to 99% ee. According to the mechanism proposed (Figure 32.10e), a cationic allyl radical A is produced from the alkene via electron-transfer oxidation of the π-bond by a photoexcited electron-donor substituted acr photocatalyst. After deprotonation to B it is intercepted by the reduced chiral chromium(II) catalyst II to give chiral allyl chromium(III) complex III. The new complex reacts with an aldehyde via a six-membered chair transition state to produce an enantiomerically-enriched chromium alkoxide IV in a syn-selective manner. After protonation, the desired product is obtained, as well as an oxidized Cr(III) complex, regenerated by the reduced form of the photocatalyst. Subsequently Glorius et al. reported that Cr(III) catalysis with chiral bisoxazoline ent-118 and the same photocatalysts could be used for the regio- and diastereoselective three-component dialkylation of 1,3-dienes 124, with aldehydes 125 and alkyl HEs 126 for the synthesis of homoallylic alcohols 127 [93]. The ees obtained in this case were also very high (Figure 32.10f). Although palladium is proven to be one of the most useful metals for catalytic applications, so far there have been very few reports of enantioselective reactions in which Pd is involved in dual catalysis with a photoredox catalyst. Zhang et al. developed a dual catalytic process, in which alkyl radicals generated from 4-alkyl-1,4-dihydropyridines (DHPs) 128 act as the coupling partners to π-allyl palladium complexes bearing chiral phosphine ligands (e.g. 130) (Figure 32.10g) [94]. A variety of allyl esters 129 are compatible with this procedure, that expands the scope of the traditional Pd-catalyzed asymmetric allylic alkylation reaction and is at the same time an alternative and potential complement to the same. The allylic alkylated products 131 were obtained with high ees, as the chiral phosphine ligand controlled the facial selectivity of π-allylpalladium complex during alkyl radical addition.

32.3.2  Dual Chiral Lewis Acid/Photoredox Catalysis The utilization of Lewis acids in photoredox catalysis is in great part linked to the cycloaddition reaction [2+2] [18]. Cycloaddition reactions have been known for a long time, with ultraviolet radiation providing the required activation. Enantioselective variants with substoichiometric amounts of catalysts proved to be very difficult to develop, mainly because uncatalyzed background photochemical processes (of catalyst-unbound substrate) were difficult to control. Free substrates produced racemic products or afforded low ees, unless the racemic reaction could be slowed down. Bach et al. devised a solution to this problem, designing reactions in which the catalyst-substrate complex absorbed light at longer wavelengths than the free substrate. A chiral hydrogen-bonding xanthone-based photosensitizer [96, 97] was utilized or a chiral Lewis acid catalyst [98], capable of inducing a bathochromic shift in the bound substrate to 366 nm and in this way high ees were obtained. The methods had a limitation that irradiation with a monochromatic light source that selectively excites the catalyst-substrate complex at a wavelength where absorption by the free substrate is minimized was required for stereocontrol. In addition, with the chiral Lewis acids high catalyst loadings had to be used to overcome the background reactions (typically approx. 50 mol%).

707

708

32  Visible Light as an Alternative Energy Source in Enantioselective Catalysis

A general strategy to eliminate completely the uncatalyzed background photochemistry had not been found until recently. The works of Back inspired Yoon et al., who found a solution in 2014 using dual chiral Lewis acid/photoredox catalysis [95]. A [2+2] cycloaddition reaction of α,βunsaturated ketones 132 (2´-hydroxychalcones) with enones 133 could be performed under visible light irradiation, with 10 mol% of Lewis acid europium(III) triflate as catalyst, a chiral Schiff base ligand (134) and Ru(ppy)3Cl2 as photocatalyst, affording cyclobutanes 135 with high ees (Figure 32.10h). Lewis acid coordination lowers dramatically the triplet energy of the chalcone substrate and the background absorption in this case was not a problem, since the uncatalyzed reaction could not proceed under these conditions. There was also no reaction in the absence of light. Subsequently, Yoon et al. applied similar methodology to the enantioselective [2+2] cycloaddition of 2´-hydroxychalcones and dienes, using scandium triflate as catalyst and a chiral bisoxazoline ligand, and Ru(bpy)3(PF6)2 as photocatalyst and visible light [99]. The products were obtained in high yields (66–84%) and ees (up to 98%) and dr 2–4:1. Later it was shown by Yoon et al. that styrenes could also be reacted with 2´-hydroxychalcones with the same set of catalysts to afford products in high yields and ees [100]. The synthesis of unsymmetrical cyclobutanes by heterodimerization of olefins is still a significant challenge, particularly in an asymmetric fashion, and these contributions represent significant advances. In 2019, the enantioselective [2+2] cycloaddition was extended to the reactions of regular cinnamate esters with styrenes [101]. Formal [3+2] cycloaddition of cyclopropyl ketones 136 with styrenes using a gadolinium Lewis acid catalyst and chiral ligand 137 was also achieved with white light activation (Figure 32.11a) [102]. Products 138 were obtained with high ees. This method expands the available technologies for [3+2] cycloadditions, which previously relied on highly activated “donor−acceptor” cyclopropanes (e.g. cyclopropane diesters [103], vinylcyclopropanes [104]), giving access to densely substituted cyclopentanes in an enantioselective manner. Yoon et al. reported later the conjugate addition reaction of a α-silylamine pronucleophile 139 to a β-substituted Michael acceptor 140 with a related dual catalysis strategy (Figure 32.11b) [105]. The reaction was postulated to proceed via an α-amino radical intermediate. The products 142 were obtained in high yields and ees. One limitation of this method was that for successful Michael addition, the presence of one N-aryl substituent was required. Aliphatic amines underwent protodesilylation, but subsequently there was no addition to the Michael acceptor. Prior to this example there had been only one report of stereocontrol in reactions of α-amino radical intermediates, which was obtained via an intramolecular conjugate addition reaction by Bach et al. using a chiral hydrogen-bonding photosensitizer [9], with pyrrole tethered quinolones, but it was achieved under UV radiation. Xiao et al. utilized a chiral octahedral complex 144 formed in situ, to catalyze the enantioselective RCA reaction of alkyl and acyl radicals derived from DHPs 126 to enones 143 (the Photo-Giese reaction) under visible light activation to obtain products 145 (Figure 32.11c). With alkyl DHPs, Fukuzumi’s acr photocatalyst Mes-Acr+ was required as well. However, with acyl DHPs, no additional external photocatalyst was required [106]. Other metals are compatible with PC and have been utilized as chiral Lewis acids, not only transition metals like scandium [105] and copper [107–109], but also aluminum [110]. A more detailed coverage is not possible due to space limitations, but the potential for future developments seems evident.

32.4  Chiral Photocatalysts In this section are considered reactions promoted by light and a single catalyst. That is, the chiral catalyst is one that possesses photoredox properties too. They may be either metal complexes chiral at the metal or chiral organocatalysts [111].

32.4  Chiral Photocatalysts

Figure 32.11  (a) Formal [3+2] cycloaddition of cyclopropyl ketones 136 with styrenes [102]. (b) Enantioselective conjugate additions of α‑amino radicals derived from α-silyl amines [105]. (c) Enantioselective conjugate additions of α-amino radicals derived from DHPs to enones [106]. (d) The enantioselective alkylation of 2-acyl imidazoles with benzyl bromides [113]. (e) Visible light-mediated enantioselective [2+2] photocycloaddition of 4-substituted quinolones bearing an olefin in the tether [121]. (f) Intermolecular enantioselective α-alkylation of cyclic ketones with alkyl bromides [126].

709

710

32  Visible Light as an Alternative Energy Source in Enantioselective Catalysis

32.4.1  Chiral-at-Metal Photocatalysts The first catalysts that performed a dual role were described by Meggers et al. in 2014 [112]. The catalysts, bis-cyclometalated iridium(III) and rhodium(III) complexes, contain only achiral ligands. They are Λ- and Δ-enantiomers (left- and right-handed propellers, respectively), with two cyclometalating 5-tert-butyl-2-phenylbenzothiazoles and two acetonitrile ligands, as well as a hexafluorophosphate counterion (e.g. 146–148) (Figure 32.11d; shows the Λ-enantiomers). The propeller shape chiral geometry is provided by the two cyclometalated ligands, and the acetonitriles are labile and can be replaced by substrate molecules, so that they function as Lewis acids. The iridium catalyst and its enantiopode, independently, were capable of catalyzing the enantioselective alkylation of 2-acyl imidazoles 149 with benzyl bromides 150 and phenacyl bromides, by Csp3–H/Csp3–H coupling, affording products 151 in 90–99% ee in the presence of visible light [113]. The 2-acyl-N-methylimidazole substrates tolerate steric, electron donating and electron accepting substituents in the phenyl moiety, but in the alkyl bromide, electron withdrawing substituents were required. In the dark there was no reaction (%98 MB (10 mg/L) removal using 1 g/L of ternary [70] photocatalyst within 120 min under simulated sunlight, photocatalytic degradation efficiency order: α-Fe2O3 > Fe2TiO5 > TiO2.

76% AP (50 mg/L) removal within 180 min of visible-light illumination at the rate constant of 0.01 min–1

88% OTC (30 mg/L) removal within 120 min using 0.8 g/L of photocatalyst at OTC natural pH under visible light

CTN (10 mg/L) complete removal and 80% TOC removal within 75 and 120 min using 0.5 g/L of photocatalyst and 10 mg/L of ozone at natural pH under simulated sunlight

Total BPA (10 mg/L) removal within 30 using 1 g/L of [59] photocatalyst at pH 7 under simulated sunlight

95.6% BPA, 97% SMX, 94% 2,4-DCP, 85% RhB, 50% [81] phenol, and 29% NPX removals using1 g/L of photocatalyst and 0.5 mm PDS within 60 min at pH 6.5 under visible light with reaction constant of 0.006 min–1

80.73% NPX (10 mg/L) removal under sunlight within 120 min at pH 4.

Remarks

Reduced graphene oxide (rGO), ethylene diamine (ED); HPV (H4[PVW11O40]⋅32H2O); single metal atom oxide (SMAO); contaminants of emerging concern (CECs); Effluents contain Organic Matter (EfOM); granular activated carbon (AC); bamboo-derived activated biochar (BAB); peroxydisulfate (PDS).

2O3

Cotinine (CTN)

Magnetic separation; negligible Fe leaching, enhanced adsorption capacity, and effective separation and transfer of e/h.

Graphene-TiO2/Fe3O4

Oxytetracycline (OTC)

Bisphenol A

BiOI/Fe3O4 (molar ratio of Magnetic separation, enhanced stability, effective 2:1) separation of e−/h+ through BiOI/Fe3O4 heterojunction.

e−/h+ recombination inhibition via Ag3PO4-Fe3O4 Bisphenol A (BPA); heterojunction; charge transfer through BAB Sulfamethoxazole (SMX); 2,4-dichlorophenol (2,4-DCP); rhodamine B; Phenol; Naproxen.

Z-scheme Ag3PO4/ Fe3O4-BAB

Naproxen (NPX)

Pollutant

Magnetic separation; enhanced visible-light activity (2.62 eV) due to Cu plasmon absorption, e−/h+ separation, and transfer.

Modification effect

ZnFe2O4/TiO2/Cu

Photocatalyst Modification type

730

33  Heterogeneous Photocatalysis for Wastewater Treatment: A MajorStep Towards Environmental Sustainability

Figure 33.3  The synergistic photocatalytic action of Fe3-xCexO4 and GO for oxytetracycline removal under visible-light irradiation. Ref [32] / Elsevier.

Figure 33.4  Plausible charge transfer route based on the CB potentials of the constituent semiconductors in the ternary Fe2TiO5/α-Fe2O3/TiO2 nanocomposite. Ref [70] / Elsevier.

33.3.1.2  Green Synthesis Routes

A sustainable and cost-effective synthesis route that has engaged enormous attention in recent years is the application of a whole plant or its derivatives as natural sources for reducing, stabilizing, and/or capping purposes [85]. These secondary metabolites have unique properties and are ecofriendly and rich sources of polyphenolic compounds that become strongly involved in the reducing of metals and preparation of metal oxides while confining size and stabilizing the structure [86]. A plasmonic photocatalyst was prepared based on Au and ZnO using the aqueous extract of cumin seeds in the absence of surfactant or any chemical agent [76]. The ions of Au4+ was reduced to zerovalent Au over the ZnO nanoparticles with the aid of phytomolecules of extract. The Schottky contact created at the interface of Au and ZnO surface resulted in higher photoactivity at the visible region coupled with robust stability of the nanocomposite (Figure 33.5).

33.3  Sustainable Photocatalysts

Figure 33.5  Suggested mechanism on the reducing and stabilizing effects of polyphenolic compounds [76] / American Chemical Society.

Application of bioingredients instead of toxic chemicals not only endow a mild and safe environment for driving surface modifications but also is a sustainable way, which reverses the undesired consequences. The fruit extract of black grape was used as the natural source of polyphenols to reduce graphene oxide sheets prior to TiO2 deposition onto the planes (Figure 33.6a) [87]. In this process, the sonically exfoliated GO in ethanol was mixed with grape extract and refluxed for several hours at 80 °C until the brownish colour changed to black, indicating the GO reduction. The resultant powder undergoes a hydrothermal reaction with tetrabutyltitanate solution to give rise to TiO2/rGO nanocomposite. The mechanism through which the GO is reduced by the polyphenolic content of grape fruit extract is represented in Figure 33.6b. The 85.4% decrease in the oxygen content of rGO compared to that of GO confirmed the successful reduction process. Under visiblelight illumination, about 80% of the TOC of bromophenol blue solution was removed over the resultant TiO2 nanocomposite, which was 30% more than that of TiO2 alone. The employed nanocomposite particles were separated by centrifuge, regenerated using a dilute acid solution, dried, and reused for fresh dye solution for six successive photocatalytic experiments. The results revealed the high stability and potential reusability of the photocatalyst.

Figure 33.6  (a) Schematic representation of the synthesis process of TiO2 deposited-grape extract mediated graphene oxide (GRGO); (b) The mechanism of GO reduction via polyphenolic content of grape fruit extract. Reproduced with permission from Ref [87].

731

732

33  Heterogeneous Photocatalysis for Wastewater Treatment: A MajorStep Towards Environmental Sustainability

33.3.2  Carbonaceous Photocatalysts Even though metal oxide-based photocatalysts have great efficiencies for the photooxidation of organic pollutants under sunlight exposure, their application has been restricted due to being expensive in some measure and their surface alteration under some gaseous media [88]. The introduction of carbonaceous materials as nature-based, abundant, economical, and metal-free catalysts opened up a new line towards sustainable photocatalysis. Carbonaceous materials offer a wide array of configurations such as carbon nanotubes (CNTs), carbon quantum dots (GQDs), fullerene, activated carbon, biochar, graphene oxide, and graphitic carbon nitride (g-C3N4) [89]. Over the past decade, there has been a tremendous increase in the studies on the photocatalytic utilization of carbon-based materials. This stems from the fact that carbon-based materials are low-cost, environmentally-benign, well-disposed, and stable candidates for the fabrication of photocatalysts with versatile features. Moreover, they can be prepared from natural precursors for the synthesis of active photocatalytic composites, which is one of the main criteria for sustainable photocatalysis. In this connection, there is a growing body of literature on the synthesis of carbonbased photocatalysts from natural materials. The preparation of CQDs from pear juice through a one-pot hydrothermal procedure [90], CQDs from corncob biomass [91], activated carbon from date seeds [92, 93], activated carbon from basil leaves (Ocimum basilicum) [94], Eriobotrya japonica leaf for biochar preparation [95], and liquefied larch sawdust for fabrication of multi-walled CNTs [96] are just a few from many of the studies on biomass-derived carbon-based materials [97–99]. The development of heterojunction/s in the carbon-based nanocomposites along with the broad visible-light absorption delivers excellent performance in photocatalytic systems. Literature is overflowing with studies on carbon-based nanomaterials with various dimensionality and accordingly unique and specific properties beneficial for photocatalytic applications. Herein, a brief statistic on a number of distinguished carbonaceous materials including carbon dots (zero-dimensional), CNTs (one-dimensional), and graphitic carbon nitride (two-dimensional), which have significant potencies for sustainable photocatalytic decontamination of wastewater is provided as follows. The fluorescent carbon nanoparticles that were noticed during the synthesis and purification of single-walled CNTs, were named CQDs (carbon quantum dots, as mentioned previously). This class of quantum dots was soon become the center of attention for a wide range of applications owing to being environmentally sound, safe, cost-effective, and easy to prepare. Though CQDs were initially synthesized by accident [100], several cost-effective and simple routes have been thereafter developed and modified. In the top-down route, the CQDs are synthesized through structural disintegration of the larger carbonaceous materials such as carbon nanotubes, graphite, or activated carbon [101]. The fabrication of CQDs using carbohydrates or some other molecular precursors via various synthesis methods is called a bottom-up approach [102]. Data on the carbon sources and synthesis methods of CQDs reported in the literature is given in Table 33.2. In general, the CQDs and their modified derivatives are fabricated via microwave heating, ultrasonication, arc-discharge, laser ablation, hydrothermal, solvothermal, and oxygen plasma treatment procedures [103]. The existence of hydroxyl and carboxylic functional groups on the surface of CQDs increases its solubility in aqueous systems and facilities the surface modifications and doping with some other chemical/biological compounds (Figure 33.7a) [104]. In addition to its high solubility in water, the powerful solar-light harvesting properties and effective charge transfer of CQDs have listed this group of carbonaceous materials among the potent candidates for the sustainable photocatalytic treatment of recalcitrant wastewater. The unique electro-optical behavior results from the integration of the quantum confinement effect of small-sized nanoparticles (≤ 10 nm) and intrinsic electrical features of carbon materials [105]. Several properties of CQDs, including

CVD

Hydrothermal

bottom-up

-

L-glutamic acid & m-phenylenediamine

Citric acid/ urea

Glucose

TiO2 NRs/ CNTs

CQDs/TiO2

CQDs (0.98 wt%)/g-C3N4

CQDs/ hydrogenated TiO2

Oil bath reflux

CVD

Toluene/ ethanol

Ti-NTs/CNTs

MB, RhB

Methyl orange (MO)

MO dCQDs < 5 nm dH-TiO2 ≈ 22.2 nm

-

dCQDs ≈ 10‒nm

[MO]0: 20 mg/L; Cat: 1.0 g/L; Visible light (Xe lamp 300 W)

[MB]0=[RhB]0: 15 mg/L; Cat: 0.33 g/L; Visible light (320–780 nm)

[MO]0: 40mg/L; Cat: T: 360 min solar (600 nm)

[MB]0: 25 mg/L; Cat: 0.02 g T: 300 min sunlight.

[108]

[129]

50% MO removal within 25 min under visible light.

(Continued)

[131]

Complete MB and RhB removal [130] within 20 and 110 min; MB and RhB removal rate constants: 0.214 and 0.048 min‒1.

70.56% of MO removal within 360 min using the 1:1 molar ratio of CQDs/TiO2

97.3% MB removal within 300 min

RhB photodegradation rate [111] Hg-Xe lamp (1000 W); −3 Cat: 1 g/L; [RhB]0: 1.6 mg/L: constants of 3.3 × 10 and −2 −1 T: 60 min dark-30 min light. 3.3 × 10 min over Ti-NTs and the NC.

Rhodamine B (RhB) λmax = 554 nm

[127]

Ref.

[128] 78.9% MB removal using 0.1%Au-2%CNT/TiO2 sample within 150 min.

98% MB removal within 150 min over ZnO/CNTs (10 wt.%)

Outcome

lamp (λmax = 254 nm): 66 W/1.37 mWcm−2; MB: 5 × 10−5 M; Cat: 0.3 g/L; T: 150 min.

[MB]0: 50 mg/L; Cat: 1 g/L; T: 60 min dark-240 min sunlight

Operational condition

MB

dCNT ≈ 10‒21 nm; MB

dCNT ≈ 100 nm; Deposition of CNTs into Ti-NTs

dCNT:18 nm; Au and TiO2 deposition on CNTs

MB dCNT: 8–15 nm; LCNT: 30–50 μm;

CVD1/sol-gel

Sonicationassisted sol-gel

Purchased

ZnO/CNTs

Pollutant

Properties

Synthesis method

Au/CNTs/TiO2 Purchased

Carbon source

Nanocomposite

Table 33.2  Carbon-based materials for photocatalytic removal of pollutants.

CNT as hard template

Electrostatic adsorption

dicyandiamide

Melamine

Porous g-C3N4

Protonated g-C3N4/CuS (20%)

dCu2O ≈ 400 nm

dg-C3N4 ≈ 2.0 nm

10–100 nm 139 m2/g Eg = 2.27 eV

-

dg-C3N4 QDs ≈ 4.4 nm

-

Properties

Melamine, SiO2 as a template

Calcination

Outcome

[RhB]0: visible light (λ > 400 nm)

[ACT]0: 10 mg/L Cat: 1 g/L; Visible light λ > 420 nm

[CIP]0: 10 mg/L Cat: 1 g/L; (420–760 nm)

[TOC]0: 27.9 mg/L visible light (λ > 420 nm)

Phenanthrene (PHT)

[TC]0: 15 mg/L; Cat: 1.0 g/L; 400 < λ < 800 nm

[PHT]0: 1 mg/L; Cat: 0.2 g/L; 400–800 nm

[132]

Ref.

4.4-fold enhancement in RhB removal rate and enlarged specific surface area (103.3 m2/g) of porous g-C3N4 vs. bulk g-C3N4 (10.5 m2/g).

98% and 97% aspirin and caffeine removals within 90 min.

92.6% and 28.5% CIP and TOC removals within 120 min of visible light exposure.

Complete TC removal within 30 min [123] under visible light

[137] 76.72% phenanthrene removal within min under visible light at rate constant of 0.355 h‒1.

[136]

[135]

[134]

[125]

65.9% TOC removal within 150 min, [133] CIP removal rate constant: 0.014 min‒1

[Cr]0: 5 mg/L; Complete Cr (VI) photoreduction within 90 min. Cat: 2.5 g/L; Visible light (Xe lamp 300 W)

Operational condition

Staphylococcus Diluted bacterial:160 μL Cat: 99.16% and 98.23% E. coli and S. 40 μL; aureus and aureus removals within 20 min. Escherichia coli xenon lamp: 0.2 W/cm2 bacteria

RhB

Aspirin and Caffeine

CIP

CIP

Cr (VI)

Pollutant

Specific surface TC area: 51.34 m2/g; pore volume: 0.226 cm3/g

Chemical vapour deposition (CVD); nanoribbons (NR); nanotube (NT); physical vapour deposition (PVD);

Mesoporous g-C3N4 N-defect

Supramolecular KOH-modified sunflower straw/ biochar/g-C3N4 Melamine-cyanuric acid self-assembly/ thermal (1:1) poly-condensation

Hydrothermal

Melamine

Hydrothermal

WO3-TiO2@gC3N4

Flake graphite powder

rGO/g-C3N4 QDs

PVD

photoreduction

Melamine

GO/g-C3N4/ BiFeO3

Synthesis method

Ag/P@g-C3N4/ Melamine BiVO4

Carbon source

Nanocomposite

Table 33.2  (Continued)

33.3  Sustainable Photocatalysts

Figure 33.7  (a) The chemical structures of single-walled carbon nanotubes (CNTs) and carbon quantum dots (CQDs). Ref [104] / IOP Publishing. (b) The Vis to UV upconversion phenomenon in CQDs/TiO2 nanocomposite. Ref [108] / IOP Publishing.

photocatalytic action, are in common with other sp2 hybridized carbon materials, which are rich in electrons. Moreover, they contain sp3 hybridized carbon atoms, which implies the presence of irregular π‒electronic conjugation in its structure [106]. Because more than 50% of sunlight is comprised of infrared radiation, the productive utilization of this portion in a photocatalytic system has long been of interest and valued. From this perspective, the upconversion nanomaterials based on quantum dots (QDs) have impressed researchers for their sustainable applications as highly active solar-light-driven photocatalysts. In this system, the solar photons of long wavelengths (near-infrared photons) are absorbed by QDs and transformed into shorter wavelengths with higher energies (UV-vis radiation) that can stimulate the neighboring semiconductor with wide band energy and thereupon launch photocatalytic reactions. This phenomenon takes place via the process of luminescence resonance energy transfer [107]. The upconversion property of CQDs together with photochemical stability and prominent role in separation and transfer of charge carriers has nominated them as one the most favoured sustainable materials for solar-triggered photocatalytic systems. In a study, the hydrothermal synthesis of anatase TiO2 with m-phenylenediamine and L-glutamic acid, as precursors of CQDs, brought about Vis‒UV upconversion CQDs/TiO2 nanocomposite [108]. The resultant photocatalyst could absorb visible light (≈ 600 nm) and convert it into UV photons (≈ 300 to 400 nm) wherein the charge carriers were in turn generated following the TiO2 excitation (Figure 33.7b). CNTs are among the widely studied carbonaceous materials since their discovery. The intriguing properties of CNTs (i.e. nontoxic, large active surface area, high tensile strength, and advantageous electrical and optical qualities) have made them great candidates for miscellaneous applications [109]. Of the routes established for the synthesis of CNTs-based nanocomposites, the direct one-step approach has received increased attention because of the uniform deposition of

735

736

33  Heterogeneous Photocatalysis for Wastewater Treatment: A MajorStep Towards Environmental Sustainability

the metal/metal oxide nanoparticles over the surface of the oxidized CNTs [110]. Opposing processes have been also applied, wherein CNTs are developed over some other materials (e.g. TiO2 nanotubes) [111]. In the present instance, an electrochemical anodization approach was used to prepare nanotubular TiO2 with the aid of Ti foil as the substrate material. Subsequently, a combination of ethanol/toluene was utilized as the carbon precursor and embedded CNTs into the TiO2 nanotubes through a chemical vapour deposition (CVD) process. The deposition of CNTs inside the nanotubular TiO2 rather than the exterior wall was mainly attributed to the inner Lewis acid sites [112]. Considering the influence of the electron-electron interactions on the nature of the gap in semiconducting CNTs [113], they demonstrate eminent efficiencies in photocatalytic systems. g-C3N4 is a rapidly rising star as a 2D polymeric carbon-based semiconductor for removing aqueous pollutants through photocatalysis [114]. The main features of g-C3N4 and corresponding synthesis approaches have been detailed in several review articles [107, 115–117]. The striking bandgap (2.7 eV) and surface properties, appealing optical and structural features, high thermochemical stability, and, more importantly, simple synthesis via thermal polymerization have encouraged researchers as ways to satisfy the sustainability criteria while conducting photocatalytic experiments [118]. However, the photocatalytic activity of pristine g-C3N4 is cramped due to the limited sunlight absorption profile, rapid recombination of charge carriers, and limited electrical conductivity [119]. Wang et al. [120] were the first researchers who reported the photocatalytic application of g-C3N4, almost 175 years from its first discovery in 1834. Afterwards, thousands of studies on the photocatalytic performance of g-C3N4 were carried out to enhance its functioning under solar irradiation. The controlled synthesis for attaining a specific morphology [121], defect induction [122, 123], tuning the bandgap through doping [124, 125], and constructing heterojunction/s with some other semiconductor material/s [50, 126] are some strategies to improve the solar photocatalytic degradation of organic molecules by g-C3N4. In an effort to fabricate a metal-free and highly efficient visible-light-driven photocatalyst, Cui et al. [121] established a cost-effective and simple approach using an in-air chemical vapour deposition (CVD) procedure to grow g-C3N4 microstructures in the form of onion rings (Figure 33.8). Despite the conventionally synthesized techniques, a thin layer of highly uniform g-C3N4 spheres (R-CN) were deposited through this method using a hard template, SiO2 microspheres of 350 nm in size, and a CVD precursor, melamine. This morphologically modified sample represented an outstanding photocatalytic activity under a broader range of visible light owing to improved bandgap (2.58 eV) and simplified separation of photoinduced charges with expanded lifespans.

(a)

(b) −SiO2

Melamine

O2

NH4HF2

CVD SiO2 microspheres

(c)

g-C3N4-SiO2

Oxidation R-CN H2PtCl6 Light

e-

200nm

c

·OH Reduction OH−

g-C3N4-SiO2

Pt/R-CN

h+ h+ h+

e-

e-

·O2

TEOA

150nm

Figure 33.8  (a) Pt-R-CN synthesis procedure. (b) Microscopic images of the samples. (c) Suggested representation for photocatalytic reactions. Reproduced with permission from Ref [121] / American Chemical Society. (Tertiary aliphatic amines [TEOA] as a sacrificial electron donor to stoke up photochemical reactions).

References

Data on some recent studies of solar-active g-C3N4 nanocomposites are presented in Table 33.2. The information on the operational conditions of the photo-oxidation of various organic pollutants through CQDs, CNTs, and g-C3N4 exhibit the sustainable chemistry practices engaged in the synthesis and photocatalysis processes.

33.4  Remarks and Future Perspectives Heterogeneous photocatalysis has shown immense potential for sustainable water decontamination. There are wide ranges of ecofriendly materials that comply with sustainable rules. Most metal oxide semiconductors, especially magnetic nanoparticles, and carbon-based semiconductors gave rise to solar-driven photocatalysts that can be synthesized easily using abundant and nature-based materials. However, in most instances, the electrical and optical properties can be modified by conjugation of a wide-bandgap metal oxide with some other semiconductor/s, transition metal/s, or carbon-based materials for powerful solar energy harvesting. In addition, the controlled synthesis procedure of most semiconductors can lead to desired properties with minimum cost and energy input. Nature-derived materials with photocatalytic properties are of great interest as the vast majority are safe, eco-friendly, cheap, and abundant. Additionally, the application of biomassderived nanomaterials as green alternatives can minimize the application of toxic substances and move closer to sustainability goals. The high porosity, large surface areas, sp2-hybridized structure, and robust stability of carbon-based composites offer great efficiencies for decontamination of wastewater under solar irradiation. On the other hand, water decontamination at minimal expense and in a sustainable manner can be accomplished via the effective utilization of solar energy. Therefore, the future directions should be to encourage the preparation of photocatalytic nanocomposites that fulfill requirements for easy preparation, high adsorption, wide solar absorption property, and excellent charge separation efficiencies along with facile separation. It must be emphasized that, despite large efforts devoted to developing photocatalysts capable of solar power conversion to chemical energy for water decontamination, there are few studies on some important factors such as photocatalyst content, photoreactor wall materials, temperature control, flow/ mixing patterns, and mass transfer efficiency and these should be also taken into account for practical and sustainable implementations.

Acknowledgments The authors are grateful for the support of the Ardabil University of Medical Sciences (IR.ARUMS. REC.1400.227) and the University of Mohaghegh Ardabili.

References 1 Shiklomanov, I.A. (2000). Appraisal and assessment of world water resources. Water Int. 25 (1): 11–32. 2 Cosgrove, W.J. and Loucks, D.P. (2015). Water Resour. Res. 51 (6): 4823–4839. 3 Rodriguez, D.J., Serrano, H.A., Delgado, A. et al. (2020). From Waste to Resource: Shifting Paradigms for Smarter Wastewater Interventions in Latin America and the Caribbean. Washington, DC: World Bank.

737

738

33  Heterogeneous Photocatalysis for Wastewater Treatment: A MajorStep Towards Environmental Sustainability

4 Connor, R. (2021). The United Nations World Water Development Report 2021, VALUING WATER World, 1–11. Italy: UNESCO World Water Assessment Programme. 5 Shannon, M.A., Bohn, W., Elimelech M. et al. (2008). Nature 452 (7185): 301–310. 6 Landrigan, P.J. et al. (2018). The Lancet 391 (10119): 462–512. 7 Lema, J.M. and Martinez, S.S. (2017). Innovative Wastewater Treatment & Resource Recovery Technologies: Impacts on Energy, Economy and Environment. London: IWA Publishing. 8 Scardozzi, G., Ismaelli, G.T., Leucci, G. et al. (2021). Archaeol. Prospect. 28 (2): 137–151. 9 Crini, G. and Lichtfouse, E. (2019). Environ. Chem. Lett. 17 (1): 145–155. 10 Rahim Pouran, S., Abdul Aziz, A.R., and Wan Daud, W.M.A. (2015). J. Ind. Eng. Chem. 21: 53–69. 11 Keller, N. Ivanez, J., Highfield,J et al. (2021). Appl. Catal. B Environ. 296: 120320. 12 Wu, Q. and Zhang, Z. (2019). Adv. Powder Technol. 30 (2): 415–422. 13 White, J.L., Baruch, M.F., Pander, J.E. et al. (2015). Chem. Rev. 115 (23): 12888–12935. 14 Schneider, J., Matsuoka, M., Takeuchi, M. et al. (2014). Chem. Rev. 114 (19): 9919–9986. 15 Pirhashemi, M., Habibi-Yangjeh, A., and Rahim Pouran, S. (2018). J. Ind. Eng. Chem. 62: 1–25. 16 Fauzi, A.A., Jalil, A.A., Hassan, N.S. et al. (2022). Chemosphere 286: 131651. 17 Ong, W.-J., Tan, L.-I., Ng, Y.H. et al. (2016). Chem. Rev. 116 (12): 7159–7329. 18 Singh, P. et al. (2019). Mater. Today Chem. 14: 100186. 19 Cao, M., Wang, P., Ao, Y. et al. (2016). J. Colloid Interface Sci. 467: 129–139. 20 Deng, J., Xiao, S., Wang, B. et al. (2020). ACS Appl. Mater. Interfaces 12 (46): 51537–51545. 21 Gao, Z., Yang, H., Li, J. et al. (2020). Appl. Catal. B Environ. 267: 118695. 22 Wang, H., Zhang, R., Yuan, D. et al. (2020). Adv. Funct. Mater. 30 (46): 2003995. 23 Zhang, C., Wu, M.-B., Wu, B.-H. et al. (2018). J. Mater. Chem. A 6 (19): 8880–8885. 24 Coronado, J.M. (2013). A historical introduction to photocatalysis. In: Design of Advanced Photocatalytic Materials for Energy and Environmental Applications (eds. J.M. Coronado, F. Fresno, M. Hernández-Alonso, and R. Portela), 1–5. London: Springer. 25 Li, K., de Rancourt de Mimérand, Y., Jin, X. et al. (2020). ACS Appl. Nano Mater. 3 (3): 2830–2845. 26 Mohd Adnan, M.A., Phoon, B.L., and Muhd Julkapli, N. (2020). J. Clean. Prod. 261: 121190. 27 Batzill, M. (2011). Energy Environ. Sci. 4 (9): 3275–3286. 28 Nunes, D., Pimentel, A., Santos, L. et al. (2019). 3 - Structural, optical, and electronic properties of metal oxide nanostructures. In: Metal Oxide Nanostructures (eds. D. Nunes, L. Santos, L. Pereira et al.), 59–102. Elsevier. 29 Pimentel, A., Rodrigues, J., Duarte, P. et al. (2015). J. Mater. Sci. 50 (17): 5777–5787. 30 Zhang, H., Zhang, Z., Liu, Y.et al. (2021). J. Phys. Chem. Lett. 12 (38): 9188–9196. 31 Raizada, P., Soni, V., Kumit, A. et al. (2021). J. Materiomics 7 (2): 388–418. 32 Hassandoost, R., Pouran, S.R., Khataee, A. et al. (2019). J. Hazard. Mater. 376: 200–211. 33 Hu, J., Zhao, R., Li, H. et al. (2022). Appl. Catal. B Environ. 303: 120869. 34 Kumar, S., Kaushik, R.D., Upadhyay, G.K. et al. (2021). J. Hazard. Mater. 406: 124300. 35 Ma, R., Zhang, S., Wen, T. et al. (2019). Catal. Today 335: 20–30. 36 Shi, Y., Ma, J., Chen, Y. et al. (2022). Sci. Total Environ. 804: 150024. 37 Chen, S., Huang, D., Xu, P. et al. (2020). ACS Catal. 10 (2): 1024–1059. 38 Kumar, R., Raizada, P., Verma, N. et al. (2021). J. Clean. Prod. 297: 126617. 39 Samuel, O., Othman, M.H.D., Kamaludin, R. et al. (2022). Ceram. Int. 48 (5): 5845–5875. 40 Sun, C., Yang, J., Xu, M. et al. (2022). Chem. Eng. J. 427: 131564. 41 Wan, X., Wang, Y., Jin, H. et al. (2019). Ceram. Int. 45 (17, Part A): 21091–21098. 42 Wei, X., Pan J., Wei, J. et al. (2018). Photonics Nanostruc. Fundam. Appl. 30: 20–24. 43 Fujishima, A. and Honda, K. (1972). Nature 238 (5358): 37–38. 44 Loeb, S.K., Alvarez, P.J.J., Brame, J.A. et al. (2019). Environ. Sci. Technol. 53 (6): 2937–2947.

References

45 Lee, S.-Y. and Park, S.-J. (2013). J. Ind. Eng. Chem. 19 (6): 1761–1769. 46 Al-Hajji, L.A. and Ismail, A.A. (2019). Superlattices Microstruct. 129: 259–267. 47 Mancuso, A., Sacco, O., Vaiano, V. et al. (2021). Catal. Today 380: 93–104. 48 Jiménez-Salcedo, M., Monge, M., and Tena, M.T. (2022). Photochem. Photobiol. Sci. 21: 337–347. 49 Vukoje, I., Kovač, K., Džunuzović, J. et al. (2016). J. Phys. Chem. C 120 (33): 18560–18569. 50 Ji, H., Du, P., Zhao, D. et al. (2020). Appl. Catal. B Environ. 263: 118357. 51 Kim, M.G., Kang, J.M., Lee, J.E. et al. (2021). ACS Omega 6 (16): 10668–10678. 52 Hsien, Y.-H., Chang, C.-F., Chen, Y.-H. et al. (2001). Appl. Catal. B Environ. 31 (4): 241–249. 53 He, J., Kumar, A., Khan, M. et al. (2021). Sci. Total Environ. 758: 143953. 54 Ding, D., Liu, K., He, S. et al. (2014). Nano Lett. 14 (11): 6731–6736. 55 Taylor, C.M., Ramirez-Canon, A., Wenk, J. et al. (2019). J. Hazard. Mater. 378: 120799. 56 Wolski, L., Grzelak, K., Muńko, M. et al. (2021). Appl. Surf. Sci. 563: 150338. 57 Das, S., Misra, A.J., Habeeb Rahman, A.P. et al. (2019). Appl. Catal. B Environ. 259: 118065. 58 Gawande, M.B., Branco, P.S., and Varma, R.S. (2013). Chem. Soc. Rev. 42 (8): 3371–3393. 59 Kim, B., Jang, J., and Lee, D.S. (2022). Chemosphere 289: 133040. 60 Esquinazi, P., Hergert, W., Spemann, D. et al. (2013). IEEE Trans. Magn. 49 (8): 4668–4674. 61 Hernández-Tecorralco, J., Meza-Montes, L., Cifuentes-Quintal, M.E. et al. (2020). J. Phys. Condens. Matter 32 (25): 255801. 62 Rahim Pouran, S., Abdul Raman, A.A., and Wan Daud, W.M.A. (2014). J. Clean. Prod. 64: 24–35. 63 Ji, W.-C., Hu, P., Wang, X-Y. et al. (2021). J. Alloys Compd. 866 (p): 158952. 64 Selvakumar, K., Wang, Y., Lu, Y. et al. (2022). Appl. Catal. B Environ. 300: 120740. 65 Brossault, D.F.F., McCoy, T.M., and Routh, A.F. (2021). J. Colloid Interface Sci. 584: 779–788. 66 Fazli, A., Khataee, A., Brigante, M.et al. (2021). Chem. Eng. J. 404: 126391. 67 Sabri, M., Habibi-Yangjeh, A., Chand, V. et al. (2021). J. Mater. Sci. Mater. Electron. 32 (4): 4272–4289. 68 Masudi, A., Harimisa, G.E., Ghafar, N.A. et al. (2020). Environ. Sci. Pollut. Res. Int. 27 (5): 4664–4682. 69 Pan, S., Yin, R.-T., Huang, W.T. et al. (2021). J. Nanosci. Nanotechnol. 21 (6): 3178–3182. 70 Bhoi, Y.P., Fang, F., Zhou, X. et al. (2020). Appl. Surf. Sci. 525: 146571. 71 Byrne, C., Dervin, S., Hermosilla, D. et al. (2021). Catal. Today 380: 199–208. 72 Mohan, H., Ramasamy, M., Ramalingam, V. et al. (2021). J. Hazard. Mater. 412: 125330. 73 Liu, X., Fang, B., Wang, Z. et al. (2021). ACS Appl. Nano Mater. 4 (4): 3742–3749. 74 Mukherjee, I., Cilamkoti, V., and Dutta, R.K. (2021). ACS Appl. Nano Mater. 4 (8): 7686–7697. 75 Dehghani, M., Nadeem, H., Singh Raghuwanshi, V. et al. (2020). ACS Appl. Nano Mater. 3 (10): 10284–10295. 76 Choudhary, M.K., Kataria, J., and Sharma, S. (2018). ACS Appl. Nano Mater. 1 (4): 1870–1878. 77 Chávez, A.M., Quiñones, D.H., Rey, A. et al. (2020). Chem. Eng. J. 398: 125642. 78 Sultana, S., Amirbahman, A., and Tripp, C.P. (2020). Appl. Catal. B Environ. 273: 118935. 79 Zhu, L., Kong, X., Yang, C. et al. (2020). J. Hazard. Mater. 381: 120910. 80 Ahmadpour, N., Sayadi, M.H., Sobhani, S. et al. (2020). J. Clean. Prod. 268: 122023. 81 Talukdar, K., Jun, B.-M., Yoon, Y. et al. (2020). J. Hazard. Mater. 398: 123025. 82 Chávez, A.M., Solís, R.R., and Beltrán, F.J. (2020). Appl. Catal. B Environ. 262: 118275. 83 Padervand, M., Ghasemi, S., Hajiahmadi, S. et al. (2021). Appl. Surf. Sci. 544: 148939. 84 He, S., Yan, C., Chen, X.-Z. et al. (2020). Appl. Catal. B Environ. 276: 119138. 85 Kumar, J.A., Krithiga, T., Manigandan, S. et al. (2021). J. Clean. Prod. 324: 129198. 86 Bayrami, A., Alioghli, S., Rahim Pouran, S. et al. (2019). Ultrason. Sonochem. 55: 57–66. 87 Ramanathan, S., Moorthy, S., Ramasundaram, S. et al. (2021). ACS Omega 6 (23): 14734–14747.

739

740

33  Heterogeneous Photocatalysis for Wastewater Treatment: A MajorStep Towards Environmental Sustainability

88 Liu, X. and Dai, L. (2016). Nat. Rev. Mater 1 (11): 16064. 89 Gopinath, K.P., Vo, D.-V.N., Gnana Prakash, D. et al. (2021). Environ. Chem. Lett. 19 (1): 557–582. 90 Das, G.S., Shim, J.P., Bhatnagar, A. et al. (2019). Sci. Rep. 9 (1): 15084. 91 Xie, X., Li, S., Qi, K. et al. (2021). Chem. Eng. J. 420: 129705. 92 Faisal, M., Alsaiari, M., Rashed, M.A. et al. (2021). J. Taiwan Inst. Chem. Eng. 120: 313–324. 93 Alsaiari, M. (2021). Arab. J. Chem. 14 (8): 103258. 94 Bayahia, H. (2022). J. Saudi Chem. Soc. 26 (2): 101432. 95 Yu, C., Tang, J., Liu, F. et al. (2021). Chemosphere 284: 131237. 96 Zhang, Y., Sun, J., Tan, J. et al. (2021). Fuel 305: 121622. 97 Omoriyekomwan, J.E., Tahmasebi, A., Dou, J. et al. (2021). Fuel Process. Technol. 214: 106686. 98 Hoang, A.T., Nižetić, S., Cheng, C.K. et al. (2022). Chemosphere 287: 131959. 99 Ahuja, V., Bhatt, A.K., Varjani, S. et al. (2022). Chemosphere 293: 133564. 100 Xu, X., Ray, R., Gu, Y. et al. (2004). J. Am. Chem. Soc. 126 (40): 12736–12737. 101 Hagiwara, K., Horikoshi, S., and Serpone, N. (2021). Chem. A Eur. J. 27 (37): 9466–9481. 102 Lim, S.Y., Shen, W., and Gao, Z. (2015). Chem. Soc. Rev. 44 (1): 362–381. 103 Pirsaheb, M., Asadi, A., Sillanpää, M. et al. (2018). J. Mol. Liq. 271: 857–871. 104 Demchenko, A.P. and Dekaliuk, M.O. (2013). Methods Appl. Fluoresc. 1 (4): 042001. 105 Choppadandi, M., Guduru, A.T., Gondaliya, P. et al. (2021). Mater. Sci. Eng. C 129: 112366. 106 Zhai, Y., Zhang, B., Shi, R. et al. (2022). Adv. Energy Mater. 12 (6): 2103426. 107 Jiang, L., Yang, J., Zhou, S. et al. (2021). Coord. Chem. Rev. 439: 213947. 108 Deng, Y., Cheng, M., Cheng, G. et al. (2021). ACS Omega 6 (6): 4247–4254. 109 Wu, L., Wu, T., Liu, Z. et al. (2022). J. Hazard. Mater. 431: 128536. 110 Chen, W., Pan, X., Willinger, M.-G. et al. (2006). J. Am. Chem. Soc. 128 (10): 3136–3137. 111 Alsawat, M., Altalhi, T., Gulati, K. et al. (2015). ACS Appl. Mater. Interfaces 7 (51): 28361–28368. 112 Eswaramoorthi, I. and Hwang, L.-P. (2007). Diam. Relat. Mater. 16 (8): 1571–1578. 113 Aspitarte, L., McCulley, D.R., Bertoni, A. et al. (2017). Sci. Rep. 7 (1): 8828. 114 Mousavi, M., Habibi-Yangjeh, A., and Pouran, S.R. (2018). J. Mater. Sci. Mater. Electron. 29 (3): 1719–1747. 115 Zhu, J., Xiao, P., Li, H., et al. (2014). ACS Appl. Mater. Interfaces 6 (19): 16449–16465. 116 Wang, L., Wang, K., He, T. et al. (2020). ACS Sustain. Chem. Eng. 8 (43): 16048–16085. 117 Huang, R., Wu, J., Zhang, M. et al. (2021). Mater. Des. 210: 110040. 118 Zhao, Z., Sun, Y., and Dong, F. (2015). Nanoscale 7 (1): 15–37. 119 Ye, C., Li, J.-X., Li, Z.-J. et al. (2015). ACS Catal. 5 (11): 6973–6979. 120 Wang, X., Maeda, K., Thomas, A. et al. (2009). Nat. Mater. 8 (1): 76–80. 121 Cui, L., Song, J., McGuire, A.F. et al. (2018). ACS Nano 12 (6): 5551–5558. 122 Zhang, M., Duan, Y., Jia, H. et al. (2017). Catal. Sci. Technol. 7 (2): 452–458. 123 Razavi-Esfali, M., Mahvelati-Shamsabadi, T., Fattahimoghaddam, H. et al. (2021). Chem. Eng. J. 419: 129503. 124 Guo, C., Chen, M., Wu, L. et al. (2019). ACS Appl. Nano Mater. 2 (5): 2817–2829. 125 Deng, Y., Tang, L., Feng, C. et al. (2018). J. Hazard. Mater. 344: 758–769. 126 Asadzadeh-Khaneghah, S. and Habibi-Yangjeh, A. (2020). J. Clean. Prod. 276: 124319. 127 Phin, H.-Y., Ong, Y.-T., and Sin, J.-C. (2020). J. Environ. Chem. Eng. 8 (3): 103222. 128 Chinh, V.D., Hung, L.X., Di Palma, L. et al. (2019). Chem. Eng. Technol. 42 (2): 308–315. 129 Shaban, M., Ashraf, A.M., and Abukhadra, M.R. (2018). Sci. Rep. 8 (1): 781. 130 Zhang, L., Zhang, J., Xia, Y. et al. (2020). Int. J. Mol. Sci. 21 (3). 131 Tian, J., Leng, Y., Zhao, Z. et al. (2015). Nano Energy 11: 419–427. 132 Hu, X., Wang, W., Xie, G. et al. (2019). Chemosphere 216: 733–741.

References

133 Zhao, P., Jin, B., Yan, J. et al. (2021). RSC Adv. 11 (56): 35147–35155. 134 Tahir, M.B., Sagir, M., and Shahzad, K. (2019). J. Hazard. Mater. 363: 205–213. 135 Chen, X., Sagir, M., and Shahzad, K. (2019). ChemistrySelect 4 (20): 6123–6129. 136 Ding, H., Han, D., Han, Y. et al. (2020). J. Hazard. Mater. 393: 122423. 137 Lin, M., Li, F., Cheng, W. et al. (2022). Chemosphere 288: 132620.

741

743

34 Sustainable Homogeneous Catalytic Oxidative Processes for the Desulfurization of Fuels Federica Sabuzi1, Giuseppe Pomarico2,3, Pierluca Galloni1, and Valeria Conte1 1

Department of Chemical Science and Technologies, University of Rome Tor Vergata, Via della Ricerca Scientifica snc, Rome, Italy Department of Molecular and Translational Medicine, University of Brescia, Viale Europa, 11, Brescia, Italy 3 CSGI, Research Center for Colloids and Nanoscience, Via della Lastruccia, 3, Sesto Fiorentino, Firenze, Italy 2

34.1  Introduction At present, in addition to CO2 and nitrogen oxides, sulfide oxides (SOx) emissions must be strictly controlled, because SOx are responsible for severe environmental and human health issues. Therefore, there is a strong demand for ultra-low-sulfur fuels to limit the amount of SOx generated during combustion. Oxidative fuel desulfurization (ODS) currently represents one of the most valuable approaches to lower S-content in fuel, as it can be carried out in sustainable mild conditions, with simple equipment and low operating costs [1, 2]. In ODS, sulfides are oxidized to the corresponding sulfoxides and/or sulfones by an oxidant, in the presence of a catalyst; oxidation products are easily removed from the fuel by liquid-liquid extraction or adsorption on solid materials. The major aromatic organosulfur compounds in fuels (usually resistant to other desulfurization methods such as hydrodesulfurization) are benzothiophene (BT), dibenzothiophene (DBT), 4-methyldibenzothiophene (MDBT) and 4,6-dimethyldibenzothiophene (DMDBT) (Figure 34.1). In this chapter, recent developments in oxidative fuel desulfurization coupled with liquid-liquid extraction methods are reported. In particular, it focuses on the application of homogeneous metal catalysis to obtain fuels with ultra-low content of sulfur. The key features of ODS, i.e. reaction selectivity for sulfides, catalysts and oxidant recovery and reuse, as well as the recyclability of extraction solvent, will be discussed, in order to assess the potential industrial application of the proposed systems.

34.2  Vanadium Vanadium exerts a chief role in oxidation reactions; next to the conventional commercially available V-catalysts, to date, a plethora of different V-complexes have been synthesized to catalyze the oxidation of alkanes, alkenes, arenes, alcohols, as well as sulfides [3–6]. Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies 2V Set, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

744

34  Sustainable Homogeneous Catalytic Oxidative Processes for the Desulfurization of Fuels

Figure 34.1  Principal aromatic organosulfur compounds in fuel.

The first examples of V-catalyzed homogeneous oxidative systems for fuel desulfurization involved the use of vanadyl acetylacetonate (VO(acac)2) and H2O2. Reactions were performed in biphasic system dissolving DBT in a model fuel, while the proper amount of VO(acac)2 and a large excess of H2O2 were dissolved in an immiscible solvent, like acetonitrile. Results showed that DBT was almost completely oxidized in two hours at 40 °C, and oxidation products were extracted in the acetonitrile layer [7]. Later on, eco-friendly alternatives to conventional organic solvents have been proposed as extractants. The ionic liquids (ILs) 1-butyl-3-methylimidazolium tetrafluoroborate ([bmim]BF4) and 1-butyl-3-methylimidazolium bis(trifluoromethanesulfonyl)imide ([bmim] Tf2N) successfully accomplished more than 95% of sulfur removal from model oils, in the presence of 5% mol of VO(acac)2 [8, 9]. Interestingly, catalyst and IL could be recycled without significant loss of activity [9]. Promising results have been obtained applying microwave irradiation (MW) during ODS. Indeed, S-removal could be efficiently achieved using VO(acac)2, H2O2, and N-carboxymethylpyridine hydrosulphate ionic liquid ([CH2CO2HPy]HSO4) with MW power of 500 W in 90 s, at 80 °C [10]. In addition, the system could be improved using a combination of H2O2/H2SO4 [10] or HNO3/H2SO4 with pyridinium phosphate ([HPy]H2PO4) as extractant IL [11]. However, the presence of very strong acids limits the industrial application of such protocol. The catalytic oxidation of DBT and DMDBT has been accomplished also using more elaborated structures (Figure 34.2). OxovanadiumIV complex V1 with a tetradentate N2O2-donor ligand has

Figure 34.2  Structure of selected V-complexes used in oxidative fuel desulfurization (ODS).

34.2  Vanadium

been adopted to perform ODS in mild conditions [5, 12]. Reactions performed on a model oil, prepared dissolving DBT and DMDBT in a mixture toluene–hexane in the presence of an excess of tert-butyl hydroperoxide (TBHP), at 40 °C, led to more than 60% of DMBT and c. 87% of DBT oxidation in six hours. The system was even upgraded, supporting V1 on a polystyrene-based polymer that improved catalyst stability and reaction selectivity to sulfone. Likewise, preliminary studies on DBT oxidation in acetonitrile have been performed using V2 and V3 catalysts [13, 14], with excess of TBHP [13] or H2O2 [14] as oxidants. High conversions and good selectivity to sulfone were achieved, but products extraction from the model oil was not reported. S-removal from a model fuel was also accomplished after DBT and DMDBT oxidation with TBHP, using oxovanadium complexes V4 and V5 [15]. Here, the beneficial role of 1-butyl-3-methylimidazolium hexafluorophosphate ([bmim]PF6) as extractive solvent was highlighted, and recyclability studies confirmed the robustness of such catalysts, which could be recycled with only slight decreasing of activity. V-oxodiperoxidocomplexes with pyridine, 2- and 4-picoline ligands have been found to be promising catalysts in desulfurization of a diesel fuel in a biphasic system (CH2Cl2/H2O), using cationic surfactants and H2O2 as the oxidant [16]. On the contrary, V-peroxido complexes with 2,6-pyridinedicarboxylic acid derivatives as ligands led to unsatisfactory results in diesel desulfurization [17]. Notably, salophen and salen resulted amongst the most promising ligands for V catalyzed ODS [18–20]: in a recent example, a three-phase system (fuel/H2O2/IL) was proposed (Figure 34.3), in which DBT was dissolved in a model fuel, while 0.5% of VV-catalyst (V6-V8) was dissolved in an immiscible IL and the primary oxidant (i.e. H2O2) resided between the two phases. Importantly, the use of H2O2 is preferred to TBHP for improving the atom economy of the process. Here, V6 resulted the most active catalyst because it led to almost quantitative DBT conversion under conventional heating or with MW irradiation at 35 W and 100 °C [18, 21]. The successful feature of such a protocol was the remarkable selectivity for sulfides oxidation, because cyclooctene added in the mixture as a competitive substrate was not significantly oxidized [18]. Indeed, it has proved to be a particularly good and economical alternative to hydrodesulfurization methods, being able to strongly reduce S-content in fuels while leaving alkenes amount intact, thus being suitable for industrial purposes.

Figure 34.3  Three phase system for oxidative fuel desulfurization (ODS) [18] / Elsevier.

745

746

34  Sustainable Homogeneous Catalytic Oxidative Processes for the Desulfurization of Fuels

34.3  Manganese Taking inspiration from oxygenase enzymes, which can convert sulfides in sulfoxides and sulfones under mild conditions, the catalytic activity of Mn-porphyrins has been evaluated in ODS. Desulfurization of a model oil prepared dissolving BT, 2-methyl-BT, 3-methyl-BT and DBT in hexane has been accomplished using [5,10,15,20-tetrakis(2,6-dichlorophenyl)porphyrinato]manganeseIII chloride as the catalyst and ammonium acetate as co-catalyst, both dissolved in CH3CN [22, 23]. Reactions performed in the dark, at 22–25 °C, using H2O2 as oxidant, led to almost complete substrates oxidation after two hours.

34.4  Iron Fe-TAML® catalyst (Figure 34.4; TAML: tetraamidomacrocyclic ligand) is a class of iron catalysts with ligand(s) designed to be used under different reaction conditions in term of pH, temperature, and solvents. Fe1 [24] was investigated both in homogeneous medium and in a two-phase system. Homogeneous reaction was performed in water/tert-butyl alcohol, where the latter was chosen to solubilize DBT. The reaction was affected by water/t-BuOH ratio, where 7:3 v/v gave the fastest reaction but no conversion was reported with lighter alcohols. Under these conditions DBT is the most reactive species followed by MDBT and DMDBT. In the two-phase reaction, performed with decane as model for diesel, oxidation requires longer time for completion (three hours vs 30 minutes) mainly because of the oxidant was added in six portions to limit its decomposition, thus extending the overall reaction time. The properties of μ-oxo dinuclear iron complex in sulfide oxidation were tested and compared to those of the structurally related mononuclear FeIII complex [25]. Dimeric iron-complex dissolved in CH3CN with H2O2 quickly reacts with DBT, as suggested by the blue color disappearance of the peroxido intermediate. DBT can directly react with the peroxido-intermediate or with some decomposition products, yielding DBTO, whereas sulfone derivative appears after the addition of eight equivalents of H2O2, ruling out direct formation of DBTO2. Investigation of the reaction mechanism performed with H218O and mass analysis suggested the involvement of a FeIV(O) species formed upon the homolytic cleavage of the peroxido intermediate. Zhou et al. [26] reported the oxidation of DBT in decalin with O2 at 1.5 MPa and 160 °C for five hours; oxidation occurred only in the presence of a porphyrin containing catalyst, Fe2 (Figure 34.5). Results by the product characterization were in agreement with the formation of DBTO2. The molecular structure of the catalyst affects its reactivity that increases with electron-withdrawing groups and decreases with donating substituents. However, the required harsh reaction conditions may lead to catalyst degradation and deactivation. As far as reaction mechanism is concerned, radical process was ruled out, while a two-step nucleophilic addition was proposed. The catalytic activity of iron μ-oxo dimer of Fe3 was investigated by Aguiar [27] in biphasic system, made by equal volume of the model oil and immiscible extracting solvent. The process was performed in two steps, where the first consists in the extraction of sulfur-based compounds, followed by H2O2 addition to start Figure 34.4  Fe1 structure [24] / Elsevier. oxidation reaction.

34.4  Iron

Figure 34.5  Structure of selected iron porphyrins used in oxidative fuel desulfurization (ODS); axial ligand is Cl− unless otherwise stated.

Although highest extraction value was achieved by DMF, no oxidation products were detected in this solvent, while CH3CN and CH3OH yielded satisfactory results after two hours. Little loss of activity was reported after three consecutive cycles. Pires et al. [28] explored the ODS reaction by iron-porphyrin bearing electron-withdrawing group on peripheral phenyl rings (Fe4). They focused on the sustainability of the reaction such as the choice of a green solvent (methanol or ethanol) and the ability of Fe-porphyrin to promote the reaction without co-catalyst (only H2O2 was used). Once the efficacy toward thioanisole had been proven, this catalyst was applied to more refractory substrates and generally good results were obtained. The need for the greenest possible systems prompted Zhao et al. [29] to explore ILs as solvent. FeIII porphyrin and H2O2 in different ILs as extractants allowed to remove at least more than 97% (complex Fe5–Fe8 yielded different values) of DBT added to octane as model oil, while lower amount of DBT was removed with no catalyst added. Under optimal reaction conditions, BT removal reached 94.7% with Fe7 and [bmim]PF6. High efficiency of this IL could be due to the stabilization effect of both ions toward intermediates and it could be recycled for six times with no loss of catalytic activity. Similar results were obtained on desulfurization of real diesel fuel. The same authors [30] detailed the role of anionic axial ligand of Fe5 in the catalytic activity enhancement. Native chloride was replaced by PF6− or BF4− which bind the metal center more loosely, favoring the formation of active intermediates, FeIII hydroperoxides porphyrin complex or highvalent FeIV oxo-porphyrin cation radical. As a result, better conversion rate of DBT in DBTO2 was obtained. Fe-phthalocyanine complex Fe9 (Figure 34.6) was initially investigated by Zhou et al. [31] with O2 as oxidant in decalin. In the presence of 1% of the catalyst, DBTO2 was produced. Conversion was enhanced by O2 pressure (solubility in decalin increases) and temperature increase. Reactivity and stability of phthalocyanines bearing different functional groups was investigated and the general trend is that the lower the electron density, the higher the reactivity. In biphasic water/octane system, Liu et al. [32] investigated oxidation of DBT and related compounds to sulfone with Fe10. Under mild conditions (30 °C, 40 equivalents of H2O2), DMDBT was more

747

748

34  Sustainable Homogeneous Catalytic Oxidative Processes for the Desulfurization of Fuels

Figure 34.6  Structure of iron phthalocyanines used in oxidative fuel desulfurization (ODS).

reactive than DBT and BT, despite steric hindrance. Noteworthy, nitrogen-based compounds such as pyridine or quinoline, common contaminants of fuels, facilitate the formation of the active intermediate high-valent ironIV oxo species. Dimeric O-bridged complex of Fe11 was chosen by Fang to investigate extractive-oxidative desulfurization in 9:1 mL of ethanol / water mixture and 5 mL of model oil with H2O2 as oxidant [33]. Catalyst improved ODS from about 37% to more than 98%. Further improvement was obtained in the presence of 4-mercaptopyridine, by its coordination to iron center. In this case, more hindered DMDBT resulted the species with lowest activity.

34.5  Cobalt Tripathi et al. [34] elaborated a water-soluble N-benzylated cobalt phthalocyaninetetrasulfonamide, to obtain an oil-soluble derivative, more suitable for oxidative desulfurization. The optimal catalyst/DBT molar ratio was found 1:10; lower amount of catalyst did not allow oxidation of substrate, while larger excess led to its aggregation. Some other parameters were explored, such as H2O2/DBT ratio and CH3CN (extracting solvent)/oil ratio. At the best, 85% of extraction efficiency was reached. The same author [35] investigated desulfurization by a CoII 5,10,15,20-tetrarylporphyrinate. Reactions were performed in dodecane as model oil, while CH3CN was used as extracting solvent; 91% desulfurization was achieved with DBT/catalyst 15:1 molar ratio and 20 equivalents (respect to DBT) of H2O2 at 50 °C. Porphyrin with different substituents on phenyl rings marginally affected the oxidation process. Properties of three CoII salen complexes were also investigated [36]. By using the same reaction conditions, ligand with 3,5-di-tert-butyl groups showed the highest activity. The bulky groups near the metal center enhance catalytic activity and increase solubility in the organic solvent. In general, yield of sulfur oxidation was in the 60–76% range depending on catalyst, H2O2 molar ratio, and solvent volume.

34.6  Molybdenum

34.6  Molybdenum Mo resulted a promising catalyst in the oxidation of alkyl and aryl sulphides, in homogeneous and heterogeneous processes [37], and thus its activity in fuel desulfurization has been explored (Figure 34.7). Peroxido-molybdenum amino acid complexes Mo1–Mo3 have been screened in the oxidation of DBT from a model fuel, with an excess of H2O2 [38]. More than 99% of S-removal was achieved with 10% mol of Mo1, at 70 °C, using [bmim]BF4 and [bmim]PF6 as extractant after oxidation. The same system could be also applied for DMDBT and BT removal from fuel: full elimination of DMDBT was accomplished in three hours, whereas BT removal reached 90%. The use of dioxomolybdenumVI complexes with 4,4′-di-tert-butyl-2,2′-dipyridyl ligand (Mo4) [39], N,N-dimethylbenzamide (Mo5) [40] or N,N′-diethyloxamide (Mo6) [40], cyclopentadienyl molybdenum tricarbonyl complex (Mo7) [41] or indenyl molybdenum tricarbonyl complex (Mo8) [41] as pre-catalysts led to remarkable results in term of S-removal from model and real diesels. In particular, preliminary desulfurization studies were performed using octane enriched with BT, DBT, MDBT, and DMDBT as the model oil. Reactions were performed in biphasic system, using acetonitrile or [bmim]BF4 as extractive solvents, in which an excess of H2O2 was dissolved, or in three-phase system (diesel/H2O2/IL) using [bmim]PF6 [39, 41, 43]. In all cases, excellent results were achieved, but [bmim]PF6 resulted the most promising solvent, because it led to more than 94% of S-removal in two hours at 50 °C with all of the catalysts. Catalysts and IL could be efficiently recycled and reused; furthermore, their application to a commercial diesel confirmed the efficiency of the protocol. A similar approach has been developed to simultaneously achieve S- and N-removal from fuel [42]. In particular, hybrid MoVI-bipyridine based catalysts allowed to remove 99.9% of S and 97% of N from a model diesel, using [bmim]PF6, in the presence of H2O2, at 70°C. Reusability and recyclability tests confirmed the possible use of catalysts and IL for consecutive desulfurization and denitrogenation run. Interestingly, Mo5 and Mo6 have been also exploited in ODS using two polyethylene glycol (PEG)-based deep eutectic solvents (DESs) as extraction solvents and reaction media [43]. DESs were obtained by combining PEG as hydrogen bond donor with tetrabutylammonium chloride or choline chloride as hydrogen bond acceptor. Reactions run in biphasic conditions, at 70 °C, led to

Figure 34.7  Structure of selected Mo-complexes.

749

750

34  Sustainable Homogeneous Catalytic Oxidative Processes for the Desulfurization of Fuels

S-free diesel after two hours. In addition, 82% desulfurization was achieved after the treatment on a real fuel, demonstrating the applicability of DESs as sustainable extractive solvents in ODS, even though their industrial scalability is still controversial.

34.7  Tungsten In a recent example, ODS catalyzed by Na2WO4.2H2O, in the presence of acetic acid and H2O2 as oxidants, has been explored [45]. However, to achieve good conversion results, extraction with methanol and an additional alumina adsorption step were required, leading to approx. 20% diesel loss. Nevertheless, Na2WO4 was previously adopted as catalyst in an interesting example of ultradeep oxidative desulfurization: a model diesel prepared dissolving DBT in tetradecane, was added to a mixture of Na2WO4.2H2O and an excess of H2O2 dissolved in 1-(4-sulfonic acid)butyl-3-methylimidazolium p-toluenesulfonate ([(CH2)4SO3HMIm]Tos) [46]. The reaction performed at 50 °C for three hours led to > 99% of S-removal, with the formed sulfone extracted in the IL phase. Interestingly, the active catalytic species was a peroxytungstate-ionic liquid complex, that made the system homogenous, and it could be even recycled by adding fresh diesel and H2O2. Remarkably, at the working temperature, aromatic compounds, essential components in fuel, were not oxidized. Encouraging results have been achieved also using hexatungstates (i.e. [(C4H9)3NCH3]2W6O19, [(C8H17)3NCH3]2W6O19, or [(C12H25)3NCH3]2W6O19) in water-in-IL emulsion systems with H2O2 [47]. Specifically, the quaternary ammonium ion of the catalyst and IL cation (1-octyl-3-methylimidazolium) promote the formation of emulsion droplets and extract S-compounds from the model oil to the IL. Here, IL anion (i.e. PF6-) promotes the formation of the active W-peroxidocomplex. Indeed, with such system, almost complete DBT and DMDBT removal was achieved at 60 °C in 60 and 80 minutes, respectively, while BT removal was c. 75%; in addition, quite promising results have been achieved on real gasoline, where 94% S-removal was obtained in 4 desulfurization steps. Interestingly, the catalytic activity of phosphotungstic acid with H2O2 for ODS has been evaluated using DESs as extractants. Reactions performed with ChCl/2PEG (obtained by combining choline chloride (ChCl) with 2 eqs. of PEG), led to > 99% of DBT removal at 50 °C within three hours [48].

34.8  Polyoxometalates Polyoxometalates (POM) are a family of anionic metal-oxide clusters, generally with octahedral structures, exploited in several fields ranging from catalysis to biology to energy and material science [49]. Due to the large selectivity, efficiency, and versatility, POMs have been used in ODS as homogeneous catalysts with H2O2 or O2 as oxidant and organic solvents or ILs as extractants. Two recent reviews detailed desulfurization by using POMs [9, 50, 51]. Even more recently, the so called Venturello compounds (Bu4N)3[PO4{MO(O2)2}4], where M is Mo or W, were investigated under different reaction conditions (catalyst and oxidant amount, temperature) and with different extracting solvents [52, 53]. In the case of Mo-derivative, highest extracting ability was achieved for the solvent-free method (respect to CH3CN or [bmim]PF6), that actually is a heterogeneous system [53]. Lately, both W and Mo based POMs were compared using CH3CN, [bmim]PF6 or DES (tetrabutylammonium/PEG) as extracting systems. After H2O2 addition, complete desulfurization within 40 minutes was achieved with peroxidotungstate and [bmim]PF6. The catalyst has stronger acid sites while [bmim]PF6 allowed the formation of a triphasic system that in some way protects catalyst from deactivation/decomposition.

34.9  Ionic Liquids

Mn polyoxotungstate was investigated by Duarte et al. [54]. With H2O2 as oxidant and CH3CN as extracting solvent, such POM oxidizes sulfur-based substrates into the corresponding sulfones, operating at room temperature with almost 100% conversion with a relatively low S/C molar ratio (150:1). System sustainability was then improved by using ethanol/water mixture to remove sulfone from hexane (chosen as model oil); after 20 minutes, 98 mol% of the organosulfur compounds were removed.

34.9  Ionic Liquids Ionic liquids, which initially were exploited as non-flammable and less toxic alternative to many organic solvents to extract S-based impurities from fuels, are currently being explored as promising eco-friendly catalysts in ODS. Among the others, the catalytic activity of metal-based ILs is receiving great attention. IL catalytic activity depends by the Lewis-acid character of the organic cation. As general strategy, metals halides are added to ILs leading to the formation of a [organic cation](metal halide anion) moiety, while H2O2 is used as oxidant. Accordingly, Lewis acid ionic liquids, with alkylated 1,8-diazabicyclo[5.4.0]undec-7-ene cation and ZnCl2-based complex anion have been explored as catalysts in ODS with H2O2 (Figure 34.8) [55]. ZnCl2 coordination to sulfide lone pair promotes ODS, that can be efficiently accomplished, at 50 °C in two hours. IL can be reused for six cycles with only a slight decrease of activity and more than 99% of S-removal can be achieved on a hydrogenated diesel via a one step process. Similarly, a series of metal-based IL having triethylamine hydrochloride cation and a series of anhydrous metal chlorides anions has been explored [56]. Among the others, [Et3NH]FeCl4 exhibited the best performance, leading to > 99% of S-removal in mild conditions, using O2 as the oxidant, under UV-light irradiation. To note, such protocol resulted particularly suitable for gasoline with low olefin concentration, since alkenes in solution lowers S-removal. Jiang et al. [57] reported that the oxidative desulfurization ability for ILs containing different metal ions strongly depends by the nature of the metals. They compared [(C8H17)3CH3N]Cl/MClx where MClx was CuCl2, SnCl2, ZnCl2, or FeCl3; in the absence of H2O2, sulfur removal was in the range of 20–30%, whereas the addition of the oxidant allowed an increase in this value close to 98% with 0.5 or 1 equivalent of FeCl3. This system was recycled six times with no significant loss of activity charging fresh H2O2 and model oil before starting a new cycle. 1-octyl-3-methylimidazolium tetrachloroferrate ([omim] FeCl4) was found to be a valuable IL for the extraction and catalytic oxidative desulfurization of a model fuel, reaching complete removal of BT, DBT, and DMDBT, with H2O2, at 25 °C in 15 minutes [58]. The IL exhibits a beneficial role in promoting sulfides extraction from the oil through the long hydrophobic alkyl chain, π-π interactions between imidazolium and Figure 34.8  Structure of a thiophenic ring, and interactions between S lone pairs and Zn-based ionic liquid (IL) [55] / Elsevier. Fe3+. In addition, Fe3+ catalyzes the formation of reactive oxygen species from H2O2, that are involved in sulfides oxidation. Even though [omim]FeCl4 retained its activity when recycled seven times, desulfurization of a real fuel was less effective. Likewise, a magnetic IL [C4(mim)2Cl2/2FeCl3], Figure 34.9) showed high desulfurization performances, with the Figure 34.9  Structure of advantage that it could be efficiently separated from model oil [C4(mim)2]Cl2/FeCl3 [59].

751

752

34  Sustainable Homogeneous Catalytic Oxidative Processes for the Desulfurization of Fuels

Figure 34.10  Structure of [Cn3MPy]FeCl4 [60] / Elsevier.

after reaction by applying an external magnetic field; thus, recycling could be accomplished with no significant loss of activity [59]. Comparisons of the effects of alkyl chain length in ILs containing of alkylpyridinium cation, like [C43MPy]FeCl4, [C63MPy]FeCl4, [C83MPy]FeCl4 (Figure 34.10), were made by Nie et al. [58]; they reported that the longer the alkyl chain, the higher the sulfur removal, a result owing to better extractive performances. Addition of H2O2 increased the sulfur removal due to the larger efficacy of extraction and oxidation process; [C83MPy]FeCl4 removes more than 99% of sulfur in all the H2O2/DBT ratio tested. After three cycles, efficacy decreases because of the DBTO2 accumulation in ILs. [C6mim] Cl/FeCl3 and [C63MPy]nFeCl4 with n = 0.5, 1, 2, 3 were investigated by Dong et al. [61]. With one equivalent of metal halide and H2O2/DBT = 4, authors reported a 100% yield. However, to achieve almost complete S-removal from a real gasoline, seven ODS runs were required, with the addition of fresh IL and H2O2 after each cycle, raising costs and limiting the industrial applicability of such protocol. Jiang et al. [62] explored a modified pyridinium based ionic liquids, [C4Py]3Fe(CN)6 and [C16Py]3Fe(CN)6 for the oxidation, coupled with 1-octyl-3-methylimidazolium hexafluorophosphate [omim]PF6 as extractant; these systems removed around 97 and 87% of sulfur respectively, more than sulfur removed by K3Fe(CN)6 when used alone. Experiments performed on real samples, demonstrated that the presence of aromatics or olefins has a negative impact on DBT oxidation. Redox couple of hexacyanoferrates were applied in the oxidation of DBT [63] and related derivatives in the form of [Cnmim]3Fe(CN)6 where n = 2, 4, and 8 as a catalyst and [bmim]BF4 as an extractant. The highest activity was measured for [bmim]3Fe(CN)6 in the presence of only 7.5% of H2O2 (almost 98% of sulfur removal); the radical anion superoxide was identified as the reactive species in the oxidation of sulfide to sulfones. Li et al. [64] used N-methyl-2-pyrrolidone (NMP) coordinating FeCl3 or ZnCl2; although bare NMP exhibits the highest extractive ability, in agreement with its structural similarity and intermiscibility with DBT, they investigated some derivatives such as C5H9NO·xFeCl3 (x = 0.1 or 0.3). As a result, C5H9NO·xFeCl3 showed very good catalytic activity (best conditions: 12 equivalents of H2O2, 30 °C, sulfur removal from 97 to 94.3% after six cycles) because it decreases the solubility of fuel oil in NMP, and allows DBT to be extracted in the more polar ILs phase. In the framework of metal-based IL, polyoxometalate-based ionic liquid catalysts are largely explored in ODS, showing excellent results in S-removal under mild conditions, with H2O2, in the presence of an extractant solvent, such as acetonitrile [65, 66], a different IL like [omim]BF4 or [omim]PF6 [67, 68], or a deep-eutectic solvent, as the one prepared from benzenesulfonic acid and PEG [69]. The peculiar catalytic activity of ILs having POM anions (i.e. [PMo12O40]3−, [PVMo11O40]4−, [PV2Mo10O40]4−, [PW12O40]3−, [PMo6W6O4 0]3−,) is usually ascribed to the formation of metal-­ peroxidocomplexes, which is pivotal for sulfides oxidation.

References

Recently, an ultrafast approach for fuel desulfurization has been proposed, using a NaClO oxidant and an ILs-based phase-transfer catalyst (i.e. 1-hexadecyl-3-methylimidazolium phosphomolybdate [C16mim]PMoO) [70]. DBT removal from octane was accomplished in 20 minutes and the authors highlighted the multiple roles of ILs: the C16 alkyl chain ensures the proper catalyst dispersibility in the oil phase, the imidazolium participates in DBT extraction from the oil through π-π interactions, and PMoO anion promotes O2•- and HO• generation from NaClO decomposition and H2O; such radicals react with the catalyst, leading to Mo-peroxido species, directly involved in DBT oxidation to sulfone. The proposed system was particularly promising, being efficiently applied also on real fuel: 99% desulfurization was obtained in 30 minutes, at 30 °C. Recyclability tests demonstrated the stability and reusability of such IL, which shows that it is suitable for potential industrial application; however, the use of greener oxidant, like H2O2 or O2, led to unsatisfactory results.

34.10  Conclusions In this chapter, a brief overview of the most efficient homogeneous catalytic systems designed to perform oxidative fuel desulfurization is offered. Several metal complexes have been successfully explored to lower the S-content in fuels in mild conditions, and the most valid protocols to be scaled-up at the industrial level have been highlighted. Of note, the use of ILs as sustainable extractive (and reusable) solvents and, possibly, as eco-friendly catalysts, is widespread and offers new perspectives to make fuels greener.

References 1 Boshagh, F., Rahmani, M., Rostami, K. et al. (2022). Energy Fuels 36 (1): 98–132. https://doi.org/ 10.1021/acs.energyfuels.1c03396. 2 Rajendran, A., Cui, T.-Y., Fan, H.-X. et al. (2020). J. Mater. Chem. A 8 (5): 2246–2285. https://doi. org/10.1039/C9TA12555H. 3 Conte, V. and Floris, B. (2010). Inorganica Chim. Acta 363 (9): 1935–1946. https://doi.org/10.1016/j. ica.2009.06.056. 4 Langeslay, R.R., Kaphan, D.M., Marshall, C.L. et al. (2019). Chem. Rev. 119 (4): 2128–2191. https:// doi.org/10.1021/acs.chemrev.8b00245. 5 Sutradhar, M., Martins, L.M.D.R.S., Guedes da Silva, M.F.C. et al. (2015). Coord. Chem. Rev. 301-302: 200–239. http://dx.doi.org/10.1016/j.ccr.2015.01.020. 6 Sutradhar, M., Pombeiro, A.J.L., and da Silva, J.A.L. (ed.) (2021). Vanadium Catalysis. Croydon: Royal Society of Chemistry. https://doi.org/10.1039/9781839160882. 7 Silva, G., Voth, S., Szymanski, P. et al. (2011). Fuel Process. Technol. 92 (8): 1656–1661. https://doi. org/10.1016/j.fuproc.2011.04.014. 8 Mota, A., Butenko, N., Hallett, J.P. et al. (2012). Catal. Today 196 (1): 119–125. https://doi. org/10.1016/j.cattod.2012.03.037. 9 Zhu, W., Xu, D., Li, H. et al. (2013). Pet. Sci. Technol. 31 (14): 1447–1453. https://doi.org/10.1080/10 916466.2010.545790. 10 Mesdour, S., Lekbir, C., Doumandji, L. et al. (2017). J. Sulphur Chem. 38 (4): 421–439. http://dx.doi. org/10.1080/17415993.2017.1304550.

753

754

34  Sustainable Homogeneous Catalytic Oxidative Processes for the Desulfurization of Fuels

11 Benmabrouka, H., Mesdour, S., Boufades, D. et al. (2019). Pet. Sci. Technol. 37 (6): 662–670. https:// doi.org/10.1080/10916466.2018.1563611. 12 Ogunlaja, A.S., Chidawanyika, W., Antunes, E. et al. (2012). Dalton Trans. 41 (42): 13908–13928. https://doi.org/10.1039/C2DT31433A. 13 Dembaremba, T.O., Correia, I., Hosten, E.C. et al. (2019). Dalton Trans. 48 (44): 16687–16704. https://doi.org/10.1039/C9DT02505G. 14 Saeedi, R., Safaei, E., Lee, Y.-I. et al. (2019). Appl. Organomet. Chem. 33 (3): e4781. https://doi. org/10.1002/aoc.4781. 15 Campitelli, P., Aschi, M., Di Nicola, C. et al. (2020). Appl. Catal. A: Gen. 599: 117622. https://doi. org/10.1016/j.apcata.2020.117622. 16 Gobara, H.M., Nessim, M.I., Zaky, M.T. et al. (2014). Catal. Lett. 144 (6): 1043–1052. https://doi. org/10.1007/s10562-014-1251-3. 17 Anisimov, A.V., Myltykbaeva, Z.B., Kairbekov, Z. et al. (2017). Theor. Found. Chem. Eng. 51 (4): 563–566. https://doi.org/10.1134/S0040579517040029. 18 Coletti, A., Sabuzi, F., Floris, B. et al. (2018). J. Fuel Chem. Technol. 46 (9): 1121–1129. https://doi. org/10.1016/S1872-5813(18)30045-8. 19 Floris, B., Sabuzi, F., Coletti, A. et al. (2017). Catal. Today 285: 49–56. https://doi.org/10.1016/j. cattod.2016.11.006. 20 Sabuzi, F., Pomarico, G., Conte, V. et al. (2021). Peroxo-vanadium complexes as sustainable catalysts in oxidations, halogenations and other organic transformations. In: Vanadium Catalysis (ed. M. Sutradhar, A.J.L. Pombeiro, and J.A.L. da Silva), 97–110. Croydon: Royal Society of Chemistry. https://doi.org/10.1039/9781839160882-00097. 21 Floris, B., Sabuzi, F., Galloni, P. et al. (2017). Catalysts 7 (9): 261. https://doi.org/10.3390/ catal7090261. 22 Calvete, M.J.F., Piñeiro, M., Dias, L.D. et al. (2018). ChemCatChem. 10 (17): 3615–3635. https://doi. org/10.1002/cctc.201800587. 23 Pires, S.M.G., Simões, M.M.Q., Santos, I.C.M.S. et al. (2012). Appl. Catal. A: Gen. 439-440: 51–56. https://doi.org/10.1016/j.apcata.2012.06.044. 24 Mondal, S., Hangun-Balkir, Y., Alexandrova, L. et al. (2006). Catal. Today 116 (4): 554–561. https:// doi.org/10.1016/j.cattod.2006.06.025. 25 Trehoux, A., Roux, Y., Guillot, R. et al. (2015). J. Mol. Catal. A: Chem. 396: 40–46. https://doi. org/10.1016/j.molcata.2014.09.030. 26 Zhou, X., Lv, S., Wang, H. et al. (2011). Appl. Catal. A: Gen. 396 (1–2): 101–106. https://doi. org/10.1016/j.apcata.2011.01.041. 27 Aguiar, A., Ribeiro, S., Silva, A.M.N. et al. (2014). Appl. Catal. A: Gen. 478: 267–274. https://doi. org/10.1016/j.apcata.2014.04.002. 28 Pires, S.M.G., Simões, M.M.Q., Santos, I.C.M.S. et al. (2014). Appl. Catal. B: Environ. 160-161: 80–88. https://doi.org/10.1016/j.apcatb.2014.05.003. 29 Zhao, R., Wang, J., Zhang, D. et al. (2017). Appl. Catal. A: Gen. 532: 26–31. https://doi. org/10.1016/j.apcata.2016.12.008. 30 Zhao, R., Wang, J., Zhang, D. et al. (2017). ACS Sustain. Chem. Eng. 5 (3): 2050–2055. https://doi. org/10.1021/acssuschemeng.6b02916. 31 Zhou, X., Li, J., Wang, X. et al. (2009). Fuel Process. Technol. 90 (2): 317–323. https://doi. org/10.1016/j.fuproc.2008.09.002. 32 Liu, H., Bao, S., Cai, Z. et al. (2017). Chem. Eng. J. 317: 1092–1098. https://doi.org/10.1016/j. cej.2017.01.086.

References

33 Fang, Z., Li, N., Zhao, Z. et al. (2022). Bio-inspired strategy to enhance catalytic oxidative desulfurization by O-bridged diiron perfluorophthalocyanine axially coordinated with 4-mercaptopyridine. Chem. Eng. J. 433 (2): 133569. https://doi.org/10.1016/j.cej.2021.133569. 34 Tripathi, D., Negi, H., Singh, R.K. et al. (2019). J. Coord. Chem. 72 (17): 2982–2996. https://doi.org/ 10.1080/00958972.2019.1683549. 35 Tripathi, D., Inderpal, Y., Negi, H. et al. (2021). J. Porphyr. Phthalocyanines 25 (01): 24–30. https:// doi.org/10.1142/S1088424620500443. 36 Tripathi, D. and Singh, R.K. (2021). Catal. Lett. 151: 713–719. https://doi.org/10.1007/ s10562-020-03343-4. 37 Thiruvengetam, P. and Chand, D.K. (2018). J. Indian Chem. Soc. 95: 781–788. https://doi. org/10.5281/zenodo.5638611. 38 Zhu, W.S., Li, H., Gu, Q.Q. et al. (2011). J. Mol. Catal. A: Chem. 336 (1–2): 16–22. https://doi. org/10.1016/j.molcata.2010.12.003. 39 Julião, D., Gomes, A.C., Pillinger, M. et al. (2016). Dalton Trans. 45 (38): 15242–15248. https://doi. org/10.1039/C6DT02065H. 40 Julião, D., Gomes, A.C., Cunha-Silva, L. et al. (2019). Catal. Commun. 128: 105704. https://doi. org/10.1016/j.catcom.2019.05.011. 41 Julião, D., Gomes, A.C., Pillinger, M. et al. (2018). Appl. Catal. B: Environ. 230: 177–183. https:// doi.org/10.1016/j.apcatb.2018.02.036. 42 Julião, D., Gomes, A.C., Pillinger, M. et al. (2020). Chem. Eng. Technol. 43 (9): 1774–1783. https:// doi.org/10.1002/ceat.201900624. 43 Julião, D., Gomes, A.C., Pillinger, M. et al. (2020). Appl. Org. Chem. 34 (4): e5490. https://doi. org/10.1002/aoc.5490. 44 Julião, D., Gomes, A.C., Pillinger, M. et al. (2020). J. Mol. Liq. 309: 113093. https://doi. org/10.1016/j.molliq.2020.113093. 45 Bourane, A., Koseoglu, O., Al‑Hajji, A. et al. (2019). React. Kinet. Mech. Catal. 126: 365–382. https://doi.org/10.1007/s11144-018-1484-z. 46 Liu, D., Gui, J., Ding, J. et al. (2011). React. Kinet. Mech. Catal. 104 (1): 111–123. https://doi. org/10.1007/s11144-011-0347-7. 47 Ding, Y., Zhu, W., Li, H. et al. (2011). Green Chem. 13 (5): 1210–1216. https://doi.org/10.1039/ C0GC00787K. 48 Liu, W., Jiang, W., Zhu, W. et al. (2016). J. Mol. Catal. A: Chem. 424: 261–268. https://doi. org/10.1016/j.molcata.2016.08.030. 49 Long, D.L., Tsunashima, R., and Cronin, L. (2010). Angew. Chem., Int. Ed. 49 (10): 1736–1758. https://doi.org/10.1002/anie.200902483. 50 Taghizadeh, M., Mehrvarz, E., and Taghipour, A. (2019). Polyoxometalate as an effective catalyst for the oxidative desulfurization of liquid fuels: a critical review. Rev. Chem. Eng. 36 (7): 831–858. https://doi.org/10.1515/revce-2018-0058. 51 Ahmadian, M. and Anbia, M. (2021). Energy Fuels 35: 10347–10373. https://doi.org/10.1021/acs. energyfuels.1c00862. 52 Gao, Y., Julião, D., Silva, F.L. et al. (2021). Mol. Catal. 505: 111515. https://doi.org/10.1016/j. mcat.2021.111515. 53 Julião, D., Gomes, A.C., Cunha-Silva, L. et al. (2020). Appl. Catal. A, Gen. 589: 117154. https://doi. org/10.1016/j.apcata.2019.117154. 54 Duarte, T.A.G., Pires, S.M.G., Santos, I.C.M.S. et al. (2016). Catal. Sci. Technol. 6: 3271. https://doi. org/10.1039/c5cy01564b.

755

756

34  Sustainable Homogeneous Catalytic Oxidative Processes for the Desulfurization of Fuels

55 Wang, J., Zhang, L., Sun, Y. et al. (2018). Fuel Process. Technol. 177: 81–88. https://doi. org/10.1016/j.fuproc.2018.04.013. 56 Wang, C., Chen, Z., Zhu, W. et al. (2017). Energy Fuels 31 (2): 1376–1382. https://doi.org/10.1021/ acs.energyfuels.6b02624. 57 Jiang, Y., Zhu, W., Li, H. et al. (2011). ChemSusChem. 4: 399–403. https://doi.org/10.1002/ cssc.201000251. 58 Andevary, H.H., Akbari, A., and Omidkhah, M. (2019). Fuel Process. Technol. 185: 8–17. https:// doi.org/10.1016/j.fuproc.2018.11.014. 59 Wang, T., Yu, W.-H., Li, T.-X. et al. (2019). N. J. Chem. 43 (48): 19232–19241. https://doi. org/10.1039/C9NJ04015C. 60 Nie, Y., Dong, Y., Lu, B. et al. (2013). Fuel 103: 997–1002. https://doi.org/10.1016/j.fuel.2012.07.071. 61 Dong, Y., Nie, Y., and Zhou, Q. (2013). Chem. Eng. Technol. 36 (3): 435–442. https://doi. org/10.1002/ceat.201200570. 62 Jiang, W., Li, H., Yin, S. et al. (2016). Appl. Organomet. Chem. 30: 753–758. https://doi.org/10.1002/ aoc.3500. 63 Jiang, W., Zhu, W., Chang, Y. et al. (2014). Energy Fuels 28: 2754–2760. https://doi.org/10.1021/ ef500082y. 64 Li, F.-T., Wu, B., Liu, R.-H. et al. (2015). Chem. Eng. J. 274: 192–199. https://doi.org/10.1016/j. cej.2015.04.027. 65 Akopyan, A., Eseva, E., Polikarpova, P. et al. (2020). Molecules 25 (3): 536. https://doi.org/10.3390/ molecules25030536. 66 Li, J., Guo, Y., Tan, J., and Hu, B. (2021). Catalysts 11 (3): 356. https://doi.org/10.3390/ catal11030356. 67 Hao, L., Sun, L., Su, T. et al. (2019). Chem. Eng. J. 358: 419–426. https://doi.org/10.1016/j. cej.2018.10.006. 68 Wang, J., Yang, B., Peng, X. et al. (2022). Chem. Eng. J. 429: 132446. https://doi.org/10.1016/j. cej.2021.132446. 69 Chi, M., Su, T., Sun, L. et al. (2020). Appl. Catal. B: Environ. 275: 119134. https://doi.org/10.1016/j. apcatb.2020.119134. 70 Li, A., Song, H., Meng, H. et al. (2020). Chem. Eng. J. 380: 122453. https://doi.org/10.1016/j. cej.2019.122453.

757

35 Heterogeneous Catalytic Desulfurization of Liquid Fuels The Present and the Future Rui G. Faria, Alexandre Viana, Carlos M. Granadeiro, Luís Cunha-Silva, and Salete S. Balula LAQV / REQUIMTE & Department of Chemistry and Biochemistry, Faculty of Sciences, University of Porto, Porto

35.1  Introduction The intensification of worldwide industrial activity has led to an ever-increasing dependence on fossil fuel combustion for energy production. The transition toward cleaner energy production technologies is crucial for ensuring environmental sustainability without significant economic impacts. Fossil fuel combustion still accounts for about 85% of the world’s energy production and the prediction for 2040 is a decrease non greater than 20% [1]. Powering civilization with fossil fuels has led to major strides in human development, but we are now aware that this entails severe consequences. Links have been developed between emissions resulting from fossil fuel combustion and severe health and environmental impacts [2]. Discussion around the environmental impacts of these fuels usually circles around the emission of carbon dioxide; however, other pollutants containing heteroatoms such as nitrogen and sulfur deserve just as much attention. The content of sulfur containing compounds (SCCs) is several times higher than that of nitrogen compounds in fossil fuels [1]. Consequently, the sulfur produced during fuel combustion (sulfur oxide[s]) contributes significantly to growing climate instability through the production of deleterious acids, and, when in contact with atmospheric moisture, by producing acid rain. Furthermore, the SSCs in fuels also lead to the corrosion of storage tanks, pipelines, and equipment, as well as the deactivation of catalysts used in refinery processes [3]. Sulfur can also chemisorb on the catalytic coverters of passenger vehicles, promoting deactivation and reducing engine efficiency [3]. These effects highlight the crucial economic and environmental need for the removal of SSCs from fuels. Strict legislation has been introduced worldwide to drastically reduce the sulfur content of transportation fuels. Since 2009, road fuels in the European Union have been limited to 10 ppm S [4]. Since 2020, sulfur content in marine fuels has been limited to 0,5% of sulfur (0,1% in the Baltic Sea, North Sea, and English Channel) by the International Marine Organization (IMO) [5]. These legislative efforts further pressure the oil processing industry into producing sulfur-free fuels. Four major groups for the SSCs in fuels can be classified, namely disulfides, mercaptans, sulfides, and thiophenes (Ts). The concentration and the nature of SSCs in fuels change over the Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 3, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels 160

120

80

40

0

1971 1973 1975 1978 1982 1985 1987 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022

Number scientific publications

758

Figure 35.1  Number of scientific publication from 1971 to date found in the Web of Science database, using the keywords “desulfurization” and “fuel”.

boiling range during petroleum distillation processes, with the heaviest fractions containing the largest amounts of sulfur species [6]. HDS is the most used method by the petroleum industry to drive down sulfur content in road transportation fuels (diesel and gasoline). This technology requires increased temperatures (300–400 C°), high pressures (3–6 MPa), and a large quantity of hydrogen. The strict regulation imposed for ultra-low sulfur fuel commercialization turn HDS into a costly choice. Large efforts have been done by researchers to develop alternative technologies with higher cost-efficiency and eco-sustainability. A total of 6,551 related scientific papers were published from 1971 to date (Figure 35.1). This high number of the scientific reports presents several possible simple or combined desulfurization technologies able to replace or complement HDS. A large part of these technologies need to use materials as the main key to achieve high desulfurization (5,212 papers found in the Web of Science database in a search incorporating the following three terms: desulfurization, fuel, and materials). These materials integrate mainly novel catalysts (2,474 papers) and/ or highly effective and selective absorbent materials. This chapter presents the most important advances given by various materials in the area of desulfurization of liquid fuels, mainly in the area of HDS, adsorptive desulfurization (ADS), oxidative desulfurization (ODS), and the combination of extractive and oxidative desulfurization (ECODS) achieved by porous powdered and membrane materials. The morphology, porosity, hydrophilicity, and nature of functional groups in the material surfaces are crucial properties in the fuel media and have a remarkable influence in the desulfurization process.

35.2  Hydrodesulfurization Hydroconversion is a broad term used to refer to the reaction of hydrocarbons with hydrogen during oil processing and includes hydrotreating, hydrocracking, and hydrogenation [7]. HDS is thus the term used to describe the catalytic desulfurization of any hydrocarbon stream during hydrotreating. This technique has been reviewed exhaustively over the past decades [7, 8], and a brief overview is given here for contextualization.

35.2  Hydrodesulfurization

HDS of lower boiling aromatics is carried out in vapor phase, but most other applications of HDS are in liquid phase, in which it is generally more efficient [9]. Oil stocks are desulfurized for many reasons, the main one being to prevent atmospheric pollution that results from the combustion of residual stocks. In mid-distillate treating, desulfurization was originally practiced just to avoid corrosion of heating equipment. In naphtha pre-treating, desulfurization is important to avoid poisoning of precious metal used in its catalytic reforming. Desulfurization is not only practiced on hydrocarbon streams that concern fuel production, as it is also key to assure purity standards in other chemical production processes. During HDS, sulfur-containing compounds from the feedstock are converted into sulfur-free hydrocarbons at high temperature and pressure conditions and in the presence of hydrogen, producing hydrogen sulfide gas. Although HDS is a well-established industrial process and is being employed for several decades, specific mechanisms of HDS reactions are still a matter of debate. In so far as these reactions are associated with the cleavage of carbon-sulfur bonds, they are categorized as hydrogenolysis reactions. It is understood that there are two possible general pathways for the HDS reaction [10]. The simplest mechanism is referred as direct desulfurization and is related to the substitution of a sulfur atom with a hydrogen atom in the hydrocarbon structure which is initially adsorbed via π-bonding in the active metal site. This happens by direct carbon-sulfur bond cleavage and is carried out without additional hydrogenation of any carbon-carbon double bond. Alternatively, in the hydrogenation route, the aromatic ring adjacent to the sulfur-containing ring is hydrogenated. The destabilization of the aromatic ring leads to the weakening of the carbon-sulfur bond before the substitution of the sulfur atom which is coordinated via π-bonding. Ultimately, hydrogenation of the aromatic ring can also be carried out after removal of the sulfur atom. It is also believed that the neighbouring sulfur-hydrogen bonds enable proton transfer in both pathways. It is accepted that sulfur removal and hydrogenation occur simultaneously on the catalyst surface, while active metal sites for both processes are considered to be the same. HDS catalysts are available in several different compositions and reactivity. The most widely applied active species are metal-sulfides, the most commonly being MoS2 or WS2 [8a]. Catalysts based on supported MoS2 are extensively used due to their low cost, high stability, and activity. These consist of two-dimensional S–Mo–S layers that are stacked to various degrees and form nano-crystalline structures related to the truncated triangle in which two types of edges are found and termed as the S-edge and the Mo-edge (Figure 35.2) [11]. The edges of these structures play the role of active sites. Because of this, catalytic activity is dependent on the orientation, form, and growth of MoS2 species over the catalyst support material. Additionally, their activity is significantly increased by the inclusion of transition metals such as cobalt and nickel which can be attached to the S-edges by coordination [12]. The support material provides enhanced surface area and mechanical stability to the catalyst and guaranties dispersion of active sites. Alumina is the most commonly employed support in HDS catalysts, others being silica, titania, or alumina-silica. Commercial HDS catalysts, and hydrotreating catalysts in general, are thus most often a porous alumina matrix impregnated with combinations of cobalt, nickel, Figure 35.2  Top-view ball model of a MoS2 nano-crystal. molybdenum and tungsten to give composites with surface area The balls denote the position of between 200 and 300 m2.g−1. Catalytic industrial processes are Mo (blue) and S (yellow and commercialized by Albemarle, ExxonMobil, Akzo Nobel, IFP, orange). Reproduced with Advanced Refining, RIPP, Haldor Topsoe, Nippon Ketjen, and permission from Ref [13] Springer Nature / CC BY 4.0. Criterion.

759

760

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

Co–Mo and Ni–Mo based catalysts are the most widely applied catalysts for HDS of most feedstock. However, their catalytic behavior is not exactly the same and differs primarily in their reactivities toward removal of heterocyclic compounds [8, 14]. Co–Mo catalysts act primarily via the direct desulfurization route and prevent the adjacent aromatic ring hydrogenation, whereas Ni–Mo catalysts have a higher selectivity for desulfurization via the hydrogenation route. The extent to which catalysis occurs via one route or the other is determining for considerations on operating hydrogen partial pressure. The two main factors for determining which catalyst should be used are pressure, as high pressure favors Ni–Mo activity, and amount of heterocyclic compounds, as low content of heterocyclic compounds favors Co–Mo catalysts [15]. Co–Mo catalysts are mostly employed in hydrotreating of straight run petroleum fractions. Ni–Mo catalysts are chosen when higher activity is required for the saturation of polynuclear aromatic compounds or for desulfurization of higher amounts of the most refractory sulfur compounds. Additionally, Ni–W catalysts are chosen when very high activity for aromatic saturation is required. State-of-the-art catalytic processes are mostly based in the combination of different catalysts. A schematic HDS process is shown in Figure 35.3. The working parameters to consider in hydrotreating processes include pressure, temperature, catalyst loading, feed flow rate, and hydrogen partial pressure. The hydrogen partial pressure must be greater than the hydrocarbon partial pressure and increasing it improves both desulfurization and denitrogenation rates. Higher temperatures will also increase the reaction rate constant and improve the kinetics. The overall desulfurization rate is ultimately determined by the rate of removal of the most refractive sulfur-containing compounds. In a standard hydrotreating process [16], the feedstock is pressurized and subsequently mixed with hydrogen gas. The feed oil can be initially preheated by a heat exchanger, but the mixture is always brought to the reaction temperature in a furnace before being fed into the catalytic reactor. The heated mixture is then hydrotreated over a catalyst bed inside the reactor, under a determined flow rate of pressurized hydrogen gas. Depending on the process design, multiple reactors can be used in processes where multi-stage conversion is employed for in-between product separation. Fixed-bed reactors are generally utilized for processing lighter feeds, whereas heavier feedstock can be processed in ebullated-bed reactors or moving-bed reactors. Different reactor systems can be integrated for the processing of complex

Figure 35.3  Process flow diagram for a hydrodesulfurization (HDS) process. Reproduced with permission from Ref [7] / Elsevier.

35.3  Adsorptive Desulfurization

feeds and many beds can be packed in a single reactor for different catalyst combinations. Both liquid and gas process fluids are cooled using a heat exchanger and separated in a series of pressure vessels. The liquid products obtained after treatment are separated into desired products in a fractionation column on the basis of their boiling points. The gas-phase mixture obtained is taken into an absorber where hydrogen sulfide gas is separated from hydrogen by using an amine scrubber.[8c]. The treated hydrogen gas is then recycled back into the reactor by using a recycle-compressor. Some of the recycle gas is purged to prevent accumulation of light hydrocarbons and to control hydrogen partial pressure. HDS presents a huge weakness since this cannot be used to marine fuels. Marine fuels have been prepared based on Heavy Fuel Oils (HFO), which present a whole set of different challenges to conventional desulfurization processes due to their complex matrices and heavy molecular composition (paraffins, cycloparaffins, aromatics, olefins, and asphaltenes, among others), with carbon numbers ranging from approximately C20 to >C50 depending on the manufacturing processes used [17] (road fuels usually contain C4 to C20 hydrocarbon chains). HDS is still employed to desulfurize HFOs [18]; however, its effectiveness is severely undermined by the aforementioned factors, along with their high metal content, combined with their coking and fouling propensity (which causes catalyst deactivation), and the molecular size and steric protection of sulfur in cyclic SCCs [19]. Other more promising technologies, involving materials as the main key to achieve the success of HFO desulfurization, have been developed, such as adsorptive and oxidative desulfurization.

35.3  Adsorptive Desulfurization ADS is a cost-effective and energetically sustainable desulfurization methodology, based on the selective removal of organic S-compounds from petroleum fractions by physicochemical adsorption processes over activated/functionalized adsorbents (Figure 35.4) [20]. Generally, ADS does not introduce impurities or affect fuel composition, and its physicochemical properties remain largely unaffected. Adsorption of S-compounds can occur through physical or chemical routes. Physical adsorption occurs essentially by Van der Walls interactions, while chemisorption can

Figure 35.4  Schematic illustration of adsorptive desulfurization (ADS) over carbon-manganese oxide nanocomposite system. Reproduced with permission from Ref [22] / Elsevier.

761

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

occur by a single mechanism or by the combination of three mechanisms: π complexation, direct sulfur-metal (S-M) interaction, and acid-base interaction. The desired characteristics for ADS adsorbents are those expected for material in other separation processes: extensive surface area and suitable pore volume, mesoporosity, surface-active sites, structural strength, and stability [21]. This chapter section focuses on the use of carbon-based materials, zeolites, mesoporous silica, and metal-organic frameworks (MOF) for ADS.

35.3.1  ADS with Carbon-based Materials C-based materials have captured broad scientific interest due to their varying properties caused by different C local bonding environments [23]. Activated carbons (ACs) are microcrystalline materials prepared from cheap and abundant precursors [24]. Processing of these starting materials under adequate conditions produces potential adsorbents with extensive surface areas, suitable porosity and unique surface properties, such as O-containing functional groups that can be modified or tuned [25]. In 1997, ACs were tested for the first time as potential adsorbents for SCCs in naphta solutions, highlighting the effects of porosity features on desulfurization efficiency [26]. More recently, it was reported that pore openings smaller than 1 nm strongly correlate with increased dibenzothiophene (DBT) adsorption capacities [27]. Furthermore, the density of acidic O-containing functional groups on the surface of ACs seems to improve ADS capacity [28]. Other investigations further revealed “pristine” ACs with high S adsorption ability but low selectivity in the presence of N compounds, which occur in real fuel matrices [29]. Thus, solvent treatment and metal loading have been attempted to increase the S adsorption capacity and selectivity of ACs. The treatment of pristine AC with NaOH or nitric acid increased the number of acidic/basic functional groups, and creating extra pores that increased ADS performance [30]. Furthermore, ACs modified with sulfuric acid after steam treatment showed different functional groups onto the AC’s surface and revealed DBT removals of over 90% (65% for pristine AC). The adsorption capacity was also correlated with the amount of surface O-containing groups [31]. Metal (M) loading on ACs aims to enhance their adsorption strength by the incorporation of M species into the C matrix. The specific role of metals for S adsorption lies in being high-energy centers for specific removal via strong interactions and π complexation with S in the confined pore space. AC doped with Cu and/or Co were tested in the ADS of model oils, revealed better removal of every SCC, notably for T, benzothiophene (BT) and 4-methylbenzothiophene (MBT), due to the strong S-M interaction (Figure 35.5) [32].

100 Removal of sulfur compounds (%)

762

Cu Co O C

Electron 250µm

Thiophene

BT

MBT

DBT

MDBT

DMDBT

90 80 70 60 50 40 30 20 10 0

AC

Co/AC Cu/AC Adsorbent

CoCu/AC

Figure 35.5  Energy dispersive x-ray (EDX) mapping image for CuCo/AC (left). Desulfurization capacity of pristine activated carbon (AC) and metal doped counterparts (right). Reproduced with permission from Ref [32] / Elsevier.

35.3  Adsorptive Desulfurization

Nanoscale C allotropes have also been explored for ADS as they possess similar surface functionalities to ACs. Graphene and carbon nanotubes (CNTs) have been tested as potential adsorbents of S from petroleum distillates. The unique mechanical properties and extensive surface areas, chemical inertness, and remarkable conductivity of graphene graphene oxide (GO) make it a prime candidate for ADS [33]. A comparative study of the adsorption of DBT on graphene and GO revealed that graphene is able of adsorbing much larger amounts of aromatic S compounds. Graphene’s hexagonal ring structure possesses a great density of π-bonds, facilitating π─π interactions with bulky Ts [34]. The density of O-containing functional groups in GO seems to disrupt its capacity to establish π-π interactions with DBT, nearly nullifying its adsorptive capacity and suggesting that S adsorption over graphene is predominantly controlled by π interactions. Despite its shortcomings as S adsorbent in its pristine state, GO has been explored to prepare hybrid nanocomposites with mixed metal oxides (MMOs) to boost its performance. A series of GO supported MMOs (MgAl, CuAl, and CoAl) were prepared and tested in ADS of DBT. The MMO-GO hybrid nanocomposites displayed remarkably superior DBT adsorption capacities, with the introduction of just as little as 5 wt% GO boosting efficiency by 170% [35]. The use of CNTs in ADS is like that of graphene, as both depend on π-π interactions, can be functionalized, and doped with nanoparticles to increase adsorption capacity. It was demonstrated that SCC adsorption over single walled CNTs depends on their diameter. Narrower tubes adsorbed larger amounts of adsorbates than their wider counterparts, as the adsorbate molecules are subjected to a larger interaction potential from the delocalized electrons of the tube, and was also verified that adsorption mainly occurs inside the nanotubes through π-interactions [36]. The comparison of the adsorption capacity of multi-walled carbon nanotubes (MWCNTs), GO, and AC, revealed that AC outperformed the other two adsorbents due to the large differences in their specific surface area (882 m2/g for AC, 217 m2/g for MWNTs and 10.8 m2/g for GO), average pore width (14.5 Å for AC, 73.8 Å for MWNTs, and 68.5 Å for GO), and micropore volume (0.487 ml/g compared to 0.286 ml/g for MWNTs and 0.021 ml/g for GO), further highlighting the importance of adequate textural characteristics for ADS [37]. In general, the C-based materials revealed a high potential for ADS; however, some limitations must still be addressed. The use of C-based adsorbents for desulfurization is extremely limited at the industrial scale due to heterogeneous pore size distributions (reducing reproducibility), poor selectivity, and difficulty of regeneration and recovery (π-interacted S is hard to remove), so further research is need to overcome these difficulties.

35.3.2  ADS with Zeolites Zeolites are microporous 3D crystalline aluminosilicates with high external surface areas, commonly used as adsorbents and catalysts, and one of the most investigated materials for ADS due to their structural features and ability to perform ion-exchange on surface active sites, which significantly boosts their adsorption capacity and selectivity toward S compounds [38]. ADS with zeolites involves all the previously mentioned adsorption mechanisms (π complexation, S-M, and Lewis acid-base interactions). Clinoptilolite is the most abundant natural zeolite and one of the most studied as an adsorbent for deep ADS. This zeolite was employed for the adsorption of SCCs after dealumination and ion exchange with Ni2+, revealing a strong correlation between Si/Al molar ratio and desulfurization capacity, and suggesting that incorporation of Ni leads to an increase in S removal through π-complexation and S-M bonds. Whereas the pristine Clinoptilolite (Si/Al = 5.65) showed a S removal percentage of 5.4%, this value increased to 68% after dealumination (Si/Al = 10.40).

763

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

Synthetic zeolites are also widely employed as adsorbents in ADS. These materials are manufactured by thermal processes, and careful control of the temperature and the composition of its precursors allows close control of the structure and the surface characteristics [39]. Furthermore, post-synthetic treatments to produce mesoporosity and the introduction of metals into the structure are the most reported methodologies to ensure maximum S adsorption. In fact, the Cu-exchanged synthetic zeolite Y selectively adsorbs T through the donation of electron charges from T to the vacant s orbital of metals and, simultaneously, back-donation of electron charges from the d orbitals of metals to T [40]. When zeolite Y was metal-exchanged with both Zn and Cu, high DBT removal efficiencies were achieved, an outcome associated to possible synergistic effects between these two metals. Removal of 4,6-dimethyldibenzothiophene (4,6-DMDBT) on Ag-Y was as effective as that of DBT, suggesting that the predominant interaction between 4,6-DMDBT and the metal-exchanged zeolite is π-complexation, as there was no significant impact from steric hindrance [41]. Modified Y-zeolites have also shown ADS capabilities for real jet fuel, with capacities of 10 mg S/g of adsorbent, at low temperatures (80 °C) [42]. β-zeolites have also been tested as potential ADS adsorbents after solvent post-synthetic treatment and ion-exchange. β-zeolite modified by alkaline treatment at different temperatures, resulted in significantly improved surface areas and pore volumes (Figure 35.6a), and revealed remarkable increases in SCC adsorption relatively to the parent zeolite (Figure 35.6b), as these bulkier species were now able to access the adsorbents’ porous structure. In addition, the ion-exchanging with Ce3+ on the previously modified β-zeolite, resulted in an adsorbent with superior deep desulfurization, attributed to the introduction of S-M interaction capabilities (Figure 35.6c) [43]. Other zeolite families like Linde type A zeolite (LTA) have been considered as potential adsorbents for ADS; however, this application fell short as LTA-zeolites do not possess adequate pore size (3–4 Å) for accommodating bulky organo-sulfur compounds [44], highlighting the importance of this crucial characteristic. With all these factors in mind, the future of ADS with zeolites is

(a) Samples Beta Beta-30 Beta-40 CeBeta-40

(wt%) 25.0 20.3 16.2 17.3

698 873 953 651

623 699 666 447

75 174 287 204

0.43 0.55 0.81 0.62

0.25 0.30 0.29 0.19

0.18 0.25 0.52 0.43

− − − 10.2

(c)

(b) 80

40

TP 3-MT BT DBT

30 Sulfur removal (%)

60 Sulfur removal (%)

764

40

20

0

beta

beta-30

beta-40

TP/toluene 3-MT/toluene BT/toluene DBT/toluene

20

10

0

beta-40

Cebeta-40

Figure 35.6  a) Textural characteristics of the investigated zeolites. b) Adsorptive desulfurization (ADS) performance of alkaline treated β-zeolites on different sulfur containing compounds (SCCs). c) ADS performance of Ce3+ ion exchanged and alkaline treated β-zeolite, compared with alkaline treated β-zeolite. Reproduced with permission from Ref [43] / Elsevier.

35.3  Adsorptive Desulfurization

strongly dependent on the development of adequate modification methodologies which endow the adsorbent with remarkable textural properties without sacrificing mechanical stability, and on the discovery of the “sweet spot” of metal ion-exchange and acid-base functionality introduction.

35.3.3  ADS with Mesoporous Silica Mesoporous silica materials (MSM) have been emerging as promising desulfurization adsorbents due to generally large specific surface area (700–1300 m2/g) and pore volume (0.5–1.2 cm3/g), tuneable pore size, narrow pore distribution, and possible synthesis in a wide range of morphologies [45]. The application of MSMs in ADS is more recent, and generally display a similar behavior to zeolites. A wide variety of MSM have been applied for ADS, with the most popular ones including Santa Barbara Amorphous-15 (SBA-15) and Mobil Composition of Matter No. 41 (MCM-41). MCM-41 with different Si/Al ratios and Ni/MCM-41 were applied for the adsorptive removal of DBT from a model oil solution, with the MCM-41 with lower Si/Al ratios revealing higher performance due to a higher density of Lewis acid sites (despite having lower surface areas). The incorporation of Ni (metal centers) into the MCM-41 structure also improved its DBT adsorption due to the creation of strong S-M interactions [46]. Also, the incorporation of distinct amounts of Ni nanoparticles on MCM-41 demonstrated that at 15% Ni loading there is a 206% increase in S adsorption capacity. However, at higher metal loadings, Ni nanoparticle agglomeration reduces the number of available Lewis acid adsorption sites due to channel blockage, reducing desulfurization efficiency. The Cu incorporation into MCM-41 by direct synthesis and incipient wetness impregnation showed that the direct synthesis method allows for better Cu dispersion across the MCM-41 structure, producing a better adsorbent than the wetness impregnation method. In terms of metal loading, it seems there is an intermediate sweet spot that ensures maximum desulfurization efficiency, as both the material with lower and the material with higher Cu loadings performed worse than the MCM-41 with intermediate metal loading [47]. Silver nanoparticles confined within the channels of MCM-41 allow the maintenance of the framework’s surface area up to a certain Ag loading, with the maximum S adsorption capacity attained at 20 wt% Ag loading (higher loadings cause silver or silver nitrate aggregates to form in the channels and block adsorption sites). These adsorbents were tested in real jet fuel, achieving S adsorption capacities four times higher than previously reported materials [48]. Mesoporous silica SBA-15 has been the subject of similar modifications. The dispersion of Ni on the hexagonal pore walls of SBA-15, resulted in a significant increase in surface area, pore volume and the number of acid sites available for S adsorption up to 20 wt% Ni loadings, with analogous disabling behavior observed for higher loadings. Similar findings were found with Cu-SBA-15 nanocomposites applied for ADS [49]. As mentioned previously for zeolites, future research into MSM should not only focus on developing novel synthetic/post-synthetic methodologies to improve textural properties, but also the investigation of strategies for the impregnation of metal species into their structures without aggregation or pore blockage.

35.3.4  ADS with Metal-Organic Frameworks MOFs have porous 3D structures generated from the rational combination of multidentate organic linkers and metal centers, producing multiple topologies, with remarkable surface areas and porosity, as well as nearly à la carte synthetic/post-synthetic tailorability [50]. The first use of a series of MOFs for the adsorptive removal of organosulfur compounds from fuels, attained better performances than those previously reported for Y zeolite [51]. It was verified that larger surface area and pore volume did not necessarily imply better desulfurization performance of MOFs.

765

766

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

MOF-177 had the highest surface area and pore volume of the five prepared materials; however, it adsorbed the least for all three organosulfur compounds, raising the question as to which structural factor influenced S adsorption the most. University of Michigan Crystalline Material-150 (UMCM-150) and MOF-505 were able to adsorb larger amounts of S due to coordinative unsaturated (or open) metal centers, taking advantage of both S-M metal bonding and physicochemical adsorption processes to ensure deep desulfurization, and demonstrating the importance of adequate functionality of the active sites (Figure 35.7) [51]. To further investigate this hypothesis, MOFs Cu-BTC, CPO-27-Ni, and ZIF-76 (ZIF = Zeolitic imidazolate framework) were prepared and tested. Despite having similar textural characteristics, Cu-BTC and CPO-27-Ni performed significantly better than ZIF-76 due to the open metal sites. Moreover, a comparative study of four MOFs, namely, MOF-5, Cu-BTC, MIL-53(Fe), and MIL101(Cr) (MIL = Matérial Institut Lavoisier), assembled from four different metal ions (Zn2+, Cu2+, Fe3+, and Cr3+) and two different organic ligands, was done to disclose the essential factors influencing adsorption performance. Interestingly, ADS performance followed the Cu-BTC ≈ MOF-5 > MIL-53(Fe) > MIL-101(Cr) order, despite MIL-101(Cr) showing the largest pore size and surface area. In fact, the ADS of aromatic S compounds can be achieved through the interactions of the metal centers in the MOF with the delocalized π electrons of the aromatic rings [52]. On the other hand, the surface functional group introduction has been reported as a strategy to improve desulfurization efficiency of MOFs. Zirconium-based MOF UiO-66 (UiO = Universitetet i Oslo) and two surface-functionalized counterparts (NH2-UiO-66 and COOH-UiO-66) were prepared and studied [53]. Despite the NH2-UiO-66 and COOH-UiO-66 showing reduced porosity relatively to UiO-66 due to the introduction of additional groups, these functionalized materials display better adsorption capabilities than the pristine framework. COOH-UiO-66 adsorbs SCCs through acid-base interactions between the carboxylate moieties and basic organosulfurs, while the NH2-UiO-66 has high density of surface NH2 groups, which can act as H donor to form hydrogen bonds with T and BT (act as H acceptors). Moreover, NH2-UiO-66 can be reused for several ADS cycles without significant loss of performance by simple ethanol washing, suggesting no leaching of functional groups.

Figure 35.7  Structures of metal-organic framework (MOF)-177, MOF-5, University of Michigan Crystalline Material (UMCM)-150, Copper benzene-1,3,5-tricarboxylate (Cu-BTC), and MOF-505, with one molecule of DBT added in the pore of each MOF to represent scale. Adsorption isotherms for different benzothiophene (BT) (left), dimethyldibenzothiophene (DMDBT) (middle), and dibenzothiophene (DBT) (right) for the different MOFs. Reproduced with permission from Ref [51] American Chemical Society.

35.4  Oxidative Desulfurization

Mixed metal loading is also a potential tool to enhance MOF S adsorption capabilities. The CuCl2-loaded MIL-47 showed a similar behavior to that observed for metal-loaded silicas: “intermediate” loading weights provide better ADS performances, which decrease with further loading increase due to the blocking of metal active adsorption sites. Comparing the porosity of pristine MIL-47 and Cu-loaded MIL-47 with their adsorption capacities, it is once again apparent that this is not the only factor impacting performance. In fact, the reduction of CuII to CuI, which occurs in the presence of V (the metallic component of MIL-47) produces synergistic effects with the framework itself which facilitate π-complexation with adsorbed SCCs. The incorporation of task-specific ionic liquids (ILs) in MOFs is another interesting approach to ADS with modified MOFs. 1-butyl-3-methylimidazolium chloride was immobilized in the pores of MIL-101(Cr), resulting in a reduction of the pore volume with gradually increasing IL amounts, suggesting they are located inside the porous framework [54]. The adsorbed S quantity by IL/MIL-101 increased up to 33% IL/MIL-101 and decreased for 50% IL/MIL-101 due to a high degree of pore blockage with excess amounts of ILs. ILs supported on MIL-101 act as acidic centers toward the basic S species, resulting in a favorable interaction between IL/MIL-101 and increasing the adsorption capacity by 71% in relation to pristine MIL-101. This aproach opens interesting possibilities into the development of ADS efficient MOFs, as there are hundreds of reported MOFs structures and ILs, thus, many MOF/IL combinations can be prepared with interesting properties favoring effective S/ adsorbent interactions. Future research on MOFs for ADS should focus on the development of structures with adequate porous systems, which allow for selective S removal, as well as further chemical functionalization strategies which can significantly enhance MOF─S interactions and allow for ultra-low sulfur diesel (ULSD) production.

35.4  Oxidative Desulfurization ODS is an energetically sustainable deep desulfurization methodology that involves a chemical reaction between an appropriate oxidant and the sulfur compounds (SCCs). This process can be performed under mild reaction conditions of temperature and pressure and using environmentally benign oxidants. The oxidation of SCCs present in the fuel matrix occurs by the presence of an oxidant that is activated by an appropriate catalyst. The reactivity of SCCs depends on their electron density and steric accessibility. Comparing the oxidative facility of T, BT, DBT, and 4,6-DMDBT under the same conditions, this follows the order DBT > 4,6-DMDBT > BT > T. The electron densities on the sulfur atom from 4,6-DMDBT, DBT, BT and T are 5.760, 5.758, 5.739 and 5.696, respectively. Therefore, compounds with higher electron densities around the S atom are more readily oxidized into their respective sulfoxides and/or sulfones [55]. The DBT is more easily oxidized than the 4,6-DMDBT despite having lower electron density. However, 4,6-DMDBT presents a spatial steric hindrance around the S atom caused by alkyl substitution, obstructing the formation of new S=O bonds [56]. The oxidative catalytic reaction can be performed in the presence or the absence of an extraction solvent that need to be immiscible with the treated liquid fuel. When the extraction solvent is used, ODS occurs simultaneously with the extractive desulfurization (EDS). This last bifunctional extractive and catalytic oxidative desulfurization process (ECODS) promotes a deep desulfurization, where the SCCs are extracted from the liquid fuel due to the difference of polarity and the higher affinity of SCCs by the higher polar extractive phase. In the latter occurs the oxidation of the SCCs to produce the respective sulfoxide and/or sulfone compounds [57].

767

768

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

35.4.1  Oxidants for (EC)ODS The choice of an adequate oxidant is crucial for the success of ODS. The first patent in the area of ODS, published in 1974, reported the use of nitrogen dioxide (NO2) as an oxidant followed by extraction with an organic solvent to remove oxidized SCCs from petroleum distillates [58]. Air has been successfully used as oxidant to convert SCCs from gasoline and diesel fuels to sulfones, in the presence of γ-butyrolactone as an oxygen transfer agent, at 140 oC, affording high conversion rates and selectivities [59]. Sundararaman et al. also proposed using air as an oxidant for the desulfurization of a commercial jet propellant-8 (JP-8) jet fuel and a commercial diesel fuel in the presence of CuO catalyst, which promote in situ generation of peroxides [60]. Other option, is the use of ozone, since this is a strong oxidant. Ma et al, used ozone produced by dielectric barrier discharge plasma combined with IL extraction [61]. The desulfurization efficiency for TS and BT was investigated and both were removed to achieve under 0.1% of S in the presence of O3/[BMIM]Ac [61]. Wang et al. used ozone and hydrogen peroxide and 99.1% of desulfurization was achieved for DBT. Using this combination of oxidants guaranteed that the DBT could be oxidized in both fuel and solvent extraction phases; while aqueous H2O2 ensure a complete oxidation in the polar extraction phase [62]. The use of tert-butyl hydroperoxide (TBHP) as an oxidant brings some disadvantageous, such as the high price and the production of t-butylalcohol as by-product. The presence of peracetic acid in ODS is easily obtained in situ from the reaction between H2O2 and carboxylic acids [63]. Dehkordi et al. reported the desulfurization of kerosone using a combination of H2O2 and acetic acid, observing that the desulfurization efficiency increased when the ratio acid/S increased, achieving as the best performance to 83.3% of S removals [64]. The solo H2O2 is often cited as the best oxidant for ODS, not only due to its high amount of active oxygen by mass unit, but also due to its superior sustainability claims, since the only by-product is water [65]. The first report on the use of H2O2 in ODS was published by Shiraishi et al. in 1999 and showed that the sulfur content in real fuel was reduced from 0.2 to 0.05 wt% even requiring 48 h of reaction [66].

35.4.2  Heterogeneous Catalysts for (EC)ODS The main goals of heterogeneous catalysis research are the development of remarkably active, selective, and recoverable/reusable catalysts, tailored to the desired reactions. Textural properties play a major role toward these goals, as different types of porosity tend to enhance/hinder different types of interactions with substrates, thus, it is desirable for heterogeneous catalysts to exhibit hierarchical porosity (i.e. combine micro/mesoporosity to increase surface area and macroporosity to enhance substrate transport capabilities) [67]. Combining supporting materials that exhibit these properties with catalytically active species as active centers allows for the production of heterogeneous catalysts that also act as molecular sieves, enhancing selectivity, with the porous system acting as micro-reactors. 35.4.2.1  (EC)ODS with Zeolites

Zeolites have recently received attention as potential heterogeneous desulfurization catalysts due to their desirable combination of micro- and mesoporosity along with excellent thermal and chemical stabilities. Titanium silicalite-1 (TS-1) is one of the most important and widely used zeolite as heterogeneous catalyst in several important reactions, including ODS, due to its remarkable efficiency and high selectivity under mild conditions [68]. Nevertheless, the microporous framework with narrow and tortuous channels strongly limits its application in catalytic reactions involving large-dimension reactants and products [69]. For that reason, several efforts have been

35.4  Oxidative Desulfurization

made to develop hierarchically structured TS-1 zeolites combining the typical micropores with additional mesoporosity [70]. Du et al. prepared titanium silicalite-1 (TS-1) zeolite using polyvinyl butyral gel as the mesopore template, producing optimized TS-1 structures that combine meso- (30 nm), micro- (0.55 nm) and super-microporosity (1.6 nm) with catalytically active tetrahedral Ti species. Commercial TS-1 reached 22% conversion of DBT after 40 minutes, due to is exclusively microporous structure which caused steric hindrance for DBT access to active sites. TS-1 derived structures obtained from mesopore templating displayed much higher catalytic activities, reaching over 95% DBT conversions in under 20 minutes, due to the introduction of abundant meso- and super-micropores in the rugged surface and interior of the zeolite, enhancing its DBT transport capabilities [71]. Kong et al. have also reported enhanced desulfurization performance of acidtreated TS-1 in the oxidation of T using H2O2 as oxidant. In fact, the HCl treated TS-1 samples were able to reach 79.6% conversion in 30 minutes while the pristine zeolite was only able to convert 37.1% of T. The superior performance of the treated samples is attributed to the removal of noncoordinated Ti from the framework which enhanced porosity and enabled substrate access to catalytically active sites [72]. Yu et al. have proposed an alternative method for the preparation of hierarchical TS-1 using a green surfactant as mesoporous template [73]. Materials with uniform intracrystalline mesopores were prepared using Triton X-100 as assisted surfactant during hydrothermal synthesis. The hierarchical TS-1 materials were evaluated as heterogeneous catalysts in the ODS of T, BT, DBT and 4,6-DMDBT using H2O2 or TBHP as oxidant. The materials exhibited enhanced catalytic activity when compared with conventional TS-1 and were able of being recycled for at least four consecutive cycles. 35.4.2.2  (EC)ODS with Metal-organic Frameworks

Owing to their aforementioned properties of excellent stability, porosity and tailorability (adjustable acidity/basicity, introduction of functional groups, etc.), it comes as no surprise that MOFs have played a key role in the research of sustainable catalytic oxidation desulfurization. MOFs exhibit 3 important components which can grant them catalytic importance: the metallic components, the organic linkers, and the porous framework structure. In the case of intrinsically catalytically active MOFs, the first two components are of utmost importance. Metallic centers in MOFs can directly coordinate with substrates or oxidants to form reactive oxygen species through the expansion of their coordination sphere, or the reversible delocalization of labile linkers [74]. The organic linkers can provide catalytic activity to a MOF through the presence of adequate functional groups [75]. While examples using V [76], Cr [77] or Co [78] MOFs have been reported, the majority of literature regarding intrinsically active MOFs for ODS is based on metals from the IV (Ti, Hf, Zr) group, due to their exceptional stability and remarkable catalytic activity. The first report of an intrinsically active Zr MOF for application in ODS was published by Granadeiro et al., using the porous Zr(IV) terephthalate UiO-66 for the desulfurization of model oil containing DBT, 4-MDBT and 4,6-DMDBT. The authors suggest that the catalytic activity of UiO-66 must be related to the formation of ZrIV-peroxo groups on the surface of the material by the interaction with H2O2, with this MOF achieving complete desulfurization of the model oil after just 30 minutes. Another important result of this work was the correlation of catalyst performance with the degree of crystallinity of the prepared UiO-66 samples, with the least crystalline structures displaying enhanced catalytic activity due to a higher number of coordinative unsaturated zirconium sites, facilitating the formation of active oxygen species [79]. Following this study, Viana et al. compared the catalytic activity of UiO-66 samples prepared under different methodologies (solvothermal and microwave assisted synthesis), noting that faster post-synthetic cooling promotes a higher incorporation of Cl anions (arising from the zinc precursor) in the UiO-66(Zr) framework, which can be correlated

769

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

with a higher number of defects in the structure and a higher catalytic activity [80]. Viana et al. also proposed a straightforward activation of low-defect pristine UiO-66 (Zr and Hf), using chloride-based salts, aiming to introduce catalytically active defects into the frameworks. The activation method performed on UiO-66(Hf) by the insertion of chloride into the framework structure led to an increase of the desulfurization performance from 48% to 91% (Figure 35.8) [81]. Another Zr-MOF obtained using trimesic acid as organic linker, MOF-808, has also been extensively reported in ODS related literature. Similarly to UiO-66, pristine MOF-808 is less active in oxidation reactions than its defect-containing analogues [82]. Gu et al. reported the acid-treatment of pristine MOF-808 aiming to create abundant defect sites in its structure by the partial removal of organic linkers. The HCl treatment did not alter the MOF-808 morphology or particle size, but caused slight improvements to its textural properties which, associated with available active metal sites from linker removal, ensured this defective MOF-808 to achieve near complete desulfurization of model oil in five minutes, a significant improvement from the performance of the pristine MOF (60% desulfurization) [83]. Similar results were reported for the preparation of defective engineered MOF-808 with enhanced catalytic activity in ODS processes [84] The methods typically involve the use of formic acid as modulator agent during synthesis followed by the post-synthetic removal of formate with the consequent generation of open metal sites in the MOF structure. Zheng et al. compared a series of Zr-based MOFs (UiO-66, UiO-67, NU-1000, and MOF-808) with intrinsic peroxidase-like activity for ultradeep ODS. All the selected Zr-MOFs possessed intrinsic peroxidase-like activity and exhibited catalytic activity in ODS with H2O2 as oxidant. Nevertheless, MOF-808 could completely remove DBT within five minutes while UiO-66, UiO-67, and NU-1000, only achieved 8.8%, 20.8%, and 11.1% respectively [85]. Ti-based MOFs such as Ti terephtalate MIL-125 and Ti-carboxylate COK-47 structures have also been reported as active desulfurization catalysts, with their activity strongly influenced by the density of structural defects. Li et al. prepared size-modulated MIL-125, inducing hierarchical porosity and structural defects on the nanocrystals. Increasing amounts of modulator led to a crystal size reduction from c. 1500 nm to 50 nm, with a significant surface area increase of the MOF due to the downsizing of the nanocrystals and the prevalence of missing-linker defects. The “defective” MIL-125 achieved complete desulfurization of model oil in 30 minutes, while pristine MIL-125 was only capable of removing 80% of DBT and 45% of 4,6-DMDBT [86]. Smolders et al. recently reported a new Ti-based MOF comprised of a layer of TiO6 octahedra forming a three-dimensional framework through the connection of bpdc2− linkers. This MOF, baptized

100

80 60

UiO-66(Zr)

40

UiO-66(Zr)-TiCl4

20

UiO-66(Zr)-ZrCl4

0

0

0.5

1

1.5

2 2.5 time/hour

3

3.5

4

total desulfurization / %

100 total desulfurization / %

770

80 60 40 UiO-66(Hf)-TiCl4

20 0

UiO-66(Hf) 0

0.5

1

1.5

2 2.5 time/hour

3

3.5

Figure 35.8  Desulfurization profiles of defect free UiO-66(Zr) and activated counterparts (left). Desulfurization performance of pristine UiO-66(Hf) and activated counterpart (right). Reproduced with permission from Ref [81] / Elsevier.

4

35.4  Oxidative Desulfurization

COK-47, possesses a remarkable density of methoxy groups (Me-O-Ti) and structural defects producing open Ti-sites, allowing for the efficient generation of reactive oxygen species when in contact with TBHP and achieving 99% DBT oxidation in 120 minutes [87]. Despite the previously described works, intrinsically active ODS MOF catalysts are, however, rare, and most literature regarding the use of MOFs for the preparation of heterogeneous desulfurization catalysts deals with their use as solid supports for catalytically active homogeneous species. 35.4.2.3  (EC)ODS with Carbon-based Materials

Haw et al. prepared a series of functionalized ACs with micro/mesoporosity features. Functionalization was achieved with phosphoric acid activation which converts carbon surface hydroxyl groups into oxygen-containing acidic moieties. H2O2/acetic acid mixture was used as the oxidant, producing peroxyacetic acid which was then decomposed by the ACs, forming •OOH free radicals, which were the main oxidant for the reaction. H2O2 can also be directly decomposed by the functionalized ACs by transfer of an electron from the reducing site of the AC to form •OH radicals which can also oxidize SCCs. The prepared functionalized-ACs can thus act both as catalysts and adsorbents in this process, providing the adsorption surface for the organosulfur compounds and oxidation reaction to take place, while an acetonitrile extraction phase ensures the continuous transfer of SCCs from the oil phase allowing their removal. The functionalized-ACs were able to reduce the sulfur content of a commercial diesel fuel from 2,189 ppm to 190 ppm of sulfur after three cycles [88]. Kampouraki et al. performed chemical modification of ACs via treatment with two different acids (HNO3 and H2SO4) aiming to introduce additional functional groups onto their surfaces. The surface chemistry of carbon materials, expressed by the density of the acidic functional groups, was found to be a critical parameter with regards to both adsorption and catalytic oxidative performance. The authors also noted that among the several tested ACs, materials with higher abundance of smaller micropores were more efficient due to increased confinement effects caused by size similarity with 4,6-DMDBT. A significant increase of the removal efficiency via catalytic oxidation can be observed for both acid-treated AC samples, reaching complete sulfur removal, further suggesting the positive effect of surface acidity on the catalytic oxidation (Figure 35.9). These surface groups can also be responsible for the strong retention of the generated oxidation products (sulfoxide and sulfones) [89].

Figure 35.9  Quantification of acidic and basic surface functional groups of the parent activated carbon (AC) and its counterparts oxidized by HNO3 and H2SO4 (left). Comparison of 4,6-dimethyldibenzothiophene (4,6-DMDBT) adsorptive and catalytic oxidation removal of the parent activated carbon (AC) and its counterparts oxidized by HNO3 and H2SO4 (right). Ref [89] / Royal Society of Chemistry / CC BY 3.0.

771

772

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

Nanoscaled carbon-based materials have also showed promising results as heterogeneous catalysts for combined adsorption/oxidation/extraction sustainable desulfurization processes. Gu et al. reported the use of metal-free reduced graphene oxide for the oxidative removal of SCCs, due to presence of chemically active defects in the graphene oxide structure which generate carbonyl groups in situ [90]. Oxygen molecules could interact with the carbon atoms adjacent to carbonyl groups to form super oxygen anion radicals, which then oxidize the adsorbed organosulfurs. The reduced graphene oxide material has proved to be a highly efficient heterogeneous catalyst for aerobic ODS reaching 96.1%, 90.5%, 100%, and 97.7% of sulfur removal for 3-methylbenzothiophene (3-MBT), BT, DBT, and 4,6-DMDBT, respectively. Recently, a new class of highly ordered porous materials has been reported using MOFs as selfsacrificial templates, the so-called MOF-Derived Porous Carbons (MDPCs) [91]. MDPCs are obtained by carbonization of MOFs under inert atmosphere resulting in uniform metal/metal oxide nanoparticles distributed in the ligand-derived carbon matrix [92]. The high specific surface area, tunable porosity, non-toxicity, lightweight and simple synthetic procedures of MDPCs makes them highly suitable for a wide range of application perspectives [93]. Hicks et al. reported a nanoporous carbon-based material obtained by pyrolysis of titaniummodified zinc-containing IRMOF-3 [94]. The pyrolyzed material exhibited higher porosity, surface area and thermal stability than the corresponding pristine MOF. Moreover, its catalytic activity in the oxidation of DBT using TBHP as oxidant was significantly enhanced as well as its recycling ability in consecutive catalytic cycles. The MIL-47(V) has also been used as template for the preparation of MDPCs for application in the oxidation of DBT [95]. The authors demonstrated that the pyrolysis temperature influences the V phase (V2O5, V2O3, VO2, V8C7) present in the final carbon material. The best desulfurization performance was attained with V carbide carbon materials showing increased catalytic activity, enhanced chemical resistance to the oxidant (TBHP) and reduced leaching of active species when compared with the other prepared heterogeneous catalysts. Another report deals with the production of novel MDPCs obtained by pyrolysis at different temperatures of hierarchical microporous/mesoporous MIL-125 frameworks prepared by a vaporassisted technique [95]. The pyrolyzed mesoporous MIL-125-derived materials were evaluated as heterogeneous catalysts in the oxidation of DBT using TBHP and their activity compared with pyrolyzed microporous MIL-125. The results reveal a direct relationship between the catalytic activity and the pyrolysis temperature as a result of the increasing Ti content. The authors suggest that the enhanced desulfurization performance can be attributed to the higher amount of mesopores and smaller Ti particle size present in the mesoporous-derived materials. Jhung et al. described the pyrolysis of Ti-loaded zinc-containing hydrophobic MAF-6 and hydrophilic MOF-74 [96]. The adequate combination of the type of MOF and solvent (hydrophobic/hydrophilic) allowed to control the location of the loaded Ti-precursors (inside/outside of the porous framework). The catalytic activity of the pyrolyzed carbon materials on the oxidation of DBT seems to be favored when the initial Ti-precursors are mainly located inside the porous framework (prior to pyrolysis) as it leads to a smaller size of TiO2 particles after the thermal treatment. In another report by Jhung et al., mesoporous MDPC materials were prepared by pyrolysis under inert atmosphere of bimetallic Zn/Ni or Zn/Mn MOF-74-type structures [97]. Several MOF precursors with different metallic ratios were prepared and its influence on the properties of the final carbonized materials was evaluated. The increasing Zn content in the MOF precursors lead to MDPCs with higher extent of mesopores and better catalytic activity in DBT oxidation using H2O2 (>95% of 1000 ppm S in 120 minutes). A similar strategy was employed for the preparation of Co-supported on nitrogen-doped porous carbon by pyrolysis of bimetallic Zn/Co MAF-6 with different metal

35.4  Oxidative Desulfurization

ratios [98]. The pyrolyzed carbonaceous materials exhibited increased porosity and enhanced catalytic activity in the ECODS removal of SCCs from model fuel compared with the parent MOFs. In particular, the MDC-6(75Zn25Co)-900 sample revealed promising results reaching 93.6% of DBT conversion after 120 minutes using H2O2 as oxidant and acetonitrile with good recyclability for five consecutive cycles. 35.4.2.4  (EC)ODS with Mesoporous Silicas

The large and accessible porous framework of mesoporous silica allows the facile diffusion of bulky reactants and products in oxidative reactions. However, the ODS activity of pristine mesoporous silicas is practically inexistent as a result of the amorphous nature of their inner channels [99]. To overcome this limitation, several efforts have been made to prepare Ti-functionalized mesoporous silica for application in the removal of SCCs [100]. By doing so, researchers aimed to combine the exceptional ODS activity of Ti-species with the channel-type porous structure of mesoporous ordered silica for fast diffusion of reactants and products. Several reports can be found in the literature dealing with the preparation of heterogeneous ODS catalysts based on Ti-functionalized hexagonal mesoporous silica (HMS), SBA-15, SBA-16, and MCM-41 [101]. Lee et al. reported a Ti-SBA-15 catalyst obtained through post-synthetic grafting of Ti using tetrabutyl orthotitanate (TBOT) as source [101c]. The Ti-SBA-15 material was tested as heterogeneous catalyst in the ODS of model feed and real light cycle oil (LCO) with TBHP as oxidant. The effect of nitrogen containing compounds (NCCs), aromatic and aprotic solvents on the catalytic activity of Ti-SBA-15 was evaluated. The results reveal that the presence of NCCs decreases the ODS activity, although the addition of aromatic (tetralin or 1-methylnaphthalene) or aprotic (acetonitrile) solvents allows to recover the catalytic activity. In these conditions, the catalyst exhibited a remarkable desulfurization performance by attaining complete ODS conversion of real LCO with 3,700 ppm S. Beltramone et al. prepared heterogeneous catalysts based on titanium-modified SBA-16 with tetrahedral or octahedral Ti coordination and compared their ODS activity in the oxidation of DBT using H2O2 as oxidant [100]. The results showed superior ODS performance of octahedral coordinated Ti-SBA-16 reaching 90% of DBT (2000 ppm) conversion after 60 minutes without loss of activity in four cycles. An alternative heterogeneous ODS catalyst based on mesoporous silica has been proposed by Plikarpova et al. [102]. Instead of the typical transition metal-functionalized silicas, the authors developed a cost-effective non-metal ODS catalyst by chemical immobilization of sulfonic groups on the surface of mesoporous MCM-41. MCM-SO3H proved to be a highly efficient catalyst reaching complete DBT and 4-MDBT conversion (500 ppm) in 120 minutes with H2O2 as an oxidant. Moreover, the chemical coordination of sulfonic groups avoids leaching of the active species and allows the catalyst to retain its catalytic activity for 10 consecutive ODS cycles. 35.4.2.5  (EC)ODS with Titanate Nanotubes

Recently, titanate nanotubes (TiNTs) have emerged as an alternative type of Ti-containing mesoporous materials for ODS applications. TiNTs have attracted the interest of the scientific community owing to their high surface area, mesoporous structure, morphology, low cost and possibility to scale-up [103]. The majority of reports deals with the use of titanate nanotubes as supports for catalytic active species, while only a few works investigated their intrinsic catalytic properties. Regarding ODS application, Lorençon et al. described the alkaline treatment of TiO2 anatase to produce sodium (Na-TiNTs) and hydrogen-titanate nanotubes (H-TiNTs) [104]. Catalytic studies revealed that the counter-ion has significant influence on the catalytic activity of TiNTs. In fact, Na-TiNTs showed no catalytic activity in the DBT conversion with H2O2 as oxidant while H-TiNTs exhibited high efficiency by achieving complete DBT removal (500 ppm) after just 60 minutes and

773

774

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

could be recycled for five cycles without significant loss of activity. The authors propose that the mechanism involves the interaction of superficial Ti(IV) sites with the oxidant to form superoxide radicals. These radical active species are able to oxidize DBT into the corresponding sulfones. Similar results have been reported by Yao et al. by also studying protonated titanate nanotubes (H-TiNTs) as heterogeneous catalysts for desulfurization of fuels [105]. Cedeño-Caero et al. have compared the catalytic activity of TiNTs (pristine and calcined) and Ti-substituted mesoporous SBA-15 in the ECODS of DBT, 4-MDBT and 4,6-DMDBT with H2O2 as oxidant [106]. The catalytic results showed a superior performance of TiNTs over Ti-SBA-15 materials which, according to the authors, might be attributed to the higher exposure of superficial Ti-species in TiNTs. Interestingly, calcined TiNTs, despite having smaller surface area than the uncalcined material, exhibited superior desulfurization performance. In fact, calcined TiNTs reached a remarkable 99% of DBT conversion in just 60 minutes while the uncalcined material converted 90.5% for the same reaction time.

35.4.3  (EC)ODS Catalyzed by Heterogeneous Polyoxometalates Transition metal compounds have been receiving considerable attention as catalytic active centers for ODS. Some of these compounds containing V, Mo, W, Nb, and Cr form peroxo-compounds by the interaction with H2O2. A particular class of transition metal compounds largely used in ODS are the polyoxometalates (POMs). These are anionic metal oxides, comprised of [MOx]n (M = Mo, W, V, Nb; x = 4–7) basic units, producing a variety of possible structures with interesting physicochemical properties [107]. The use of POMs in ODS is centered on their ability to form easily active peroxo-compounds, increasing the electrophilic character of peroxidic oxygens [108]. One of the most active POMs derived from Keggin structure [XM12O40]n−, forming easily peroxo-POM species [109]. Among the different transition metals that can incorporate POM structures, MoVI and WVI stand out due to a favorable combination of ionic radius, charge and accessibility of empty d orbitals for metal-oxygen π bonds [110]. In 1997, Collins published the first report of using POMs in ODS, using the phosphotungstic acid (PW12) for the single oxidation of DBT with H2O2 [111]. Collins suggested that there is a fast reaction between H2O2 and the catalyst to produce active species which ensures the oxidation of DBT [111]. Wang et al. noted the influence of the Brønsted acidity and the cation in the ODS performance of POMs [112]. The efficiency of POMs as catalysts for ODS systems to treat fuels led to more than 200 papers to date, in which various reaction conditions, extraction solvents, oxidants, fuel feedstock were attempted. More recent studies are dedicated to the preparation of POMs composites (i.e. their heterogenization using various support materials). This is a clever procedure to use active catalytic centers able to be recycled. Several successful examples of POMs composites for ODS using model and real fuels are reported. 35.4.3.1  Carbonaceous Composites

Recently, several commercial ACs differing in acid-base properties were tested as supports for active POM catalysts. Various Keggin-type POMs structures decomposed due to their interaction with more basic ACs, forming monomeric and/or oligomeric MVI oxo species on the carbon surface, which showed to enhance POM-AC composite activity, allowing 100% of DBT oxidation in just 30 minutes [113]. Further, these catalysts showed to be stable and recyclable for the oxidation of DBT [113]. Previously, Dizaji et al. took a different approach and immobilized different POMs on graphene oxide, using these for ECODS processes to treat multicomponent model fuel (BT, DBT and 4,6-DMDBT) [114]. SCCs were continuously extracted from model fuel and adsorbed into the

35.4  Oxidative Desulfurization

graphene oxide surface. SCCs were then oxidized and desorbed from the graphene oxide surface, freeing the active sites for further reactions. Complete desulfurization was achieved after 30 minutes [114]. In the beginning of last decade, Wang et al. prepared a Multi-walled carbon nanotube composite Cs2.5H0.5PW12O40/MWNT which was found to be very effective for the oxidative removal of DBT, with a desulfurization efficiency of up to 100%. The catalyst was reused without significant loss of activity; however, lower efficiencies were found for the most difficult to oxidize SSCs [115]. 35.4.3.2  MOF Composites

The encapsulation of homogeneous active POMs into porous MOFs has been forming high performance heterogeneous catalysts for ODS. In 2013, Ribeiro et al. presented the first published work using POM@MOF catalysts, in which the tetrabutylammonium salt TBA3PW12O40 (TBA = Tetra-n-butylammonium) was encapsulated in the chromium terephthalate metal-organic framework MIL-101(Cr), following an impregnation method [116]. This new heterogeneous catalyst showed to have a similar activity than the homogeneous POM (i.e. near complete desulfurization obtained after three hours) [116]. The composite TBA3PW12O40@MIL-101(Cr) was reused for three cycles with minimal loss of activity, attributed to the leaching of the active center [116]. The absence of POM leaching was found some years later by Wang et al. using an amine-modified NH2MIL-101 support [117]. The presence of amine-functional group in the MOF structure showed to be essential to prevent POM leaching via electrostatic interaction. More recently, Wang et al. presented a correlation between desulfurization efficiency and the pore and the window size of different MOF structures: MIL-100(Fe), UiO-66, and ZIF-8 [118]. These were used as support materials for the encapsulation of [PW12O40]3− active centers and these solid catalysts were used to treat model and real gasolines. From these MOFs, the POM size was smaller than the MOF nanocage sizes and larger than MOFs window sizes, preventing POM leaching. However, the catalytic performance in these cases was lower probably due to the mass transfer limitations [118]. The best catalytic performance was obtained for POM@MIL-100(Fe) (8.6 × 8.6 Å) achieving a desulfurization efficiency of 92 % to treat a multicomponent model fuel (BT, DBT, and 4,6-DMDBT) using H2O2 as oxidant; while UiO-66 (6.0 × 6.0 Å) displayed high desulfurization efficiency only for the smaller size sulfur compounds (i.e. BT and DBT), because these are more easily diffused into the inner pores of UiO-66. The low activity of ZIF-8 was ascribed to its narrow pores (3,4 × 3,4 Å), which limit the sulfur compounds diffusion [118]. 35.4.3.3  Zeolite Composites

Zeolites have also been modified via the introduction of POMs into their structures to produce active catalysts for organosulfur oxidation. One of the most recent works was published by Wang et al. that dispersed PW12 on the MWW zeolite, which consists of two ranges of independent porous structures (sinusoidal channels and supercages), and tested the composite catalyst for the desulfurization of model oil, straight-run gasoline, and fluid catalytic cracking gasoline [119]. The introduction of 38 wt% of W into the zeolite boosted its desulfurization efficiency from 35 to 99%. The prepared catalyst was reused for three consecutive cycles with minimal loss of activity (91.4% oxidation on the third cycle) due to the adsorption of oxidation products. These were easily removed by washing with ethanol and the composite was once again reused, achieving 98% S removal. The POM-MWW catalyst was also able to remove 98% of S content from straight-run gasoline, and 47% of organosulfurs from fluid catalytic cracking gasoline, due to its high olefin content, which hinders efficiency [119]. One previous work was reported by Zhang et al. that prepared cup-like hollow Zeolite Socony Mobil–5s (ZSM-5s) which were then used as a support for a POM@MOF composite of phosphomolybdic acid (PMo12) in MOF-199 [119]. Each ZSM-5 cup acts

775

776

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

as a micro-reactor allowing for efficient formation of reactive oxygen species by the POM@MOF composite and subsequent adsorption of DBT, leading to its oxidation. The introduction of the catalytically active centers achieved 91% of desulfurization, a significant increase from the pristine zeolite, 21% [120]. One of the first works using POM@zeolite to (EC)ODS first was presented by Wang et al. in 2011 and used a ship in the bottle methodology to encapsulate PMO12 in Y-type zeolites [121]. Using H3PO4 and MoO3 as the P and Mo precursors, respectively, this preparation method allows for in situ POM formation in the zeolite’s surface and pores. Surface POMs are then washed away with water, unblocking substrate and oxidant access to the porous structure’s microreactors. The Y-zeolite with 10% Mo loading achieved 85% DBT oxidation in 20 minutes and was recycled for more three cycles with just 0.7% loss of activity [121]. 35.4.3.4  Mesoporous Silica Composites

To guarantee an effective immobilization of active POMs on the surface of mesoporous silica supports, these hydrophilic materials need to be pre-functionalized (i.e. their surface needs to be modified by the incorporation of organic functional groups). The presence of strategic functional groups will promote an interaction via electrostatic or covalent with the POM and it will prevent leaching occurrence. In 2019, Julião et al. presented successful work that modified SBA-15 by functionalization with propyltrimethylammonium groups (TMA) to immobilize tetranuclear peroxotungstate TBA3{PO4[WO(O2)2]4} (PW4) [122]. The composite PW4@TMA-SBA-15 achieved a desulfurization efficiency similar to the homogeneous PW4 to treat model oil and real fuels (Figure 35.10). This heterogeneous catalyst could be recovered and reused for at least 10 consecutive cycles with the first to eighth cycle achieving complete desulfurization and minimal losses of activity between the eighth and tenth cycle [122]. In the same year, Ribeiro et al. performed structural modification on a Keggin-type POM, obtaining a catalytically active lacunar structure which was then supported on amine-functionalized SBA-15. Surface functionalizing with amines minimizes leaching owing to the strong interaction between the POM via dative bonding. The catalytic activity of the PW11@aptesSBA-15 composite was similar to that of the homogeneous PW11, with the heterogeneous catalyst displaying remarkable activity for the ODS of real diesel, achieving 83.4% desulfurization when applied in a tandem EDS/ODS system. The recycling capacity of the composite was confirmed for eight cycles [123]. Ribeiro et al. also tested an ethylene-bridged periodic mesoporous organosilica (PMO), functionalized with a cationic group from N-trimethoxysilypropyl-N, N, N-trimethylammonium (TMA), for the immobilization of the same PW11. The same functionalization and immobilization was

(a)

(b)

4 µm

2 µm

Figure 35.10  Scanning electron microscopy (SEM) micrographs of the solid support TMA-SBA-15 (a) and the composite material PW4@TMA-SBA-15 (b) (left, center). Comparison of the desulfurization profiles of the homogeneous and heterogeneous catalysts (right). Reproduced with permission from Ref [122] / Elsevier.

35.5  (EC)ODS Catalyzed by Membranes

performed using SBA-15 support. The PW11@TMA-SBA-15 and PW11@TMA-PMOE composites were tested as heterogeneous catalysts in the ODS of both model and real fuels. The PW11@TMASBA-15 catalyst was able to achieve complete desulfurization for DBT, 4-MDBT and 4,6-DMDBT, and 93.9% for 1-BT, after 30 min, while PW11@TMA-PMOE reached, 92.8% for 1-BT, 98.2% for DBT, 99.0% for 4-MDBT and 99.3% for 4,6-DMDBT after 60 minutes of catalytic oxidation, resulting in a total desulfurization of 96.9%. PW11@TMA-SBA-15 retained its activity for six consecutive cycles, at which point some loss of performance was observed, ascribed to active site deactivation by the presence of strongly adsorbed sulfones. This composite was also used in the desulfurization of real untreated diesel, achieving 93% S removal [124].

35.5  (EC)ODS Catalyzed by Membranes Applying membrane technology to the desulfurization of industrial fuels can be one of the most efficient approaches to mitigate the frequently reported catalyst mass losses during recycling and deactivation due to active species leaching and sintering [125]. This approach may be the defining step toward making other desulfurization technologies, such as ADS and/or (EC)ODS as true alternatives for application at an industrial level, as it dramatically enhances catalyst handling and may lower operating and energy costs, allowing for process scale-up. In fact, several reports can be find in the literature preparing mixed-matrix membranes (MMM) for desulfurization related applications, mainly for sulfur extraction. These works are mostly focused on using polymeric membranes due to their wide availability, low cost and ease of preparation [126]. These MMMs are based on the combination of organic polymers with some of the materials previously discussed in this chapter (carbon-based materials, silicas, and MOFs) and the reports mostly focus on their applications for selective adsorptive or pervaporative desulfurization. Recently, Peng et al. supported graphene oxide in a polyurethane membrane through a simple methodology of mixing each of the MMM components, followed by solvent evaporation [127]. The prepared MMMs were thiophenepermselective, displaying a significantly better separation performance over the non-modified membrane [127]. The decade before, Lin et al. used a similar approach to prepare a CuY-zeolite@ polyethylene glycol (PEG) MMM [128], and also Cao et al. to obtain a polydimethylsiloxane MMM enhanced with silver/silica core–shell microspheres, noting that the composite membrane displayed more desirable characteristics than the non-modified membrane [129]. A couple of years ago, MOFs have also been used for the enhancement of polymeric membranes. Copper benzene-1,3,5-tricarboxylate (Cu-BTC) was incorporated into PEG, producing a MMM with a permeation flux that was 100% higher than a pristine PEG membrane prepared under the same conditions [130]. MOF-505 was inserted into the same polymeric matrix by Shi et al., resulting in a 158% increase of permeation flux over the corresponding pristine membrane [131]. None of these examples used the oxidative catalytic potential that can have these membranes. In fact, a gap in the literature is found for the application of MMMs that can combine oxidative catalytic activity and suitable sulfur extraction. One of the two published work was presented by Vigolo et al. in 2016, reinforcing a poly(methylmetacrylate) (PMMA) membrane by crosslinking with ­different molar ratios of Zr or Hf oxoclusters [132]. The different prepared MMMs were tested for the oxidation of DBT in a biphasic n-octane/acetonitrile system, with H2O2 as the oxidant. The best performance was obtained using a 1:50 Hf4:polymer monomer ratio, reducing sulfur content from 300 to 25 ppm in 24 hours, a 16% higher than the obtained using the solo oxocluster as homogeneous catalyst. This increase in activity was noted for all seven prepared MMMs, highlighting the significant role

777

778

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

of the polymeric matrix in shaping a suitable catalytic environment. More recently, a second report on the preparation of oxidative catalytic MMM for desulfurization was published by Mirante et al. [133]. The researchers reported on the development and application of a membrane-supported layered coordination polymer as an advanced sustainable catalyst for desulfurization by the immobilization of a previously reported active powdered coordination polymer catalyst [45a] in the PMMA matrix [133]. This membrane was then tested as an oxidative catalyst membrane for desulfurization of a model diesel containing four SCCs (1-BT, DBT, 4-MDBT and 4,6-DMDBT), with a BMIM based IL as extraction solvent and H2O2 as the oxidant, achieving complete desulfurization after four hours of reaction [133]. The powdered coordination polymer achieved a similar result after just two hours of reaction; however, the significantly enhanced handling of the membrane catalyst, added to its ease recovery and reusing (six ECODS cycles with no loss of activity), can open a new generation of advanced catalysts with viability to the industrial application of ODS processes.

35.6  Future Perspectives The need of sustainable alternatives for clean energy production should be one of the drivers of scientific innovation for the next century. However, energy transition to clean energy future is a long-term process. In opposite, the Organization of the Petroleum Exporting Countries predicts a global economic growth doubling from today to 2040, as the number of people on the planet expands by 1.7 billion [1]. Significantly for the oil sector, which is transportation driven, another 1.2 billion people will be behind the wheel of an automobile. Commercial vehicles on the road will double, while air travel will soar. Even with a massive effort to decrease the usage of fossil fuels (mainly in occidental countries), it is reported that the use of fossil fuels could still represent 60% of primary energy consumption by 2040. Therefore, the combustion of fossil fuels still represents an enormous fraction of the planet’s energy production. According to this scenario, the investment of developing technologies to reduce the harmful environmental impacts of fossil fuels is a priority. Desulfurization processes focus on limiting the eventual consequences of the release of sulfur into the atmosphere. The enhancement of the current industrial desulfurization process (HDS) by its combination with more efficient and cost-effective processes is crucial to provide clean fuel oil. Throughout this chapter, ADS and ODS, and the tandem use of ODS and EDS through (EC)ODS, were presented as potential sustainable alternatives/complementary processes to ensure deep desulfurization; however, their potential for industrial application is still in a research and development stage for real transportation fuels. Future advances in the actual industrial desulfurization technology will pass by strategically combining the well stablished HDS with other technology that is highly efficient for desulfurizing heavy molecular aromatic sulfur compounds. This latter will principally rely on materials with capacity to combine catalytic oxidation and extraction, resulting from its adsorptive or selective porosity. A promising class of materials able to incorporate all these functionalities are the membranes. Active and robust MOF or silica bulk porous materials that integrate oxidative catalytic and adsorptive capacities need to be incorporated in membranes containing selective permeability to heavy molecular aromatic sulfur compounds present in fuels. The combination of these desulfurization technologies integrating suitable materials will open the possibility to desulfurize a large number of liquid fuels, including HFO, saving time, costs, and material loss and deactivation.

References

Acknowledgments The work developed in our research group received financial support from Portuguese national funds (FCT/MCTES, Fundação para a Ciência e a Tecnologia and Ministério da Ciência, Tecnologia e Ensino Superior) through the strategic project UIDB/50006/2020 (for LAQV / REQUIMTE). Carlos Granadeiro, Luís Cunha-Silva and Salete S. Balula thank FCT/MCTES for funding through the Individual Call to Scientific Employment Stimulus (Ref. 2022.02651.CEECIND/CP1724/ CT001, CEECIND/00793/2018 and Ref. CEECIND/03877/2018, respectively). Alexandre Viana and Rui Faria thank FCT/MCTES and ESF (European Social Fund) through POCH (Programa Operacional Capital Humano) for their PhD grants (Refs. SFRH/BD/150659/2020 and UI/ BD/151277/2021, respectively).

References 1 Johnsson, F., Kjärstad, J., and Rootzén, J. (2019). Clim. Policy 19: 258–274. 2 Lelieveld, J., Klingmüller, K., Pozzer, A. et al. (2019). Proc. Natl. Acad. Sci. 116: 7192. 3 Sikarwar, P., Gosu, V., and Subbaramaiah, V. (2019). Rev. Chem. Eng. 35: 669–705. 4 Sefoka, R.E. and Mulopo, J. (2017). Int. J. Ind. Chem. 8: 373–381. 5 (2020). Oil Energy Trends 45: 3–5. 6 Singh, D., Chopra, A., Mahendra, P.K. et al. (2016). Pet. Sci. Technol. 34: 1248–1254. 7 Fahim, M.A., Alsahhaf, T.A., and Elkilani, A. (2010). Fundamentals of Petroleum Refining (ed. M.A. Fahim, T.A. Alsahhaf, and A. Elkilani), 153–198. Amsterdam: Elsevier. 8 (a) Grange, P. (1980). Catal. Rev. 21: 135–181; (b) Babich, I.V. and Moulijn, J.A. (2003). Fuel 82 : 607–631; (c) Shafiq, I., Shafique, S., Akhter, P. et al. (2022). Catal. Rev. 64: 1–86. 9 Schuman, S.C. and Shalit, H. (1971). Catal. Rev. 4: 245–318. 10 a. Duayne Whitehurst, D., Isoda, T., and Mochida, I. (1998). In Advances in Catalysis, 42. (eds. D.D. Eley, W.O. Haag, B. Gates, and H. Knözinger), Academic Press, pp. 345–471.; b. Song, C. and Ma, X. (2003). Appl. Cataly. B Environ. 41: 207–238. 11 Hansen, L.P., Ramasse, Q.M., Kisielowski, C. et al. (2011). Angew. Chem. Int. Ed. 50: 10153–10156. 12 Zhu, Y., Ramasse, Q.M., Brorson, M. et al. (2014). Angew. Chem. Int. Ed. 53: 10723–10727. 13 Salazar, N., Rangarajan, S., Rodríguez-Fernández, J. et al. (2020). Nat. Commun. 11: 4369. 14 Ma, X., Sakanishi, K., and Mochida, I. (1996). Ind. Eng. Chem. Res. 35: 2487–2494. 15 Knudsen, K.G., Cooper, B.H., and Topsøe, H. (1999). Appl. Catal. A: Gen. 189: 205–215. 16 Song, C. (2003). Catal. Today 86: 211–263. 17 McKee, R.H., Reitman, F., Schreiner, C. et al. (2013). Int. J. Toxicol. 33: 95S–109S. 18 Rana, M.S., Sámano, V., Ancheyta, J., and Diaz, J.A.I. (2007). Fuel 86: 1216–1231. 19 Anthony, E.J., Talbot, R.E., Jia, L., and Granatstein, D.L. (2000). Energy Fuels 14: 1021–1027. 20 Dias da Silva, P., Samaniego Andrade, S.K., Zygourakis, K., and Wong, M.S. (2019). Ind. Eng. Chem. Res. 58: 19623–19632. 21 Linares, N., Silvestre-Albero, A.M., Serrano, E. et al. (2014). Chem. Soc. Rev. 43: 7681–7717. 22 Saleh, T.A., Sulaiman, K.O., Al-Hammadi, S.A. et al. (2017). J. Clean. Prod. 154: 401–412. 23 Dinadayalane, T.C. and Leszczynski, J. (2010). Struct. Chem. 21: 1155–1169. 24 Saleem, J., Shahid, U.B., Hijab, M. et al. (2019). Biomass Convers. Biorefin. 9: 775–802. 25 Sevilla, M. and Mokaya, R. (2014). Energy Environ. Sci. 7: 1250–1280.

779

780

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

26 Salem, A.B.S.H. and Hamid, H.S. (1997). Chem. Eng. Technol. 20: 342–347. 27 Shi, Y., Liu, G., Wang, L., and Zhang, X. (2015). Chem. Eng. J. 259: 771–778. 28 Ania, C.O. and Bandosz, T.J. (2005). Langmuir 21: 7752–7759. 29 a. Moreira, A.M., Brandão, H.L., Hackbarth, F.V. et al. (2017). Chem. Eng. Sci. 172: 23–31.; b. Yu, C., Qiu, J. S., Sun, Y. F. et al. (2008). J. Porous Mater. 15: 151–157. 30 Jung, B.K. and Jhung, S.H. (2015). Fuel 145: 249–255. 31 Deng, L., Lu, B., Li, J. et al. (2017). Fuel 200: 54–61. 32 Saleh, T.A. (2018). J. Clean. Prod. 172: 2123–2132. 33 a. Novoselov, K.S., Fal′ko, V.I., Colombo, L. et al. (2012). Nature 490: 192–200. b. Chen, D., Feng, H., and Li, J. (2012). Chem. Rev. 112: 6027–6053. 34 Song, H.S., Ko, C.H., Ahn, W. et al. (2012). Ind. Eng. Chem. Res. 51: 10259–10264. 35 Menzel, R., Iruretagoyena, D., Wang, Y. et al. (2016). Fuel 181: 531–536. 36 Crespo, D. and Yang, R.T. (2006). Ind. Eng. Chem. Res. 45: 5524–5530. 37 Khaled, M. (2015). Res. Chem. Intermed. 41: 9817–9833. 38 Jiang, K., Li, Z., Zheng, Z. et al. (2021). Environ. Sci. Atmos. 1: 569–576. 39 Zhao, X. (2010). Materials for Energy Efficiency and Thermal Comfort in Buildings (ed. M.R. Hall), 399–426. Sawston: Woodhead Publishing. 40 Yang Ralph, T., Hernández-Maldonado Arturo, J., and Yang Frances, H. (2003). Science 301: 79–81. 41 Zhang, Z.Y., Shi, T.B., Jia, C.Z. et al. (2008). Appl. Catal. B: Environ. 82: 1–10. 42 Neubauer, R., Husmann, M., Weinlaender, C. et al. (2017). Chem. Eng. J. 309: 840–849. 43 Tian, F., Yang, X., Shi, Y. et al. (2012). J. Nat. Gas Chem. 21: 647–652. 44 Zhou, Y., Dong, Z., Terasaki, O., and Ma, Y. (2022). Acc. Mater. Res. 3: 110–121. 45 a Mendes, R.F., Antunes, M.M., Silva, P. et al. (2016). Chem. Eur. J. 22: 13136–13146; b Narayan, R., Nayak, U.Y., Raichur, A.M., and Garg, S. (2018). Pharmaceutics 10: 118. 46 Alvarado-Perea, L., Colín-Luna, J.A., López-Gaona, A. et al. (2020). Catal. Today 353: 26–38. 47 Song, L., Chen, J., Bian, Y. et al. (2015). J. Porous Mater. 22: 379–385. 48 Palomino, J.M., Tran, D.T., Hauser, J.L. et al. (2014). J. Mater. Chem. A 2: 14890–14895. 49 Tian, W.H., Sun, L.B., Song, X.L. et al. (2010). Langmuir 26: 17398–17404. 50 Zhou, H.-C., Long, J.R., and Yagh, O.M. (2012). Chem. Rev. 112: 673–674. 51 Cychosz, K.A., Wong-Foy, A.G., and Matzger, A.J. (2008). J. Am. Chem. Soc. 130: 6938–6939. 52 Li, Y.-X., Jiang, W.-J., Tan, P. et al. (2015). J. Phys. Chem. C 119: 21969–21977. 53 Zhang, X.-F., Wang, Z., Feng, Y. et al. (2018). Fuel 234: 256–262. 54 Khan, N.A., Hasan, Z., and Jhung, S.H. (2014). Chem. Eur. J. 20: 376–380. 55 Otsuki, S., Nonaka, T., Takashima, N. et al. (2000). Energy Fuels 14: 1232–1239. 56 Zhang, M., Zhu, W., Xun, S. et al. (2013). Chem. Eng. J. 220: 328–336. 57 Zhao, H. and Baker, G.A. (2015). Front. Chem. Sci. Eng. 9: 262–279. 58 Guth, A.D.E., KVB Inc. (1974). United States Patent 1191. 1451 Nov. 12, Method for removing sulfur and nitrogen in petroleum oils. 59 Xu, X., Moulijn, J.A., Ito, E. et al. (2008). ChemSusChem. 1: 817–819. 60 Sundararaman, R., Ma, X., and Song, C. (2010). Ind. Eng. Chem. Res. 49: 5561–5568. 61 Ma, C., Dai, B., Liu, P. et al. (2014). J. Ind. Eng. Chem. 20: 2769–2774. 62 Wang, J., Zhao, D., and Li, K. (2010). Energy Fuels 24: 2527–2529. 63 Kim, J. and Huang, C.-H. (2021). ACS ES&T Water 1: 15–33. 64 Dehkordi, A.M., Sobati, M.A., and Nazem, M.A. (2009). Chin. J. Chem. Eng. 17: 869–874. 65 Akopyan, A., Eseva, E., Polikarpova, P. et al. (2020). Molecules (Basel, Switzerland) 25: 536. 66 Shiraishi, Y., Hara, H., Hirai, T., and Komasawa, I. (1999). Ind. Eng. Chem. Res. 38: 1589–1595.

References

67 Isaacs, M.A., Robinson, N., Barbero, B. et al. (2019). J. Mater. Chem. A 7: 11814–11825. 68 (a) Napanang, T. and Sooknoi, T. (2009). Catal. Commun. 11: 1–6; (b) Shen, C., Wang, Y.J., Xu, J.H., and Luo, G.S. (2015). Chem. Eng. J. 259: 552–561. 69 Lv, Q., Li, G., and Sun, H. (2014). Fuel 130: 70–75. 70 a Ding, Y., Ke, Q., Liu, T. et al. (2014). Ind. Eng. Chem. Res. 53: 13903–13909; b Li, L., Wang, W., Huang, J. et al. (2022). Appl. Cataly. A General 630: 118466. 71 Du, S., Chen, H.-M., Shen, H.-X. et al. (2020). ACS Appl. Nano Mater. 3: 9393–9400. 72 Kong, L., Li, G., and Wang, X. (2004). Catal. Today 93-95: 341–345. 73 Du, S., Li, F., Sun, Q. et al. (2016). Chem. Commun. 52: 3368–3371. 74 Hall, J.N. and Bollini, P. (2019). React. Chem. Eng. 4: 207–222. 75 Llabrés I Xamena, F.X., Luz, I., and Cirujano, F.G. (2013). Metal Organic Frameworks as Heterogeneous Catalysts. 237–267. The Royal Society of Chemistry. 76 a McNamara, N.D., Neumann, G.T., Masko, E.T. et al. (2013). J. Catal. 305: 217–226; b Li, X., Gu, Y., Chu, H. et al. (2019). Appl. Cataly. A General 584: 117152. 77 Hwang, Y.K., Hong, D.-Y., Chang, J.-S. et al. (2009). Appl. Catal. A: Gen. 358: 249–253. 78 Masoomi, M.Y., Bagheri, M., and Morsali, A. (2015). Inorg. Chem. 54: 11269–11275. 79 Granadeiro, C.M., Ribeiro, S.O., Karmaoui, M. et al. (2015). Chem. Commun. 51: 13818–13821. 80 Viana, A.M., Ribeiro, S.O., Castro, B. et al. (2019). Materials (Basel) 12: 3009. 81 Viana, A.M., Julião, D., Mirante, F. et al. (2021). Catal. Today 362: 28–34. 82 Dissegna, S., Epp, K., Heinz, W.R. et al. (2018). Adv. Mater. 30: 1704501. 83 Gu, Y., Xu, W., and Sun, Y. (2021). Catal. Today 377: 213–220. 84 (a) Gu, Y., Ye, G., Xu, W. et al. (2020). ChemistrySelect 5: 244–251; (b) Fu, G., Bueken, B., and De Vos, D. (2018). Small Methods 2: 1800203. 85 Zheng, H.-Q., Zeng, Y.-N., Chen, J. et al. (2019). Inorg. Chem. 58: 6983–6992. 86 Li, N., Zhang, Z.-W., Zhang, J.-N. et al. (2021). Dalton Trans. 50: 6506–6511. 87 Smolders, S., Willhammar, T., Krajnc, A. et al. (2019). Angew. Chem. Int. Ed. Engl. 58: 9160–9165. 88 Haw, K.-G., Bakar, W.A.W.A., Ali, R. et al. (2010). Fuel Process. Technol. 91: 1105–1112. 89 Kampouraki, Z.C., Giannakoudakis, D.A., Triantafyllidis, K.S., and Deliyanni, E.A. (2019). Green Chem. 21: 6685–6698. 90 Gu, Q., Wen, G., Ding, Y. et al. (2017). Green Chem. 19: 1175–1181. 91 Shen, K., Chen, X., Chen, J., and Li, Y. (2016). ACS Catal. 6: 5887–5903. 92 Yap, M.H., Fow, K.L., and Chen, G.Z. (2017). Green Energy Environ. 2: 218–245. 93 (a) Marpaung, F., Kim, M., Khan, J.H. et al. (2019) Chem. Asian J. 14: 1331–1343; (b) Chen, Y.-Z., Zhang, R., Jiao, L., and Jiang, H.-L. (2018). Coordinat. Chem. Rev. 362: 1–23. 94 Kim, J., McNamara, N.D., Her, T.H., and Hicks, J.C. (2013). ACS Appl. Mater. Interfaces 5: 11479–11487. 95 Kim, J., McNamara, N.D., and Hicks, J.C. (2016). Appl. Catal. A: Gen. 517: 141–150. 96 Sarker, M., Bhadra, B.N., Shin, S., and Jhung, S.H. (2019). ACS Appl. Nano Mater. 2: 191–201. 97 Bhadra, B.N. and Jhung, S.H. (2018). Nanoscale 10 (31): 15035–15047. 98 Bhadra, B.N., Khan, N.A., and Jhung, S.H. (2019). J. Mater. Chem. A 7: 17823–17833. 99 Capel-Sanchez, M.C., Campos-Martin, J.M., and Fierro, J.L.G. (2010). Energy Environ. Sci. 3: 328–333. 100 Rivoira, L.P., Vallés, V.A., Ledesma, B.C. et al. (2016). Catal. Today 271: 102–113. 101 (a) Cui, S., Ma, F., and Wang, Y. (2007) React. Kinet. Catal. Lett. 92: 155–163; (b) Kim, T.-W., Kim, M.-J., Kleitz, F. et al. (2012). ChemCatChem. 4: 687–697; (c) Cho, K.-S. and Lee, Y.-K. (2014). Appl. Cataly. B Environ. 147: 35–42; (d) Shah, A.T., Li, B., and Ali Abdalla, Z.E. (2009). J. Colloid

781

782

35  Heterogeneous Catalytic Desulfurization of Liquid Fuels

Interface Sci. 336: 707–711; (e) Chica, A., Corma, A., & Dómine, M.E. (2006). J. Catalysis 242: 299–308. 102 Polikarpova, P., Akopyan, A., Shlenova, A., and Anisimov, A. (2020). Catal. Commun. 146: 106123. 103 (a) Kasuga, T., Hiramatsu, M., Hoson, A. et al. (1998). Langmuir 14: 3160–3163; (b) Bavykin, D.V., Friedrich, J.M., and Walsh, F.C. (2006). Adv. Mater. 18: 2807–2824. 104 Lorençon, E., Alves, D.C.B., Krambrock, K. et al. (2014). Fuel 132: 53–61. 105 Lu, S.-X., Zhong, H., Mo, D.-M. et al. (2017). Green Chem. 19: 1371–1377. 106 Cedeño-Caero, L., Ramos-Luna, M., Méndez-Cruz, M., and Ramírez-Solís, J. (2011). Catal. Today 172: 189–194. 107 Rao, C., Müller, A., and Cheetham, A. (2004). The Chemistry of Nanomaterials: Synthesis, Properties and Applications, 2. (ed. C.N.R. Rao, A. Müller, and A.K. Cheetham), 761. Wiley-VCH. ISBN 3-527-30686-2 March 2004, -1. 108 Trakarnpruk, W. and Rujiraworawut, K. (2009). Fuel Process. Technol. 90: 411–414. 109 Te, M., Fairbridge, C., and Ring, Z. (2001). Appl. Catal. A: Gen. 219: 267–280. 110 Ivanova, S. (2014). ISRN Chem. Eng. 2014: 963792. 111 Collins, F.M., Lucy, A.R., and Sharp, C. (1997). J. Mol. Catal. A Chem. 117: 397–403. 112 (a) Wang, R., Zhang, G., and Zhao, H. (2010). Catal. Today 149: 117–121; (b) Misra, A., Kozma, K., Streb, C., and Nyman, M. (2020). Angewandte Chemie Int. Ed. 59: 596–612. 113 Ghubayra, R., Yahya, R., Kozhevnikova, E.F., and Kozhevnikov, I.V. (2021). Fuel 301: 121083. 114 Khodadadi Dizaji, A., Mortaheb, H.R., and Mokhtarani, B. (2019). Catal. Lett. 149: 259–271. 115 Wang, R., Yu, F., Zhang, G., and Zhao, H. (2010). Catal. Today 150: 37–41. 116 Ribeiro, S., Barbosa, A.D.S., Gomes, A.C. et al. (2013). Fuel Process. Technol. 116: 350–357. 117 Wang, X.S., Huang, Y.B., Lin, Z.J., and Cao, R. (2014). Dalton Trans. 43: 11950–11958. 118 Wang, X.-S., Li, L., Liang, J. et al. (2017). ChemCatChem. 9: 971–979. 119 Wang, H., Jibrin, I., and Zeng, X. (2020). Front. Chem. Sci. Eng. 14: 546–560. 120 Zhang, Y., Zhang, W., Zhang, J. et al. (2018). RSC Adv. 8: 31979–31983. 121 Wang, H. and Wang, R. (2011). Collect. Czechoslov. Chem. Commun. 76: 1595–1605. 122 Julião, D., Mirante, F., Ribeiro, S.O. et al. (2019). Fuel 241: 616–624. 123 Ribeiro, S.O., Granadeiro, C.M., Almeida, P.L. et al. (2019). Catal. Today 333: 226–236. 124 Ribeiro, S.O., Granadeiro, C.M., Corvo, M.C. et al. (2019). Front. Chem. 7: 756. 125 Argyle, M.D. and Bartholomew, C.H. (2015). Catalysts 5: 949–954. 126 Fihri, A., Mahfouz, R., Shahrani, A. et al. (2016). Chem. Eng. Process.- Process Intensif. 107: 94–105. 127 Peng, P., Lan, Y., Zhang, Q., and Luo, J. (2022). J. Appl. Polym. Sci. 139: 51514. 128 Lin, L., Zhang, Y., and Li, H. (2010). J. Colloid. Interface Sci. 350: 355–360. 129 Cao, R., Zhang, X., Wu, H. et al. (2011). J. Hazard. Mater. 187: 324–332. 130 Cai, C., Fan, X., Han, X. et al. (2020). Polymers 12: 414. 131 Shi, W., Han, X., Bai, F. et al. (2021). Sep. Purif. Technol. 272: 118924. 132 Vigolo, M., Borsacchi, S., Sorarù, A. et al. (2016). Appl. Catal. B: Environ. 182: 636–644. 133 Mirante, F., Mendes, R.F., Faria, R.G. et al. (2021). Molecules 26: 2404.

783

Part VII Hydrogen Formation, Storage, and Utilization

785

36 Paraformaldehyde Opportunities as a C1-Building Block and H2 Source for Sustainable Organic Synthesis Ana Maria Faísca Phillips1, Maximilian N. Kopylovich1, Leandro Helgueira de Andrade2, and Martin H.G. Prechtl1,3 1

Centro de Química Estrutural, Institute of Molecular Sciences, Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal Departamento de Química Fundamental, Instituto de Química, Universidade de São Paulo, São Paulo, Brazil 3 Department of Synthesis and Analysis, Albert Hofmann Institute for Physiochemical Sustainability, Vlotho, Germany 2

36.1  Introduction In recent years, environmental concerns related to the production of fine chemicals has become a major issue, because hazardous reagents are commonly applied in many industrial processes and this results in large amounts of polluted waste and requires expensive reaction setups for safe processing [1–3]. A modernization of traditional reaction schemes considers the use of catalysts for higher selectivitis, shorter reaction times, and lower reaction temperatures combined with the use of less toxic and renewable solvents, as well as reagents which can be derived from renewable sources. Another consideration is related to the need to minimize the risk related to the use and handling of hydrogen gas and CO under elevated pressure and/or at high concentrations. To overcome these issues for synthetic setups, liquid/solid H2 and CO sources are attractive alternatives [1–6]. Among these are small molecules that can be derived from CH4 and CO2 with a relatively high hydrogen content (in wt-%), such as C1-molecules like methanol (12.5 wt-%), formaldehyde (6.7 wt-%), methanediol (8.4 wt-%), or formic acid (4.3 wt-%), which are catalytically convertible to H2 and CO2, H2 and CO or CO, and H2O (Figure 36.1) [1, 4, 7]. One of these C1-molecules, formaldehyde, is known to form oligomers and polymers of different chain lengths in the absence of water and paraformaldehyde (PFA) can act as a source of formaldehyde and methanediol, respectively (Figure 36.1) [1–3, 5, 8–10]. Formaldehyde can be synthesized by the oxidation/dehydrogenation of methanol via the Formox process at elevated temperature, applying a transition metal catalyst [3, 11, 12], or by the conversion of syngas (CO:H2 = 1:2) in the aqueous phase with metal catalysts [3, 11, 13]. PFA [(CH2O)n], the corresponding colorless oligomer/polymer solid powder (n = 8–100), is formed by dehydration of aqueous formaldehyde and, respectively, methanediol. In this chapter, we give a brief overview about selected examples of homogeneous metal-catalyzed reactions with PFA which can be used as a source of hydrogen gas and C1-building blocks in Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 3, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

786

36  Paraformaldehyde

addition to its potential use as a CO surrogate, depending on the reaction conditions. In particular, PFA in water can be used as source of H2 for transfer hydrogenation reactions such as for the reduction of C=C, C=O, –CC–, and –CN bonds (Figure 36.2a) [9, 10, 14, 15]. Underlining that water acts as an H2 source as well because it reacts with PFA to yeild methanediol, which is subsequently dehydrogenated to yield two eq. of H2 and one eq. of CO2 (Figure 36.3) [1, 8]. PFA itself is produced from methanol, as by-product of formaldehyde [3, 11, 12], which can be produced from CO2 [16–20], and therefore PFA could become a renewable molecule in the future as well. Moreover, PFA is a convenient CO-surrogate and C1 building block; therefore by the use of appropriate metal complex catalysts, it can be applied in hydroformylation [6], N-formylation [5], carbonylation [21], methylation [2], hydroxymethylation [22], or C–H activation [22–24], among many other reactions which will be outlined further in this chapter (Figure 36.2b). Figure 36.1  Interconversion of C1-molecules by shuttling hydrogen and oxygen in the presence of catalysts.

Figure 36.2  a) Metal-catalyzed transfer-hydrogenation of multiple-bonds with paraformaldehyde (PFA) as an H2 source [9, 10, 14, 15]. b) Metal-catalyzed N-formylation and (hydroxyl)methylation [2, 5, 6, 22].

36.2  Carbonylation and Related Reactions

Figure 36.3  Transition metal catalyzed hydrogen evolution from paraformaldehyde (PFA) via methanediol [1, 3, 12, 25, 26].

Gaseous reagents such as dihydrogen, syngas, and CO require the use of more sophisticated and expensive laboratory infrastructure to handle volatile, hazardous, explosive, and/or toxic gases. In contrast, the solid form and simple use of PFA shows advantages for application in research labs [2, 5, 9, 10]. The adjustment of the required quantity of the H2 carrier, respectively C1-surrogate and water, can be easily conducted gravimetrically and these reactions can be run in simple glassware suitable for low pressure reactions (e.g. headspace vials). Expensive stainless steel pressure reactors are not required. In addition, in several cases air and moisture stable catalysts and reagents are used, thus expensive and inert conditions, and likewise glovebox or Schlenk techniques, are not necessary. The robustness of the catalysts, the use of more environmentally benign solvents, and PFA and water as hydrogen carriers, give access to more sustainable synthetic methodologies.

36.2  Carbonylation and Related Reactions The incorporation of a C=O unit into a molecule, known as carbonylation, provides access to a variety of useful carbonyl-containing compounds, including aldehydes, esters, acids, or amides [27]. These products may themselves be subsequently transformed into a whole range of fine chemicals, due to the high synthetic versatility of the carbonyl function. Carbon monoxide (CO) gas is the cheapest source of this C1 unit and carbonylation reactions are amongst the largest applications of homogeneous catalysis in industry [27]. An example is the synthesis of propanoic acid, produced on a multi-ton scale by the hydroxycarbonylation of ethene, as an intermediate for the production of plastics, pharmaceuticals, and pesticides [28]. Methyl methacrylate (MMA), produced by the palladium-catalyzed methoxycarbonylation of ethene (the Lucite process), followed by reaction of the methyl propanoate obtained with formaldehyde and elimination, is another example. Methyl methacrylate is used to prepare poly(methyl methacrylate), a transparent and strong polymer with many industrial applications. The carbonylation of alkenes is also highly important in the pharmaceutical industry [6, 29] and it can even be used to convert the very inert alkanes from the petrochemical and natural gas industries into value-added products [6, 30]. Although CO is usually used for carbonylation, it is flammable and highly toxic and therefore many attempts are underway to find surrogates that are easier to handle and safer, without compromising the efficiency of the reactions, particularly (but not only) for laboratory-scale applications [28]. A few organic and inorganic chemical compounds have been utilized to perform this role, although these compounds have other drawbacks, related either to their atom efficiency, reactivity, toxicity, or price. For example, metal carbonyls such as Mo(CO)6 may require stoichiometric amounts of other transition metals and harsh conditions for the release of CO. Organic reagents include pivaloyl chloride, silicon-based carboxylic acids, N-formylsaccharin, formic acid, and formamides such as dimethylformamide (DMF). Waste generation, harsh reaction conditions, poor selectivity in the presence of other nucleophiles, low atom efficiency, and even the need for external CO pressure in some cases, can be some of their disadvantages, which make the search for other CO sources a subject of continued interest [28]. Formaldehyde is the most economic

787

788

36  Paraformaldehyde

surrogate; it has suitable reactivity and the added advantage that the percentage by molecular weight of the CO unit is 93%, the highest of all surrogates used so far. High yields and high levels of selectivity have been obtained with its solid form, paraformaldehyde, in the alkoxycarbonylation of alkenes, in N-carbonylation, in the carbonylation of aryl halides, and even in cascade reactions, to name a few, as described next.

36.2.1  Alkoxycarbonylation of Olefins The alkoxycarbonylation of olefins allows the synthesis of carboxylic acid esters from readily available olefins, alcohols, and CO [31, 32]. The main issues to overcome are i) the regioselectivity (i.e. whether linear or branched products are formed), ii) the activity of the catalyst, and iii) the rate of depolymerization of PFA. At present, a number of catalysts are available to promote this transformation, based on Ru or Pd, and the nature of the ligand used has a strong impact on activity and selectivity. An early example, in which PFA was used as the CO source [33], relied on the utilization of triruthenium dodecacarbonyl [Ru3(CO)12] as catalyst, and an electron-rich monodentate phosphine ligand, PCy3 (L1, Cy=cyclohexyl), in order for high yields to be obtained. The utilization of 1-butyl-3-methylimidazolium chloride (BMIMCl) as an additive helped to improve conversion and selectivity. In the Lucite process for the synthesis of MMA, CO gas is used, and methane sulfonic acid is required as an additive to ensure high catalyst activity. The new PFAbased method by-passed the need for high pressure reactors, as well as that of acid, which may cause reactor vessel corrosion problems. As the examples in Table 36.1 show, the Ru/L1 system afforded carboxylate esters from a variety of terminal and internal alkenes. Methyl propionate, the Table 36.1  The alkoxycarbonylation of olefins.

No

Alkene

Major product

Metal/ ligand1

1

[Ru]/L12 [Pd]/L2

2 3

Yield [%] (linear: branched selectivity)

Ref

51 99

[33] [28]

[Ru]/L1 [Pd]/L2

51 (70:30) 90 (78:22)

[33] [28]

[Ru]/L1 [Pd]/L3

89 (> 99:1) 97 (> 99:1)

[33] [28]

36.2  Carbonylation and Related Reactions

Table 36.1  (Continued) 4

[Ru]/L1 [Pd]/L2 [Pd]/L3

68 43 (>99:1) 92 (>99:1)

[33] [28] [34]

5

[Ru]/L1 [Pd]/L2

43 89 (>99:1)

[33] [28]

6

[Ru]/L1 [Pd]/L3

53 89 (> 99:1)

[33] [34]

7

[Ru]/L1 [Pd]/L3

88 94

[33] [34]

8

[Ru]/L1 [Pd]/L3

90 91

[33] [34]

9

[Ru]/L1

82

[28]

10

[Ru]/L1

74 (51:49) 93 (95:5)

[33] [28]

11

[Pd]/L2

93 (94:6)

[28]

12

[Pd]/L2

75 (89:11)

[28]

13

[Pd]/L2

85 (>99:1)

[28]

14

[Pd]/L2

92 (>99:1)

[28]

15

[Pd]/L2

89 (>99:1)

[28]

16

[Pd]/L3

93 (>99:1)

[34]

17

[Pd]/L3

42 (89:11)

[34]

18

[Pd]/L3

93 (>99:1)

[34]

19

[Pd]/L3

93 (>99:1)

[34]

20

[Pd]/L3

42

[34]

1)  Conditions: [Ru]/L1: [Ru3(CO)12] (1.5 mol%), PCy3 (4.5 mol%), NMP, BMIMCl (2 equiv), 130 °C; [Pd]/L2: [Pd(OAc)2] (1 mol%), dtbpx (4 mol%), PTSA (5 mol%), 100 °C; [Pd]/L3: 1.0 mmol substrate, 1.0 mol% Pd(OAc)2, 4.0 mol% L1, 5.0 mol% PTSA·H2O, 200 mg (CH2O)n, 2.0 mL MeOH, 120 °C, 72 h; NMP = N-methylpyrrolidone. 2)  [Ru3(CO)12] 0.7 mol%.

789

790

36  Paraformaldehyde

intermediate in the synthesis of methyl methacrylate, could be obtained in 51% yield from ethylene, with an even lower catalyst loading of 0.7 mol%. With styrene derivatives, terminal aliphatic esters were obtained (entries 3 and 4), and the use of terminal and internal aliphatic alkenes produced a mixture of branched and linear esters with almost the same yields and selectivities (e.g. entry 10). The major side reaction observed was alkene hydrogenation, but, in the reactions of aliphatic alkenes, hydrogenated and isomerized byproducts (n-octane and iso-octenes) were also observed. In this study, other alcohols (e.g. ethanol, butanol, benzyl alcohol) were also reacted with cyclohexene, and the corresponding esters were obtained in moderate to excellent yields, which decreased as the length of the carbon chain increased. An isotope-labeling experiment with 13C-labeled methanol and cyclohexene showed that paraformaldehyde was the single source of CO in the reaction, because 13C only appeared in the methyl ester portion of the product, not in the CO group. In addition, palladium catalysis appeared to be suitable for this type of conversion [28], and the use of palladium acetate in combination with α,α′-bis(di-tertbutylphosphino)-o-xylene (L2), a more hindered bidentate diphosphine ligand, provided very high linear selectivities (up to >99:1). In general, linear selective (anti-Markovnikov) reactions are preferred, because linear carboxylate esters are favored by industry. Terminal aliphatic olefins yielded esters in very high yields and n-selectivity (e.g. entry 10, Table 36.1). Aliphatic internal olefins were converted into the corresponding linear esters with high yields, by double bond isomerization followed by alkoxycarbonylation (entries 11 and 12). This isomerization is a useful feature of this process, because it allows the use of mixtures of double bond isomers as starting materials to obtain a single product, which can be of interest for industrial applications because olefin mixtures are usually less expensive than pure isomers. Fatty acids could also be functionalized to yield linear diesters (entries 14 and 15). This palladium-catalyzed process required an acid additive (p-toluene sulfonic acid [PTSA] or methanesulfonic acid [MSA]) for high yields to be obtained. With other acids (e.g. H2SO4, acetic acid [AcOH], trifluoroacetic acid [TFA]), there was no product formation. In this reaction, 13C isotope labeling experiments showed different results from those obtained with the Ru-catalyzed process. In this case, both 13C-labeled paraformaldehyde and the independent use of 13 CH3OH as the solvent produced a 13C-labeled methyl nonanoate, suggesting that both formaldehyde and methanol act as the carbonyl source in this process simultaneously. Despite the good results obtained with the Pd/L2 system, tri- and tetra-substituted olefins did not react under these conditions. However, an optimized system to convert these highly substituted olefins into the corresponding products could be realized [34]. The key to success was the development of the ligand L3, which incorporated pyridyl substituents. The success of this ligand is related to the ability of the amphoteric group to act as a proton shuttle to form the active palladium hydride, as well as for enabling an N-assisted alcoholysis step. The pyridyl substituents also helped to improve olefin isomerization. The desired methyl esters could be obtained with very good yields and selectivitites, as shown in Table 36.1, even when tetrasubstituted olefins were reacted (entries 19 and 20). (2-Pyridylmethylene)cyclobutanes were subjected to rhodium-catalyzed carbonylation with gaseous CO in xylene [35]. CO insertion in the presence of the Rh(I) catalyst proceeded via pyridinedirected C–C oxidative addition. The overall result is cyclobutane to cyclopentanone ring expansion. Good to high yields of products (55–95%) and selectivities in the range 50:50–26:74 were obtained with the internal double bonded products predominating over those with an exocyclic double bond.

36.2.2  Carbonylation of Aryl Halides The reaction of aryl halides with CO gas has been known for a while, but the utilization of paraformaldehyde as a CO surrogate has been much less explored [6]. N-(o-bromoaryl)amides were used

36.2  Carbonylation and Related Reactions

as substrates for the preparation of substituted benzoxazinones heterocycles (36–86% yield) by a simple palladium-catalyzed carbonylation with PFA (Figure 36.4a) [36]. The presence of electronwithdrawing substituents on the amidoaryl moiety had a detrimental effect on the yield. 4,5-Bis(diphenylphosphino)-9,9-dimethylxanthene (Xantphos) was found to be the best ligand for this process. An experiment with 13C-labeled paraformaldehyde yielded a 4-13C-labeled benzoxazinone derivative that has many important pharmaceutical and biological applications. In addition 2-bromobiphenyls could be transformed into a range of fluoren-9-one derivatives in good to high yields, via palladium catalysis (Figure 36.4b) [37]. The fluoren-9-one structure is found in many optoelectronically and biologically active compounds, as well as in key synthetic intermediates. In this process, there is not only aryl halide activation, but also the much more difficult to achieve direct C–H carbonylation by cleavage of a C–H bond. The presence of a base

Figure 36.4  The carbonylation of aryl halides. (a) Synthesis of benzoxazinones by Pd-catalyzed carbonylation with paraformaldehyde (PFA) [36]. (b) The synthesis of fluorinones [37]. (c) The reaction mechanism proposed for the synthesis of fluorinones [37].

791

792

36  Paraformaldehyde

(Na2CO3) and a dehydrating agent (MgSO4) are important in this process, the latter for suppressing the reductive hydrodebromination of the aryl halide. The mechanism postulated is shown in Figure 36.4c [37]. Oxidative addition of the aryl bromide to Pd(0) obtained from Pd(II) gives rise to complex A. CO formed by decarbonylation of formaldehyde under Pd(0) catalysis coordinates to A, then undergoes migratory insertion giving rise to acyl palladium species B. C–H bond cleavage can occur in one of two ways: either by electrophilic aromatic substitution or via concerted metallation-deprotonation to yield palladacycle C. Reductive elimination releases the product and Pd(0) for another catalytic cycle. The cyclocarbonylation of 2-bromo-1,1,2-triphenylethylene was possible under the same conditions, to afford 2,3-diphenylinden-1-one in 52% yield. Rhodium catalysis can be used for the aryloxycarbonylation of aryl iodides with PFA and phenols [38]. However, in this case, good yields are only obtained with phenyl iodide, although several substituents may be present in the phenolic reaction partner. The corresponding reaction performed with free CO was less sensitive to phenol substitution and good yields could be obtained (Table 36.2). The reaction conditions had to be modified for the paraformaldehyde reaction. The addition of ethyl acetate as co-solvent improved the solubility of the in situ formed formaldehyde, although the catalyst activity dropped substantially if too much was added. Side products were Table 36.2  The Rh-catalyzed aryloxycarbonylation of aryl iodides with PFA.1

Entry

R1

R2

Conditions

1

F

H

A

80

72

B

21

19

A

95

88

B

67

3

A

97

88

B

43

35

A

>99

94

B

>99

86

2 3

Cl

H

4 5

Me

H

6 7

H

F

8 9

H

iPr

10 H 1

CF3

Conv. [%]

Yield [%]

A

98

95

B

65

61

A

98

93

B

49

43

Acac = acetylacetone; xantphos = 4,5-Bis(diphenylphosphino)-9,9-dimethylxanthene;

36.2  Carbonylation and Related Reactions

were observed in all of the reactions to a small extent, resulting from reduction of the iodobenzene, which led to benzene formation or ester hydrolysis leading to carboxylic acid formation.

36.2.3  Cascade C–H Activation/carbonylation/cyclization Reactions and Related Processes: The Synthesis of Heterocycles The synthesis of quinolines and their derivatives has been recently achieved via cascade processes involving carbonylation with paraformaldehyde using rhodium catalysis [23]. In this strategy, unprotected anilines are reacted with electron-deficient alkynes to yield initially C–C bonded products instead of the more common C–N bonded species often used in the synthesis of these compounds. For example, anilines usually undergo Michael-type addition with electron-deficient alkynes, leading to C–N bonded adducts (Figure 36.5a) [23]. A C–N bond is also formed in

Figure 36.5  (a) Rh-catalyzed synthesis of quinolines and (b) the mechanism proposed;

793

794

36  Paraformaldehyde

TM-catalyzed hydroamination of alkynes [39]. This reversal in reactivity, in which there is TM-catalyzed C–H bond activation of unprotected or directing group-free amines, is very rare, and this appears to be the first example of non-directed catalytic C–H bond activation of unprotected primary anilines. The synthesis of these carboannulated products could also be achieved using pre-prepared ortho-vinylanilines, PFA, and the same rhodium catalyst (Figure 36.5a). In that case, 1 mol% of catalyst was enough for high yields to be obtained (up to 86% of isolated compound). 3-Methylaniline showed poor reactivity (6% product yield after 24 hours), which was attributed to a lower arene electron density: less η2-coordination and greater Lewis basic coordination to the metal center. The fact that water can be used as a solvent adds to the environmental friendliness of the procedure. Labelling experiments with 13C-labelled PFA showed unambiguously that CO originates from the PFA. A mechanism was proposed for the reaction (Figure 36.5b) [23]. Non-directed C–H activation by the rhodium catalyst results in C–C coupling after coordination of the alkyne to rhodium activated species A and migratory insertion. C, the product of these initial steps, reacts with CO obtained from PFA. CO insertion can lead to intermediate E via D. Protodemetalation of intermediate E and intramolecular imination affords the desired quinoline. The synthesis of quinolines from anilines has also been made possible by the reaction with ketones and PFA by Yi et al. using cobalt catalysis [40]. A C–H activation/carbonylation/cyclization reaction takes place to give products in good to very high yields (Figure 36.6). Exclusive siteor/and region-selectivity was also observed when meta-substituted anilines were reacted with unsymmetrical ketones, and a large range of substrates were compatible with the reaction conditions. As in the previous example, no directing groups were required. The only by-products in this atom and step-economical synthesis are H2O and H2. It was postulated that AgNTf2 has the dual role of activating the catalyst (i.e. from Cp*Co(CO)I2 to Cp*Co(CO)NTf2) and the ketone enolate, and that C–H bond activation probably occurs via an internal electrophilic substitution (IES)-type mechanism. This rules out the probability of there being ring closure through an electrophilic aromatic substitution (EAS) pathway. Palladium-catalyzed carbonylative C–H activation of arenes in the presence of norbornene has been used to obtain various 5-(pyridin-2-yl)-hexahydro-7,10-methanophenanthridin-6(5H)-ones in moderate yields (Figure 36.6b) [24]. Although molybdenum hexacarbonyl was the main source of CO, PFA could also be utilized under similar conditions to provide a product in moderate yield (Figure 36.6).

36.2.4  Hydroformylation of Alkenes The hydroformylation of alkenes with syn gas is widely known as an important industrial process [6, 41]. It normally requires a high pressure of H2/CO and a metal catalyst, of which Rh, Co or Ru are commonly used. The interest to find a replacement that can work under pressure free conditions, namely paraformaldehyde, has also attracted much interest. PFA can sometimes provide complementary regioselectivity to the gaseous hydroformylation reactions [6]. Hydroformylation performed with PFA was already reported in 1982 [42], where 0.5 mol% RhH2(O2COH)[P(i-Pr)3]2 catalyzed the conversion of olefins to aldehydes, with the later being obtained in low to good yields (11–67%) (Table 36.3). With α,β-unsaturated alkenes, C=C bond reduction products predominated. It was found that β-oxygen-substituted olefins (i.e. allyl alcohol and methyl acrylate) were more reactive and yielded higher n/iso ratio of products than nonoxygenated alkenes when RhH(CO)(PPh2)2/PPh3 was used as catalyst, although the highest yield obtained was only 23% [43]. Employing [Rh(CO)2(acac)]/2dppe as the catalytic system, obtained a modest n/iso ratio with 1-hexane (n/iso = 2, turnover number [TON] = 200), whereas a high

36.2  Carbonylation and Related Reactions

Figure 36.6  (a) Co-catalyzed synthesis of quinolines from anilines, ketones and PFA and the mechanism proposed [40]. (b) Pd-catalyzed carbonylative activation of arenes and norbonene [24].

iso/n ratio was observed for allyl alcohol (iso/n = 21, TON = 129) [44]. A highly efficient (up to 95%) and regioselective (l/b=up to 98/2) hydroformylation of 1-alkenes can be performed by using a combination of two ligands (2,2ʹ-bis(diphenylphosphino)-1,1ʹ-biphenyl [BIPHEP] and Nixantphos [4,6-bis(diphenylphosphino)phenoxazine]) for a Rh-catalyzed process that operated under mild conditions [45]. The yields obtained under similar conditions using a formalin solution instead of PFA were generally 20% higher, except for entry 3. The same process could be applied under microwave conditions in domino processes involving hydroformylation [46]. An asymmetric hydoformylation process using the same metal complex, PFA, and the chiral phosphine ligand (S,S)-Ph-BPE can be realized [47]. Good enantio- and regioselectivity was obtained in the hydroformylation of a range of styrenes, with branched products being obtained preferentially. However, the yields obtained with a formalin solution were >20% higher, except for entry 7.

795

796

36  Paraformaldehyde

Table 36.3  The hydroformylation of alkenes with paraformaldehyde (PFA) as a CO source.

Entry

1

R

Conditions

Selectivity

Yield of the major product [%]1; %ee

Ref

n-C6H3

[Rh(COD)Cl]2 (1 mol%), BIPHEP, Nixantphos (2 mol% each), toluene, 90 °C

97:3 (l:b)

72

[31]

2

BnO(CH2)6

96:4 (l:b)

47

[31]

3

PivO(CH2)8

98:2 (l:b)

73

[31]

4

Ph

93:7 (b:l)

92%; 90% ee

[33]

[Rh(COD)Cl]2 (0.5 mol%), [(S,S)-Ph-BPE] (1.2 mol%), tolune, 80 °C

5

4-OMe-C6H4

96:4

63%; 95% ee

[33]

6

4-F-C6H4

96:4

69%; 95% ee

[33]

7

4-CF3-C6H4

8

Ph

9

4-OMe-C6H4

10

t-Bu

11

CO2Bu

88:12 [Rh(COD)Cl]2 (0.1 mol%), DPPP (0.4 mol%), toluene, 120 °C, H2 (10 bar)

96%; 67% ee

[33]

3:5

83

[35]

2:5

74

[35]

22:1

88

[35]

3.7:1

17

[35]

1

Conversions are shown, not yields; BIPHEP=2,2ʹ-bis(diphenylphosphino)-1,1ʹ-biphenyl; Nixantphos = 4,6-bis(diphenylphosphino)phenoxazine; COD = 1,5-cyclooctadiene.

The same Rh catalyst with dppp as ligand in the hydroformylation of styrenes with PFA [48], required higher operating temperature of 120 °C and the use of gaseous H2 to provide a positive pressure that allowed higher yields to be obtained. In this case, the linear products were favored. Aliphatic alkenes (e.g. cyclooctene, butylacrylate) provided a linear product in a highly regioselective and efficient manner. The enantioselective hydroformylation of Z-alkenes using PFA as CO source [49] uses [Rh(acac) (CO)2] (2 mol%) in combination with 3 mol% Ph-BPE. The aldehyde products could be obtained in good yields, as a 65:35 mixture of regioisomers, ees between 90–98%, from unsymmetrically substituted silbenes were reported. Using a dual catalyst system (i.e. catalysts A, [RhCl(cod)]2 with (S,S)Ph-BPE) was shown by separate experiments to be responsible for syngas formation and catalyst B, [Rh(acac)(CO)2] with (Sax,S,S)-BOBPHOS as a ligand produced the best regioselectivity (69:31) and a good ee (Figure 36.7). With this system, there was a slow release of syngas with catalyst A,

36.2  Carbonylation and Related Reactions

Figure 36.7  The Rh-catalyzed hydroformylation of allylbenzene and the synthesis of cyclic hemiacetal [49]. Catalyst A = [RhCl(cod)]2 with (S,S)-Ph-BPE; catalyst B = [Rh(acac)(CO)2] with (Sax,S,S)-BOBPHOS.

whereas catalyst B promoted the asymmetric hydroformylation reaction. The same catalyst combination was used with allylbenzene, which is a challenging example because isomerisation to the internal conjugated alkene, prop-1-en-yl benzene, is thermodynamically favored. Most catalysts give either no selectivity or favor the linear aldehyde in hydroformylation of alkenes of this type. Catalysts with the BOBPHOS ligand (Figure 36.7) are exceptions and 75–80% selectivity can be observed. With the catalysts A and B in combination, a 69:31 b:l ratio could be obtained at high conversion with a high ee of 79% (Figure 36.7). The same strategy and reaction conditions were used to convert (S)-N,N-dibenzyl allylglycinol into a cyclic hemiacetal also with very high conversion (>99%), good regioselectivity (b : l, 82 : 14) and d.r (86 : 14). It was found that if the alkene and catalyst B were joined to the solution of A and paraformaldehyde after a certain time (30 minutes), presumably enough for CO and H2 formation, the b:l ratio improved. While the search for novel applications of PFA in synthesis continues, other studies have provided important clues that may help with future developments. One of them was the use of in situ Raman spectroscopy to follow paraformaldehyde depolymerization at different temperatures [50]. In these experiments the concentration of H2C(OMe)2, the product of the paraformaldehyde depolymerization in methanol, was monitored by the change in intensity of the νS(OC–O) vibration at 913 cm–1. It was found that the presence of even small (catalytic) amounts (0.15 mol%) of the common carbonylation catalyst [Pd(dtbpx)(MeOH)2]2+, when present in solution, can slow down the depolymerization process. [Pd(dtbpx)]2+ fragments formed from [Pd(dtbpx)(MeOH)2]2+ act as inhibitors in this reaction. They change the electron density at its reactive sites. The development of ways to control this inhibitory effect could help to develop better processes. More recent studies using in situ nuclear magnetic resonance (NMR) spectroscopy and quantum mechanical calculations were used to unravel the mechanism of the alkoxycarbonylation of alkenes with different surrogate molecules, including PFA [21]. Interestingly, no free CO could be found in these studies, although free CO resulting from the depolymerization of (CHO)n and its interaction with the metal catalysts, is postulated to be formed in many of these reactions. The authors found that, instead, the reaction proceeds via the C–H activation of in situ generated methyl formate, the key intermediate. The authors proposed a reaction mechanism based on the fact that Pd-hydride, Pd-formyl, and Pd-acyl species could be observed by NMR, the last one for the first time. Also of interest in the field of carbonylation are recent developments on the application of [13C]PFA for the introduction of isotope labels [51].

797

798

36  Paraformaldehyde

36.2.5  N-formylation Formamides represent a specific class of amides that are classically synthesized by using activated carboxylic acid derivatives or more recently using methanol under dehydrogenative or oxidative conditions in presence of amines [5, 52]. An early report for the utilization of PFA for the N-formylation of primary and secondary (a)cyclic amines uses [(Cp*IrI2)2] (1 mol%) as a catalyst in water at elevated temperature for 5–10 hours, giving the corresponding formamides in moderate to good yields (Figure 36.8; 13 examples, yields 41–95%) [53]. One of the latest findings demonstrated that PFA, as well as methanol, undergoes N-formylation with amines in presence of a bicatalytic system consisting of Cu(I)/TEMPO (2,2,6,6-tetramethylpiperidine 1-oxyl radical) catalysts that are well known for the selective oxidation of alcohols to aldehydes with oxygen as terminal oxidant (Figure 36.8) [5]. Interestingly, the reaction proceeds at low temperature and selectively to form N-formamides using PFA or methanol for the formylation. This observation is specific for these reagents and stays in vast contrast to the conversion other alcohols or aldehydes in presence of amines that form selectively imines using Cu/TEMPO [5]. Primary, secondary, (a)cyclic, aliphatic, and benzylic amines react to formamides in low to good yields (17 examples; yields 12–97%). In contrast, simple aniline is not suitable for the conversion to formamides because it undergoes homo-coupling under these conditions to form azobenzene. However, other aromatic amines can be N-formylated. Furthermore, in another study it has been demonstrated that catalytic

Figure 36.8  a) N-formylation of amines with paraformaldehyde (PFA) with an iridium catalyst. b) N-formylation of amines with PFA with Cu/TEMPO as catalysts. c) N-formylation of lactams with Shvo’s catalysts yielding N-formylimides under hydrogen evolution.

36.3  Methylation and Related Reactions

N-formylation of lactams (14 examples; yields 39–99%) is possible with PFA by the utilization of Shvo’s catalyst at elevated temperature (150 °C) within 1h (Figure 36.8) [54]. Various lactams were smoothly converted into N-formylimides without the use of a stoichiometric activating reagent and hydrogen as the only byproduct.

36.3  Methylation and Related Reactions For a long time, formalin and PFA has been applied to insert methyl (-CH3), methylene (-CH2-) and hydroxymethyl (-CH2OH) moieties into various substrates (Figure 36.9) [6, 27]. Thus, aldol condensation, Baeyer diarylmethane synthesis, Blanc and Mannich reactions, Petasis, PrinsKriewitz, Tiffeneau, and calixarenes syntheses were widely used to introduce C1 terminal or bridging block into numerous scaffolds (Figure 36.9) [27]. For example, in the classical Mannich

Figure 36.9  Some established synthetic protocols to introduce C1 blocks into various substrates.

799

800

36  Paraformaldehyde

reactions of formalin, an amine and an enolizable carbonyl compound propargyl amines can be obtained (Figure 36.9) [55]. To expand the substrate scope and improve yields, many metal catalyst-based systems have been developed [56–59]. The hydrohydroxymethylation of terminal alkynes is an important reaction for the synthesis of propargylic alcohols (Figure 36.9). Thus, CuIcatalyzed alkynylation of formaldehyde with formation of primary propargyl alcohols has been developed [60]. Compared to the former procedures, gaseous formaldehyde and moisture sensitive Grignard reagents were omitted, while the yields were higher and the reaction proceeded at a faster rate. As early as 1979, the first example of a copper bromide-catalyzed synthesis of terminal allenes from alkynes, PFA and diisopropylamine was reported [61, 62], and this direction has been greatly developed since then [63, 64]. PFA is also known as a methylating agent in a Rh-catalyzed methylation of ketones with CO/ formaldehyde/water system (Figure 36.9) [65]. Under similar reaction conditions, the N-methylation of amines can be also performed. For example, morpholine can be methylated with 92% yield using [Rh] as a catalyst [66], while under high CO pressure PFA was employed as the formylation reagent for amines in the presence of a cobalt complex [67]. In this case, the imine was first formed and then reduced by hydrogen from the water-gas shift in the presence of a cobalt catalyst. In a more recent study, it has been demonstrated that a redox self-sufficient reductive amination occurs employing PFA both as a carbon source and reducing agent [2]. The methylation of primary and secondary amines is possible in the presence of 0.5 mol% of various ruthenium p-cymene dimers, giving the best results with the [(Ru(p-cymene)I2)2], and with excellent yields for many substrates after only two hours (Figure 36.10). In the case of primary amines, the dimethylated products have been formed. In addition, biphasic reductive N-methylation and catalyst recycling has been successfully demonstrated. The reductive methylation of amines is also related to transfer-hydrogenation, respectively, to the concept of the borrowing hydrogen principle, because the hydrogen is generated from one of the substrates which are incorporated into the target product [68–70]. In the traditional Baeyer diarylmethane synthesis (Figure 36.9), strong acids are usually needed in large amounts. However, in the presence of an [In] catalyst, the reactions of electron-rich arenes proceeded smoothly with trioxane to give the corresponding diarylmethanes in good yields [71]. And a selective palladium-catalyzed cross-coupling reaction between arylalkenes and aminals, which were generated in situ from amines and paraformaldehyde (Figure 36.11), can be realized as well [72]. PFA can be used for regioselective metal-catalyzed reductive coupling with 1,3-dienes, allenes, and alkynes [6]. The products of these reactions cannot be formed in a selective manner by

Figure 36.10  Redox self-sufficient reductive N-methylation.

36.3  Methylation and Related Reactions

Figure 36.11  Pd-catalyzed synthesis of cinnarizine via a one-pot vinylation of amines via aminals to allylic amines.

conventional hydroformylation. The complete inversion of the regioselectivity is currently possible in the reactions of dienes and of alkynes. Examples are shown in Figure 36.12. Coupling of PFA to 2-substituted dienes can occur at the C1, C2, C3 or even the C4 carbon atoms in a regioselective manner, by a variation in catalysts and reaction conditions (Figure 36.12a). Nickel and ruthenium-based catalysts have been found to work well for C1 coupling, a process involving Ni(0) catalysis [73]. Formaldehyde plays the role of carbonyl electrophile and terminal reductant. Continuing studies on the reactions of 2-substituted dienes with higher aldehydes using silanes, boranes, and organozinc reagents as the terminal reductants [74], isolated the metallacycles involved and showed that metallacycle formation was reversible, whereas the formation of the C1 adducts, by oxidative coupling, was kinetically preferred [75]. Such oxidative pathways also appear to take place in the corresponding diene–PFA reductive couplings, with the observed C1 regioselectivity suggesting a kinetically controlled process (Figure 36.12, Eq 1.). Accordingly, it is expectable that 2-substituents on the diene capable of weakening the newly formed C–C bond of the C1 adduct might enable equilibration between π-allyl A and π-allyl B. This would lead to the formation of C4 adducts, e.g. eqs 1 and 4. Trialkylsilyl or trialkylstannyl substituents at position 2, in which hyperconjugation of the C–Si or C–Sn σ-bond with the σ-antibonding orbital of the newly formed C–C bond at the C1 position can occur, would be candidates for this type of coupling, which was indeed observed with Ni(0) catalysis. A neutral ruthenium catalyst allows the synthesis of C3 coupled products [76], whereas cationic Ru catalysts hydrometalate reversibly at all diene positions, providing access to the C2 coupling products bearing all-carbon quaternary centers (Eqs 2 and 3, Figure 36.12b) [77]. In the ruthenium-catalyzed reactions to form the C3 and C2 coupling products, isopropanol and PFA both contribute as terminal reductants. Alkyne hydroxymethylation has been achieved with PFA and formic acid as the terminal reductant [78]. Primary allylic alcohols were obtained with good to complete levels of regioselectivity (Figure 36.12c). The use of nickelcatalysis under conditions of reductive coupling provides the isomeric allylic alcohols with good to complete levels of regioselectivity. The initial products are formate esters, which are hydrolyzed during isolation. 1,1-Disubstituted allenes and PFA can react under conditions of reductive coupling with ruthenium catalysis and isopropanol as terminal reductant, to provide the primary neopentyl alcohols with complete levels of branch regioselectivity [79]. When CF3-substituted allenes were employed, alcohols containing CF3-bearing all-carbon quaternary stereocenters could be obtained (Figure 36.12d) [80]. These transformations provide a greener alternative to conventional methods employed in alkyne functionalization that require metallic, pyrophoric, or highly mass intensive terminal reductants (e.g. ZnR2, BEt3, HSiR3) [6]. Cycloaminomethylation of pyrrole or indole involving their СН- and NH-reactive sites can be accomplished with a 1:2 mixture of primary alkyl(phenyl)amine and PFA in the presence of 5 mol% zirconium or nickel catalysts (Figure 36.13) [81]. Thus, performing the reaction in the

801

802

36  Paraformaldehyde

Figure 36.12  (a) Hydroxymethylation of dienes. (b) C2 vs. C3 regioselectivity in the ruthenium-catalyzed reductive coupling of 2-substituted dienes with paraformaldehyde (PFA). (c) Hydroxymethylation of alkynes; (d) Hydroxymethylation of allenes.

presence of ZrOCl2·8H2O catalyst resulted in a mixture of aminomethylated products with up to 86% combined yield (Figure 36.13a), whereas cycloaminomethylation of pyrrole with NiCl2·6H2O or [Ni(Py)4Cl2]·0.76H2O catalysts proceeds at positions 2 and 5 of the pyrrole ring, leading to the formation of piperazinopyrroles reaching maximum of 48% yield (Figure 36.13b). Cycloaminomethylation of indole in the presence of ZrOCl2·8H2O catalyst was accomplished effectively at positions 1 and 3, giving rise to 3-alkyl(phenyl)-3,4-dihydro-2H-1,5-(metheno)[1,3] benzodiazepines (Figure 36.13c). Addressing the essential need in a sustainable strategy for the synthesis of quinazoline compounds from aniline derivatives and PFA, a recyclable GO@Fe3O4@APTES@FeL (GOTESFe)

36.3  Methylation and Related Reactions

Figure 36.13  Cycloaminomethylation of pyrrole (a, b) and indole (c) (R = n-Pr, n-Bu, t-Bu, Hex, Ph).

catalytic composite was developed by grafting the Fe(III)−Schiff base complex (FeL) (L = 2,2-dimethylpropane-1,3-diyl)bis(azanylylidene)bis-(methanylylidene)bis(2,4-Cl-phenol) onto 3-aminopropyltrie-thoxysilane (APTES)-coated Fe3O4 nanoparticle-decorated graphene oxide (GO). The catalytic composite was exploited for the synthesis of dihydroquinazoline-based compounds with acetonitrile/tetrahydrofuran (THF) as the solvent and paraformaldehyde as the carbon source under mild and acid free conditions (Figure 36.14) [82]. The magnetic property of the GOTESFe composite enables the catalyst to be recycled and reused up to five times without a visible loss in catalytic activity. The synthetically useful catalytic preparation of primary propargylic alcohols from terminal alkynes and formaldehyde can be realized with a nanosilver catalyst and PFA (Figure 36.15) [83]. To devise such catalytic particles, titanium oxide, and nanosilver were dispersed on polymethylhydrosiloxane-based cross-linked semi-interpenetrating networks (PMHS-SIPNs) to form supported nanoAg@TiO2@PMHSIPN composite. Although the aliphatic and silicon-linked alkynes were not suitable substrates in this reaction, this procedure deserves consideration for the preparation of aromatic propargylic alcohols due to its practicality and simplicity of operation. The modification of the the Mannich condensation of α,ω-diacetylenes with secondary diamines and aldehydes has been found as an efficient approach to diazaalkatetraynes and tetraazatetraacetylenic macrocycles [84]. In this study, the reaction of α,ω-diacetylenes with PFA and acyclic and cyclic diamines (2 : 2 : 1) in the presence of CuCl catalyst (10 mol.%) in toluene at 100  °С within eight hours produces diazatetraacetylenic compounds in 29–68% yields (Figure 36.16a). The catalytic cyclocondensation of diaza tetraacetylenes with formaldehyde and piperazine (1 : 2 : 1 molar ratio) in dioxane in the presence of CuCl affords tetraazatetraacetylenic macrocycles in 34–39% yields (Figure 36.16b).

803

804

36  Paraformaldehyde

Figure 36.14  Synthesis of substituted dihydroquinazolines from aniline derivatives and paraformaldehyde (PFA) with GO@Fe3O4@APTES@FeL (GOTESFe) composite as a catalyst.

36.3  Methylation and Related Reactions

Figure 36.15  Synthesis of primary propargylic alcohols from terminal alkyne and formaldehyde catalyzed with nanosilver Ag@TiO2@PMHSIPN composite.

Figure 36.16  Catalytic aminomethylation of diacetylenes with secondary diamines and paraformaldehyde (PFA) to give diazaalkatetraynes (a) and tetraazatetraacetylenic macrocycles (b).

Medicinally relevant and biologically active chiral tertiary alcohols can be prepared by asymmetric decarboxylative cycloaddition of vinylethylene carbonates (VECs) with formaldehyde under the cooperative catalysis of achiral palladium complex and chiral squaramide [85]. With combination of a palladium complex generated in situ from Pd2(dba)3•CHCl3 (2.5 mol%), an achiral phosphine ligand (10 mol%) and chiral squaramide OC (25 mol%; OC = organocatalyst) as cooperative catalysts, the reaction of VECs with PFA (10 equiv.) proceeded smoothly to give desired tertiary alcohol derivatives (acetal protected diols) in good yields (51%–65%) with moderate enantioselectivities (62%–79% ee) (Figure 36.17). The reaction conditions are also suitable for the reaction of VECs with electronic deficient arylaldehydes to afford desired products in high yields with good enantioselectivities, although the catalytic system is less effective for the control of the selectivities.

805

806

36  Paraformaldehyde

Figure 36.17  Asymmetric decarboxylative cycloaddition of vinylethylene carbonates (VECs) with formaldehyde for the construction of acetal protected diols with a chiral tertiary alcohol.

The azaindole ring system is one of the most valuable heterocyclic moieties; however, only limited methods have been developed for functionalization of 7-azaindoles. Recently a facile synthesis of hydroxymethylated N-aryl-azaindoles was developed via a Ru(II)-catalyzed regioselective C-H addition to PFA (Figure 36.18) [22]. The reaction is compatible with air, shows high functional group tolerance and regioselectivity, and is an environmentally benign method without any undesired byproduct. The C-H hydroxymethylation proceeded efficiently over a broad range of substrates irrespective of their electronic nature. A variety of substrates containing electron-donating and electron-withdrawing groups at the meta- or para- position of the phenyl ring were proved to be productive substrates for this coupling reaction, affording the corresponding products in moderate to good yields. Synthetically important functional groups, such as halogen, methoxy, ester, and ketone groups were well tolerated, enabling further functionalization. Moreover, for substrate bearing meta-Me group, the C-H hydroxymethylation exhibited excellent regioselectivity in favor of the sterically more accessible C-H bond. Several aryl and aliphatic aldehydes were also examined by using the current catalytic system. A Rh(II)-catalyzed multicomponent reaction is able to trap the in situ generated α-imino enols, and rapidly affords α-amino-β-indole ketones [86]. In particular, 1-sulfonyl-1,2,3-triazoles can be transformed via α-imino metal carbene species by vinylimine ions using C(2)-substituted indoles and PFA as precursors in the presence of a Rh(II) catalyst (Figure 36.19). A broad range of triazoles provided the corresponding products in moderate to good yields. 1-Mesyl–substituted triazoles bearing a bromide or other electron-rich groups at the C-4′ of the phenyl ring produced good yields of the corresponding products. 1-Mesyl–substituted triazole featuring a heterocyclic group also produced a good yield of the desired product. Bulky triazoles were found to be tolerated despite the steric effect and can equally furnish the corresponding products. This transformation was also applicable to a range of C(2)-substituted indoles . An evaluation of the substituents on the indolic nitrogen revealed that N-H and N-Bn indoles were both compatible with the reaction conditions to

36.3  Methylation and Related Reactions

Figure 36.18  Ru(II)-catalyzed C-H hydroxymethylation of 7-azaindoles.

Figure 36.19  Synthesis of α-amino-β-indole ketones from 1-sulfonyl-1,2,3-triazoles, C(2)-substituted indoles and paraformaldehyde in the presence of a rhodium(II) catalyst.

provide the products in 55 and 62% yield, respectively. The resulting products and density functional theory (DFT) calculations indicated that the enolic carbon had a stronger nucleophilicity than the traditional enamic carbon in the trapping process. Dearomatization reactions have garnered considerable interest because they represent an useful way to prepare complex three-dimensional molecular scaffolds from readily available aromatic compounds. In particular, dearomative methods that form both C–H and C–C bonds represent an important class of reaction that are particularly versatile and efficient. In this view, the recently developed single point activation of pyridines, using an electron-deficient benzyl group, facilitates the ruthenium-catalyzed dearomative functionalization of a range of electronically diverse pyridine derivatives [87]. The key aspect involves the use of a ruthenium catalyst in conjunction with a highly electron-deficient 2,4-bistrifluoromethylbenzyl activating group (Figure 36.20). This tailored activating group can be readily removed to furnish the free amine. Ranging from 4-arylpyridines to the electron-rich 4-methoxypyridine, the hydroxymethylation delivers hydroxymethylated piperidines in good yields. A noteworthy feature of this work is that paraformaldehyde acts as both a hydride donor and an electrophile in the reaction, enabling the use of cheap and readily available feedstock chemicals. Removal of the activating group can be achieved readily, furnishing the free NH compound in only two steps. Mechanistic work has shown that the metal catalyst oxidizes formaldehyde in methanol to methyl formate, forming a metal hydride in the process. The synthetic utility of the method was illustrated in the synthesis of paroxetine. Since the last century, the challenge has remained to produce the valuable ethylene glycol (EG) directly from the C1 building block formaldehyde in a single step. In the classical systems, the reaction conditions were very harsh, often with pressures above 400 bar. However, under milder

807

808

36  Paraformaldehyde

Figure 36.20  Ruthenium-catalyzed dearomative hydroxymethylation of pyridines toward hydroxymethylated piperidines.

Figure 36.21  [Rh-(pincer ligand)] catalyzed carbonylation of paraformaldehyde (PFA) toward ethylene glycol.

conditions, the selectivity was on the side of glycol aldehyde (GA) and the hydrogenation product methanol. However, Rh-catalyzed method for the carbonylation of PFA, which allows the direct one pot synthesis of EG from PFA at relatively mild conditions (70 bar, 100 °C) with yields up to 40% has been described (Figure 36.21) [88]. Application of a symmetrical pyridine-based PNP ligand was essential. Moreover, a straightforward and convenient ruthenium(II)-catalyzed synthesis of 3-unsubstituted phthalides from aryl amides and PFA via C–H activation can be realized (Figure 36.22) [89]. The reaction proceeds through tandem ortho-hydroxymethylation of aryl amide and subsequent intramolecular lactonization. An one-pot copper(I)-catalyzed synthesis of multi-substituted 2-azolylimidazole derivatives is possible with N-propargylcarbodiimides, azoles, PFA and secondary amines through a domino addition/A3 coupling/ cyclization process (Figure 36.23) [90]. The N-propargyl-N′arylcarbodiimides bearing different substituents underwent the reaction smoothly. Reactions of the N-propargylcarbodiimides containing an electron-donating or electron-withdrawing phenyl group proceeded well to afford the corresponding 2-imidazolylimidazoles in moderate to good yields. The N-propargylcarbodiimides with a naphthyl group could also successfully participate in the one-pot reaction to deliver the desired products. The substrates containing a pendant o-tolyl, o-iodophenyl or α-naphthyl group were found to have some influence on the outcome of the reaction, probably due to the steric hindrance. The reaction of an N-propargyl-N′-alkylcarbodiimide gave only a 37% yield of the desired 2-imidazolylimidazole. Then the one-pot reactions of the substrates bearing different branched chains (such as n-hexyl, benzyl, β-phenylethyl, and c-hexyl chains) at the α-position of the propargyl group were also tested, and moderate to good yields of the desired multi-substituted 2-imidazolylimidazole derivatives were isolated. A variety of secondary amines including linear amines [such as NH(iPr)2 and NH(nBu)2] and cyclic amines

36.3  Methylation and Related Reactions

Figure 36.22  Ru (II)-catalyzed synthesis of 3-unsubstituted phthalides from aryl amides and paraformaldehyde (PFA).

Figure 36.23  Synthesis of multi-substituted 2-azolylimidazole derivatives from N-propargylcarbodiimides, azoles, paraformaldehyde (PFA),and secondary amines.

(such as piperidine, morpholine, N-methylpiperazine, and 1,2,3,4-tetrahydroiso-quinoline) were shown to be good partners for this one-pot reaction. The reactions with substituted imidazoles and benzimidazoles also furnished the desired 2-azolylimidazole derivatives successfully. However, the use of other azoles such as pyrrole, indole, 1,2,3-triazole, and 1,2,4-triazole failed to afford the desired 2-azolylimidazoles, probably because of their weaker nucleophilicity. The carbon-carbon bond-forming reaction between two carbonyl-containing substances to yield an α-hydroxyketone or derivative thereof is of great synthetic value. Product dehydration, selfcondensation, or polymerization are potential problems. The aldol reaction can be catalyzed either by acid or base. In the Mukaiyama-aldol version a silyl enol ether (or a chemical equivalent) is used. Formaldehyde is the smallest aldehyde that can be employed in the aldol reaction. An obvious advantage is that only the other reaction partner can be enolized and act as the nucleophile, thus limiting the number of possible products that can be obtained. Its very high reactivity can, however, be an obstacle, that the use of PFA can help. In this case the issue becomes how to generate the reactive monomer efficiently. Both metal- and organocatalyzed processes have been developed with the aid of PFA, including asymmetric reactions [91]. PFA is also an efficient hydroxymethylating agent in an enantioselective synthesis of chiral 4,5-dihydrooxazoles from α-isocyanocarboxylates catalyzed by a complex formed in situ between bis(cyclohexylisocyanide)gold (I) tetrafluoroborate and a diphosphine ligand (Figure 36.24a) [92]. This reaction proceeded in dichloromethane at room temperature, and ees of 81% could be

809

810

36  Paraformaldehyde

Figure 36.24  (a) Asymmetric aldol reaction of α-isocyanocarboxylates with paraformaldehyde (PFA). (b) Rh-catalyzed aldol reaction of methyl 2-cyanopropionate with PFA. (c) Pd-catalyzed hydroxymethylation of α-substituted β-ketoesters.

obtained. The initially obtained 4,5-dihydrooxazoles were subsequently hydrolyzed to α-alkyl serines. Moreover, methyl 2-cyanopropionate could also be reacted with PFA via Rh(I) catalysis, provided by 1 mol% of a rhodium(I) complex generated in situ from Rh(acac)(CO)2 and triphenylphosphine to yield aldol product at rt in 1 h, in 97% yield (Figure 36.24b) [93]. A catalytic asymmetric hydroxymethylation of α-substituted β-ketoesters is possible with chiral Pd(II) and Pt(II)-(R)-BINAP (BINAP = 2,2-bis(diphenylphosphino)-1,1-binaphthyl) [94]. Aldol products bearing a chiral quaternary center were obtained in high yields and ees (Figure 36.24c). It is suggested that a nucleophilic chiral square planar M(II)-enolate intermediate was the active species in this process, which did not require protection from oxygen (air) or moisture.

36.4  Hydrogen Generation and Transfer-hydrogenation Reactions In addition to the use of PFA as CO-surrogate and C1-building block (see previous sub-chapter) [6, 27], the application of PFA, respectively aq. formaldehyde for homogenously catalyzed hydrogen generation is another field of application (Figure 36.25a) [1, 95]. [(Ru(p-cymene)Cl2) 2] has been identified as catalyst precursor (13 mol%) at 95 °C, under acidic conditions for PFA depolymerisation, resulting in 75% conversion within 74 minutes following the sequence depicted in Figure 36.3. In addition to the application of PFA, the use of aq. formaldehyde solution resulted in a TON of 700 and turnover frequency (TOF) >3000 h–1 [96]. The catalytically active species formed in situ has been identified as [(Ru(p-cymene))2μ-H(μ-HCO2)μ-Cl]+ by NMR [1] and MS with multiple isotopelabelling using 2H, 13C labelled PFA respectively 2H and 18O labelled water [1, 9, 10]. Subsequently, more transition metal catalysts have been developed and identified for the hydrogen generation from PFA and/or aq. formaldehyde [25, 26, 97–99]. Other

36.4  Hydrogen Generation and Transfer-hydrogenation Reactions

cationic IrCp* complexes with cooperative ligand site which were used under basic conditions (pH 11) with moderate activity for the hydrogen evolution in the aqueous phase with formalin, respectively PFA (Figure 25b-c) [26, 100]. A promising activity at low temperature (60 °C) has been revealed using an anionic ruthenium complex known to be active for methanol dehydrogenation [12, 25, 101]; with a TON >1700 and initial TOFs of >20000 h–1 (Figure 36.25d). However, this

Figure 36.25  Examples of metal complexes suitable for H2 generation in aq. formalin and aq. paraformaldehyde (PFA) solutions [1, 25, 26, 96–100].

811

812

36  Paraformaldehyde

setup requires strong basic conditions of pH >12. The highest conversion has been observed by the application of a water-soluble biphenyldiamine ruthenium complex for hydrogen production from aq. formaldehyde under additive- and base-free conditions at pH 7 and 95 °C (Figure 36.25e) [97]. A maximum TON of 24, 000 was observed for the catalytic system after 100 hours. Moreover, the dimeric [(Ru(p-cymene)Cl2)2] complex which is frequently used as precursor for the synthesis of monomeric ruthenium amine complexes, forms by treatment with imidazole the corresponding monomeric complex (Figure 36.25f) [98]. This complex is active for hydrogen production from formaldehyde and paraformaldehyde in water with no external base at 95 oC. The catalytic system demonstrated high activities with TONs >12000 and TOFs 5175 h–1. Also, recyclability experiments of the catalyst exhibited high activity after eight consecutive cycles [98]. Interestingly, immobilized heterogeneous bifunctional catalyst materials have been realized based on coordination polymers (Figure 36.25g). The most active bifunctional catalyst contains two catalytically active sites, a sulfonic acid group and an organometallic ruthenium entity. The sulfonic acid is responsible to catalyze the hydrolysis of paraformaldehyde into formaldehyde. Then, the ruthenium catalyst unit can catalyze the formaldehyde-water shift reaction generating hydrogen and formic acid which is then decomposed into hydrogen and carbon dioxide. The TOF is up to 685 h–1 at 363 K [99]. Following the demonstrations of acceptorless metal-catalyzed PFA degradation to H2 and CO2, such approaches have been further developed for transfer-hydrogenation (including isotopelabelling with deuterated PFA and D2O) reactions for the selective conversion of enones [14], aldehydes [15], alkynes [10], and nitriles [9] and for the formation of methanol [8]. A commercially available ruthenium complex as catalyst [RuCl2(PPh3)3] is suitable for transferhydrogenation with PFA in water to reduce C=C bond of α,β-enones in a chemoselective fashion (Figure 36.26). The optimized conditions required 2 mol% of catalyst, five equvalents of PFA and K2CO3 as base for 18 hours at 110 oC [14]. Similarly transfer-hydrogenation with PFA in aqueous dimethyl sulfoxide (DMSO) can reduce aromatic aldehydes with an iron complex as catalyst (Figure 36.26). The optimized conditions required 3 mol% of the iron-catalyst, 10 eq. PFA and Na2CO3 as base for 24 hours at 120  oC [15]. The performance of the reaction in a two-chamber system to understand the nature of the production of H2 from PFA and water confirmed the in situ

Figure 36.26  Transfer-hydrogenation of enones and aldehydes [14, 15].

36.4  Hydrogen Generation and Transfer-hydrogenation Reactions

formation of formic acid [1] previous to the decomposition to CO2 and H2, which is the reducing agent to transform aromatic aldehydes to alcohols. Several aromatic aldehydes were applied as substrates, yielding the corresponding alcohols in good yields (up to 89%). Applying PFA for alkyne transfer-hydrogenation, it has been demonstrated that E-alkenes are formed with up to >99% E/Z selectivity via Ru-catalyzed partial hydrogenation of different aliphatic and aromatic alkynes (Figure 36.27). Best results were obtained with [Ru(p-cymene)Cl2]2 complex as pre-catalyst in combination with BINAP as a ligand (1:1 ratio per Ru monomer to ligand). Mechanistic investigations indicate that the E-selectivity is due to the fast Z to E isomerization of alkenes. This method is a complementary procedure to the well-known Z-selective Lindlar reduction in late-stage syntheses and also suitable for the production of deuterated alkenes simply using d2-paraformaldehyde and D2O mixtures. The use of pFA and the base-free conditions results in a high functional group tolerance and many different substrates can be converted into E-alkenes with this protocol. In the context of transfer hydrogenation reactions applying PFA in aqueous media, it turned out that a very mild method for the selective reductive deamination of nitriles to primary alcohols under very mild conditions can be realized (Figure 36.28) [9]. This reaction represents an artificial pathway which goes even further than nitrilases go during the natural detoxification of nitriles by the conversion to carboxylic acids and NH3 [102–104]. Usually, in synthetic systems, nitriles are reduced to primary amines with secondary/tertiary amines as side-products. Additionally, nitriles are often hydrolyzed to amides or even carboxylic acids in the presence of water. In contrast, the present protocol reduces nitriles to imines that react in situ to the corresponding aminals in presence of water, and under ammonia evolution, the corresponding aldehydes are formed in situ that are then further reduced to the corresponding alcohols (Figure 36.28). The protocol is applicable for a wide range of nitriles (20 examples) and neither expensive nor air or moisture sensitive chemicals are required for the conversion of nitriles to alcohols. A broad substrate scope showed very good to excellent yields under the optimised conditions. [Ru(p-cymene)Cl2]2 acts as the catalyst precursor in the presence of pFA resulting in its degradation to CO2 and H2. Nitriles play a dual role as substrate and as ligand where the binuclear catalyst structure converts to monomeric ones upon coordination of nitrile molecules.

Figure 36.27  E-selective hydrogenation of alkynes with paraformaldehyde (PFA) in presence of water.

Figure 36.28  Catalytic deaminative conversion of nitriles to alcohols with paraformaldehyde (PFA) in water as hydrogen source.

813

814

36  Paraformaldehyde

Inspired by nature, it has been shown that in addition to formaldehyde dehydrogenase mimicry, it is possible to mimic formaldehyde metabolism with a second pathway using dismutases, independent of external sacrificial redox partners (Figure 36.29, right) [8]. On one side, these formalin disproportionating enzymes exhibit a considerable structural and functional resemblance to the glutathione- independent zinc-containing dehydrogenases, with a sequence similarity greater 70%. Considering the analogy of NADH as biological hydrogen carrier, and our lately described bioinspired process featuring acceptorless H2 liberation [25], it was proposed that modification of the organometallic species and/or the reaction environment will lead to a novel homogeneously catalyzed formaldehyde-to-methanol converting system, (Figure 36.29, right), as yet unprecedented in the context of abiotic C1-valorization pathways [8]. Indeed, it was possible by slight modification of the catalyst backbone and the addition of larger amounts of phosphate buffer [8]. To exclude further reaction pathways that would lead to methanol formation, the pH-dependency and base-catalyzed disproportionation (Cannizzaro reaction), and isotope-labelling experiments (2H and 13CO2) with labelled PFA were investigated to exclude CO2-coupled or H2-coupled disproportionation (Figure 36.29). The variation of the pH showed that a Cannizzaro reaction occurs only at a pH >9.5 and, since no 12C/13C- and 1H/2H-scrambeling occurred during the reaction, we could exclude CO2-coupled or H2-coupled disproportionation. A dismutase pathway would require a H2-decoupled disproportionation, or, in other words, a direct hydride transfer would imitate the enzymatic dismutation. Under optimized conditions, a dismutase mimic reached >90% selectivity with an imidazolium-tagged ruthenium complex [8]. In a similar study, an IrCp* complex was evaluated for the PFA conversion to methanol (Figure 36.29) [105]. The cationic iridium complex, Cp*IrL(OH2)2+ (Cp* = pentamethylcyclopentadienyl, L = 2,2′,6,6′-tetrahydroxy-4,4′-bipyrimidine), is very active and highly selective for PFA to methanol conversion at 25 °C with a TOF of 4120 h–1 yielding methanol in high amounts (93%) with a high TON of 18,200. In the Ir-catalyzed reaction, the base-catalyzed Cannizzaro reaction does not play a role. In addition to the PFA-to-methanol conversion (see Figure 36.29), it has been demonstrated that the conversion of PFA with CO2 and H2 to ethanol can be catalyzed by a mixture of Ru(acac)3 and CoBr2 in 1,3-dimethyl-2-imidazolidinone (DMI) as the solvent with lithium iodide (LiI) employed as promoter (Figure 36.30) [106]. The optimized conditions and recycling tests used 2.3 mol% Ru(acac)3 and 14 mol% μmol CoBr2, 0.9 equivalent LiI, 2 mL DMI, and 3.2 mmol PFA under a pressure of 3:5

Figure 36.29  Investigations toward the realization of an organometallic dismutase mimic.

References

Figure 36.30  Conversion of paraformaldehyde (PFA) to ethanol under CO2 (3 MPa) and H2 (5 MPa) pressure.

MPa (CO2:H2) during nine hours at 180 °C, resulting in a TOF of 17.9 h–1 for the conversion to ethanol with a product selectivity of 50.9% and a total TON of 805 after five cycles [106].

36.5  Summary and Outlook In summary, the described efforts to use paraformaldehyde as a versatile reagent in the field of transition metal catalyzed organic synthesis are very promising for future applications in a wide range of applications. Paraformaldehyde can be used under mild conditions for i) transfer-hydrogenation reactions for the conversion of several functional groups acting as hydrogen source, ii) carbonylation reactions in the function of a CO surrogate and iii) as a C1-building block to incorporate methyl, methylene, and other groups. Moreover, the methods are suitable for the synthesis of isotope-labelled compounds with hydrogen and carbon isotopes. The mild conditions such as low-temperature, robustness of the catalytic systems (non-inert conditions under air in presence of moisture or even in the aqueous phase), and high chemo-selectivities are often accompanied with a good functional group tolerance, which are promising for further applications. For the future, one can expect that more base-metal catalysts will be identified for PFA activation in the so far precious metal dominated field. Moreover, the future synthesis of PFA from renewable sources and from carbon dioxide will open further possibilities for the application in organic synthesis and as energy storage molecule for the generation of hydrogen for paraformaldehyde reforming fuel cell applications.

Acknowledgement The authors acknowledge the Fundação para Ciência e Tecnologia (FCT, EXPL/QUI-QOR/1079/2021; PTDC/QUI-QOR/02069/2022, UIDB/00100/2020, UIDP/00100/2020, LA/P/0056/2020; AMFP, MNK, and MHGP), the Deutsche Forschungsgemeinschaft (DFG 411475421; MHGP) the Centro de Química Estrutural / Institute of Molecular Sciences (AMFP, MNK, MHGP), Instituto de Química da Universidade de São Paulo (IQ-USP; LHA), Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPES; 2019/10762-1; LHA), and Conselho Nacional de Desenvolvimento Científico e Tecnológico / Conselho Nacional de Pesquisas(CNPq, 312751/2018-4; LHA). Moreover acknowledgement goes to the Insituto Politécnico de Lisboa for funding the IPL/IDI&CA2023/SMARTCAT_ISEL project (MHGP).

References 1 Heim, L.E., Schlorer, N.E., Choi, J.H., and Prechtl, M.H.G. (2014). Nat. Commun. 5: 3621. 2 van der Waals, D., Heim, L.E., Gedig, C. et al. (2016). Chemsuschem 9: 2343–2347. 3 Heim, L.E., Konnerth, H., and Prechtl, M.H.G. (2017). Green Chem. 19: 2347–2355. 4 Heim, L.E., Thiel, D., Gedig, C. et al. (2015). Angew. Chem. Int. Edit. 54: 10308–10312. 5 Pichardo, M.C., Tavakoli, G., Armstrong, J.E. et al. (2020). Chemsuschem 13: 882–887. 6 Sam, B., Breit, B., and Krische, M.J. (2015). Angew. Chem. Int. Edit. 54: 3267–3274.

815

816

36  Paraformaldehyde

7 Scholten, J.D., Prechtl, M.H.G., and Dupont, J. (2010). Chemcatchem 2: 1265–1270. 8 van der Waals, D., Heim, L.E., Vallazza, S. et al. (2016). Chem. Eur. J. 22: 11568–11573. 9 Tavakoli, G. and Prechtl, M.H.G. (2019). Catal. Sci. Technol. 9: 6092–6101. 10 Fetzer, M.N.A., Tavakoli, G., Klein, A., and Prechtl, M.H.G. (2021). Chemcatchem 13: 1317–1325. 11 Heim, L.E., Konnerth, H., and Prechtl, M.H.G. (2016). Chemsuschem 9: 2905–2907. 12 Trincado, M., Grutzmacher, H., and Prechtl, M.H.G. (2018). Phys. Sci. Rev. 3: psr-2017-0013. 13 Bahmanpour, A.M., Hoadley, A., and Tanksale, A. (2015). Green Chem. 17: 3500–3507. 14 Li, W.F. and Wu, X.F. (2015). Eur. J. Org. Chem. 2015: 331–335. 15 Natte, K., Li, W.F., Zhou, S.L. et al. (2015). Tetrahedron Lett. 56: 1118–1121. 16 Leopold, M., Siebert, M., Siegle, A.F., and Trapp, O. (2021). Chemcatchem 13: 2807–2814. 17 Seibicke, M., Siebert, M., Siegle, A.F. et al. (2019). Organometallics 38: 1809–1814. 18 Siebert, M., Seibicke, M., Siegle, A.F. et al. (2019). J. Amer. Chem. Soc. 141: 334–341. 19 Schieweck, B.G. and Klankermayer, J. (2017). Angew. Chem. Int. Edit. 56: 10854–10857. 20 Thenert, K., Beydoun, K., Wiesenthal, J. et al. (2016). Angew. Chem. Int. Edit. 55: 12266–12269. 21 Geitner, R., Gurinov, A., Huang, T.B. et al. (2021). Angew. Chem. Int. Edit. 60: 3422–3427. 22 Li, S.Q., Yu, Y., Yang, Y.X., and Zhou, B. (2020). Heterocycles 100: 934–945. 23 Midya, S.P., Sahoo, M.K., Landge, V.G. et al. (2015). Nat. Commun. 6. 24 Chen, J.B., Natte, K., and Wu, X.F. (2016). J. Organom. Chem. 803: 9–12. 25 Trincado, M., Sinha, V., Rodriguez-Lugo, R.E. et al. (2017). Nat. Commun. 8: 14990. 26 Suenobu, T., Isaka, Y., Shibata, S., and Fukuzumi, S. (2015). Chem. Commun. 51: 1670–1672. 27 Lia, W.F. and Wu, X.F. (2015). Adv. Synth. Catal. 357: 3393–3418. 28 Liu, Q., Yuan, K.D., Arockiam, P.B. et al. (2015). Angew. Chem. Int. Edit. 54: 4493–4497. 29 Martins, L.M.D.R.S., Phillips, A.M.F., and Pombeiro, A.J.L. (2018). In: Sustainable Synthesis of Pharmaceuticals: Using Transition Metals as Catalysts (eds. M.M. Pereira and M.J.F. Calvete), 193–229. Oxford: Royal Society of Chemistry. 30 Faisca Phillips, A.M. and Pombeiro, A.J.L. (2019). In: Alkane Functionalization (ed. A.J.L. Pombeiro and M.F.C.G. da Silva), 476–513. Chichester: John Wiley & Sons. 31 Wu, L.P., Liu, Q., Jackstell, R., and Beller, M. (2014). Angew. Chem. Int. Edit. 53: 6310–6320. 32 Sang, R., Hu, Y.Y., Razzaq, R. et al. (2021). Org. Chem. Front. 8: 799–811. 33 Liu, Q., Wu, L.P., Jackstell, R., and Beller, M. (2014). Chemcatchem 6: 2805–2809. 34 Sang, R., Schneider, C., Razzaq, R. et al. (2020). Org. Chem. Front. 7: 3681–3685. 35 Matsuda, T., Fukuhara, K., Yonekubo, N., and Oyama, S. (2017). Chem. Lett. 46: 1721–1723. 36 Li, W.F. and Wu, X.F. (2014). J. Org. Chem. 79: 10410–10416. 37 Furusawa, T., Morimoto, T., Oka, N. et al. (2016). Chem. Lett. 45: 406–408. 38 Abu Seni, A., Kollar, L., Mika, L.T., and Pongracz, P. (2018). Mol. Catal. 457: 67–73. 39 Martins, L.M.D.R.S., Faisca Phillips, A.M.M.M., and Pombeiro, A.J.L. (2020). In: Synthetic Approaches to Nonaromatic Nitrogen Heterocycles (ed. A.M.M.M.F. Phillips), 119–160. Chichester: John Wiley & Sons. 40 Xu, X.F., Yang, Y.R., Chen, X. et al. (2017). Org. Biomol. Chem. 15: 9061–9065. 41 Chakrabortty, S., Almasalma, A.A., and de Vries, J.G. (2021). Catal. Sci. Technol. 11: 5388–5411. 42 Okano, T., Kobayashi, T., Konishi, H., and Kiji, J. (1982). Tetrahedron Lett. 23: 4967–4968. 43 Ahn, H.S., Han, S.H., Uhm, S.J. et al. (1999). J. Mol. Catal. A - Chem. 144: 295–306. 44 Rosales, M., Gonzalez, A., Gonzalez, B. et al. (2005). J. Organomet. Chem. 690: 3095–3098. 45 Makado, G., Morimoto, T., Sugimoto, Y. et al. (2010). Adv. Synth. Catal. 352: 299–304. 46 Cini, E., Airiau, E., Girard, N. et al. (2011). Synlett 42: 199–202. 47 Morimoto, T., Fujii, T., Miyoshi, K. et al. (2015). Org. Biomol. Chem. 13: 4632–4636. 48 Uhlemann, M., Doerfelt, S., and Borner, A. (2013). Tetrahedron Lett. 54: 2209–2211.

References

49 Pittaway, R., Dingwall, P., Fuentes, J.A., and Clarke, M.L. (2019). Adv. Synth. Catal. 361: 4334–4341. 50 Geitner, R. and Weckhuysen, B.M. (2020). Chem. Eur. J. 26: 5297–5302. 51 Nielsen, D.U., Neumann, K.T., Lindhardt, A.T., and Skrydstrup, T. (2018). J. Labelled Comp. Radiopharma. 61: 949–987. 52 Ortega, N., Richter, C., and Glorius, F. (2013). Org. Lett. 15: 1776–1779. 53 Saidi, O., Bamford, M.J., Blacker, A.J. et al. (2010). Tetrahedron Lett. 51: 5804–5806. 54 Lee, H., Kang, B., Lee, S.I., and Hong, S.H. (2015). Synlett 26: 1077–1080. 55 Tramontini, M. (1973). Synthesis 1973 (12): 703–775. 56 Kabalka, G.W., Zhou, L.L., Wang, L., and Pagni, R.M. (2006). Tetrahedron 62: 857–867. 57 Li, P.H. and Wang, L. (2007). Tetrahedron 63: 5455–5459. 58 Ohta, Y., Oishi, S., Fujii, N., and Ohno, H. (2008). Chem. Commun. 835–837. 59 Feng, H.D., Jia, H.H., and Sun, Z.H. (2015). Adv. Synth. Catal. 357: 2447–2452. 60 Kundu, S.K., Mitra, K., and Majee, A. (2015). RSC Adv. 5: 13220–13223. 61 Crabbe, P., Fillion, H., Andre, D., and Luche, J.L. (1979). J. Chem. Soc. - Chem. Commun. 859–860. 62 Searles, S., Li, Y., Nassim, B. et al. (1984). J. Chem. Soc. - Perkin Trans. 1: 747–751. 63 Yu, S.C. and Ma, S.M. (2012). Angew. Chem. Int. Edit. 51: 3074–3112. 64 Luo, H.W. and Ma, S.M. (2013). Eur. J. Org. Chem. 2013: 3041–3048. 65 Watanabe, Y., Shimizu, Y., Takatsuki, K., and Takegami, Y. (1978). Chem. Lett. 215–216. 66 Watanabe, Y., Yamamoto, M., Mitsudo, T., and Takegami, Y. (1978). Tetrahedron Lett. 1289–1290. 67 Sugi, Y., Matsuda, A., Bando, K.I., and Murata, K. (1979). Chem. Lett. 363–364. 68 Watson, A.J.A. and Williams, J.M.J. (2010). Science 329: 635–636. 69 Edwards, M.G., Jazzar, R.F.R., Paine, B.M. et al. (2004). Chem. Commun. 90–91. 70 Edwards, M.G. and Williams, J.M.J. (2002). Angew. Chem. Int. Ed. 41: 4740–4743. 71 Sun, H.B., Hua, R.M., and Yin, Y.W. (2006). Tetrahedron Lett. 47: 2291–2294. 72 Xie, Y.J., Hu, J.H., Wang, Y.Y. et al. (2012). J. Amer. Chem. Soc. 134: 20613–20616. 73 Kopfer, A., Sam, B., Breit, B., and Krische, M.J. (2013). Chem. Sci. 4: 1876–1880. 74 Kimura, M. and Tamaru, Y. (2007). Top. Curr. Chem. 279: 173–207. 75 Ogoshi, S., Tonomori, K., Oka, M., and Kurosawa, H. (2006). J. Amer. Chem. Soc. 128: 7077–7086. 76 Smejkal, T., Han, H., Breit, B., and Krische, M.J. (2009). J. Amer. Chem. Soc 131: 10366–10367. 77 Han, H. and Krische, M.J. (2010). Org. Lett. 12: 2844–2846. 78 Bausch, C.C., Patman, R.L., Breit, B., and Krische, M.J. (2011). Angew. Chem. Int. Edit. 50: 5686–5689. 79 Ngai, M.Y., Rucas, E., and Krische, M.J. (2008). Org. Lett. 10: 2705–2708. 80 Sam, B., Montgomery, T.P., and Krische, M.J. (2013). Org. Lett. 15: 3790–3793. 81 Akhmetova, V.R., Bikbulatova, E.M., Akhmadiev, N.S. et al. (2018). Chem. Heterocycl. Comp. 54: 520–527. 82 Chakraborty, A., Chowdhury, T., Menendez, M.I., and Chattopadhyay, T. (2020). ACS Appl. Mater. Interf. 12: 38530–38545. 83 Dong, X.Y., Gao, L.X., Zhang, W.Q. et al. (2016). Chemistryselect 1: 4034–4043. 84 Khabibullina, G.R., Zaynullina, F.T., Tyumkina, T.V. et al. (2019). Russ. Chem. Bull. 68: 1407–1413. 85 Khan, I., Li, H.F., Wu, X., and Zhang, Y.J. (2018). Acta Chim. Sin. 76: 874–877. 86 Liu, S.Y., Yao, W.F., Liu, Y. et al. (2017). Sci. Adv. 3. 87 Marinic, B., Hepburn, H.B., Grozavu, A. et al. (2021). Chem. Sci. 12: 742–746. 88 Meyer, T., Konrath, R., Kamer, P.C.J., and Wu, X.F. (2021). Asian J. Org. Chem. 10: 245–250. 89 Zhou, C., Zhao, J.Q., Chen, W.K. et al. (2020). Eur. J. Org. Chem. 2020: 6485–6488.

817

818

36  Paraformaldehyde

90 Lin, Y.F., Li, E.F., Wu, X. et al. (2020). Org. Biomol. Chem. 18: 1476–1486. 91 Meninno, S. and Lattanzi, A. (2016). Chem. Rec. 16: 2016–2030. 92 Ito, Y., Sawamura, M., Kobayashi, M., and Hayashi, T. (1988). Tetrahedron Lett. 29: 6321–6324. 93 Kuwano, R., Miyazaki, H., and Ito, Y. (2000). J. Organomet. Chem. 603: 18–29. 94 Fukuchi, I., Hamashima, Y., and Sodeoka, M. (2007). Adv. Synth. Catal. 349: 509–512. 95 Prechtl, M.H.G., Heim, L.E., and Schloerer, N.E. (2013). H2-Produktion—Verwendung von Formaldehyd, Paraformaldehyd, Wasser und Methandiol (Formaldehydhydrat) zur Wasserstoffproduktion (H2, HD, D2, T2) und Energiespeicherung. German patent 10 2013 011 379 2 (2013). 96 Heim, L.E., Vallazza, S., van der Waals, D., and Prechtl, M.H.G. (2016). Green Chem. 18: 1469–1474. 97 Wang, L., Ertem, M.Z., Kanega, R. et al. (2018). ACS Catal. 8: 8600–8605. 98 Awasthi, M.K. and Singh, S.K. (2021). Sustain. Energy Fuels 5: 549–555. 99 Shen, Y.B., Bai, C., Zhan, Y.L. et al. (2020). Chempluschem 85: 1646–1654. 100 Fujita, K., Kawahara, R., Aikawa, T., and Yamaguchi, R. (2015). Angew. Chem. Int. Edit. 54: 9057–9060. 101 Rodriguez-Lugo, R.E., Trincado, M., Vogt, M. et al. (2013). Nat. Chem. 5: 342–347. 102 Gong, J.S., Lu, Z.M., Li, H. et al. (2012). Microb. Cell Fact. 11: 142. 103 Pawar, S.V. and Yadav, G.D. (2014). Ind. Eng. Chem. Res. 53: 7986–7991. 104 Sugai, T., Yamazaki, T., Yokoyama, M., and Ohta, H. (1997). Biosci. Biotech. Bioch. 61: 1419–1427. 105 Wang, L., Ertem, M.Z., Murata, K. et al. (2018). ACS Catal. 8: 5233–5239. 106 Zhang, J.J., Qian, Q.L., Cui, M. et al. (2017). Green Chem. 19: 4396–4401.

819

37 Hydrogen Storage and Recovery with the Use of Chemical Batteries Henrietta Horváth, Gábor Papp, Ágnes Kathó, and Ferenc Joó Department of Physical Chemistry, University of Debrecen, P.O. Box 400, Debrecen, Hungary

37.1  Introduction Except for nuclear and tidal energy, the ultimate source of all energy presently available on Earth is the Sun. The energy of sunshine which made possible animal and plant life on Earth many million years ago and has been preserved for us in the so-called fossilic energy sources (coal, oil, and gas). However, whatever rich deposits of such fossilic energy carriers are available, their total amount is limited (and unevenly distributed). An even greater problem of our times is that burning (or other chemical use) of coal, oil, and natural gas severely contributes to disrupting the fine equilibrium between the absorption (fixation) and emission of atmospheric carbon dioxide by the global ecosystem, including human activity. Transportation and various industrial processes (such as the production of portland cement) yield huge amounts of carbon dioxide. In May 2022, the CO2 concentration in the atmosphere passed over 420 ppm (Mauna Loa Observatory, Hawaii, USA), in contrast to 280 ppm prior to the Industrial Revolution. Such a huge increase in the concentration of this greenhouse gas has already brought about unfavorable climate changes (commonly referred to as global warming), and a continuing exponential accumulation of CO2 in the atmosphere ultimately may lead to global climate catastrophes with severe detrimental consequences on civilization. An ideal solution to this problem would be the capture and utilization of at least as much atmospheric CO2 during a given time as is released to the atmosphere in the same time period. Note that such a steady-state situation would only stabilize the atmospheric CO2 level at the present concentration and would not result in elimination of the climate problems caused so far by global human activity. It is brutally shown by the data of environmental monitoring stations, that despite international efforts, even this steady-state of CO2 emission/fixation could not be achieved in the last few decades. More, and more effective, measures have to be taken! Science has long realized the need for procedures for decreasing CO2 emissions to the atmosphere and several strategic directions have been identified. One of those is the replacement of fossilic fuels by renewable ones. Coal, oil, and natural gas store and concentrate the energy of the early Sun. Apart from their other uses as chemical raw materials, these concentrated energy sources can be conveniently stored, transported, and used as fuels, too, according to demand. Harnessing the Sun today is still a possibility either directly in photovoltaic devices, or indirectly (e.g. by means of wind, gravitational, hydroelectric, tide, or underwater current power plants). However, the Catalysis for a Sustainable Environment: Reactions, Processes and Applied Technologies Volume 3, First Edition. Edited by Armando J. L. Pombeiro, Manas Sutradhar, and Elisabete C. B. A. Alegria. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd.

820

37  Hydrogen Storage and Recovery with the Use of Chemical Batteries

production of such renewable forms of energy fluctuates over time and needs concentration and storage in forms (preferably as liquids or gases) that can be used in a manner adapted to the actual energy demand. The term renewable is somewhat misleading, because nothing is renewed here; instead, the million years old fossilic storage materials of solar energy are replaced by new storage forms of recently captured energy of unceasing sunlight. Nevertheless, this approach is, indeed, suitable for mitigation of atmospheric CO2 emisssions, provided that high capacity storage solutions and specific devices are available both for local and mobile applications. It should also be borne in mind that the capture and storage of renewable energy has its price (which also can be expressed in quantities of CO2); obviously, this price must compare favorably to the obviated CO2 emission. It is also worth mentioning here that although energy storage may critically contribute to the success of atmospheric CO2 depletion via replacement of fossil fuels, its role is much more diverse and includes, for example, the synthesis of high-energy materials, construction of small and large electric batteries for myriad uses, and other applications, some of which may be even accompanied by zero or negative net carbon dioxide emission. In the complex system of questions and answers for replacing fossilic fuels with renewable ones, one of the key problems is energy storage. Energy can be stored in several forms (e.g. electric, magnetic, gravitational), including storage in suitable chemicals (i.e. as the energy of chemical bonds). The interested reader can find a large number of books and reviews describing various aspects of this important topic [1–5]. In this chapter, we treat only a small section of the field, namely chemical energy storage with the use of hydrogen batteries.

37.2  Hydrogen as an Energy Storage Material At present, hydrogen is used in several large-cale industrial processes (e.g. steel, petrochemical, fertilizer industries). For practical reasons, 95% of industrial H2 is produced from fossilic raw materials, mainly by decomposition of methane (natural gas) in the presence of steam, a process which results in formation of carbon dioxide, too. Hydrogen has long been advocated as a general energy carrier of the future, playing a central role in the so-called Hydrogen Economy. This suggestion is based on two very important properties of hydrogen. First, it can be produced by electrolysis of water (or direct photochemical water fission, not yet realized on the technological scale). Second, its direct oxidation in fuel cells produces electricity accompanied by formation of water as the sole product [6]. Provided that cheap electricity is available (from renewable, or nuclear/geothermal sources), electrolysis of water results in storage of electric energy as the energy content of the generated hydrogen. The energy can be liberated in fuel cells based upon the oxidation of H2 by O2 with no obvious harm to the environment since only water is produced as unique product. This way the overall process converts electric energy first to H2 and, after some time of storage, back to electric energy (Figure 37.1). Efficient methods of storage are needed for mitigation of daily or seasonal fluctuations of H2 availability from renewable sources, or for transportation of energy in another form than electricity. In addition, the H2O → H2 → H2O cycle does not involve carbon-containing compounds and does not contribute to the increase of atmospheric CO2 level (i.e. it is carbon neutral). Not to forget, though, that production and operation of the various elements of the process (electrolyzers, fuel cels, energy production facilities) have their carbon footsteps. The produced hydrogen can be used in many applications, and, beause the location and time of electrolysis and the end use of H2 can be separated, in principle, hydrogen can be a major chemical for energy storage. At 1 bar pressure, hydrogen is a diatomic gas with a molecular mass Mw = 2.016 × 10–3 kg (2.016 g) and a molar volume Vm = 22.41 × 10–3 m3 (22.41 L) at 0 °C, and Vm = 24.05 × 10–3 m3 (at 20 °C).

37.3  Chemical Hydrogen Storage

Figure 37.1  The role of a hydrogen storage unit in mitigation of the fluctuations of electric energy from renewable sources.

Because the heat of combustion of H2 is –285.82 kJ/mol (–141.80 MJ/kg), burning of 1 L H2 in a fuel cell supplies 11.88 kJ/mol heat energy, as a theoretical maximum [7]. The energy need of a 60 kW compact car is 216 MJ/h, which would require the burning of 756 mol (1.523 kg) i.e. 18.1 m3 (!) of H2 stored at room temperature. In comparison, the mass and volume of gasoline with the same energy content (heat of combustion of 47.31 MJ/kg, average density 0.75 L/kg) are only 4.567 kg or 6.1 L, respectively [8]. These data show clearly that H2 at atmospheric pressure is unsuitable for powering transportation vehicles and that for fuel purposes H2 must be stored in more condensed form. Storage of hydrogen is a complex problem that, however, can be solved on many ways [2, 3]. Hydrogen can be stored with the use of physical methods (as a liquid or compressed gas, adsorbed on the surface of various solids, including cryoadsorption, or in interstitial hydrides). Chemical methods include the use of hydrolyzable hydrides, hydrogenation-dehydrogenation of hydrocarbons or nitroaromatic compounds, alcohols, formic acid, and others. Solutions of solids such as formate salts, or carbohydrates are also suitable for hydrogen storage and generation. A few biological methods are also known which apply enzymes or whole cells for catalysis of H2 fixation in suitable compounds. The advantages and disadvantages of the individual methods should be carefully analyzed not only with respect to storage economics, but also with consideration of the requirements of the actual application. A very large field of (conceived) practical use concerns onboard generation of H2 from suitable storage systems, while a similarly important approach aims at temporary storage of electric energy available from renewable sources (such as solar, wind, and other power plants). Typically, such stationary H2 storage/delivery facilities have to deal with enormous amounts of energy storage materials. As will be described later, several reactions, applicable for chemical storage of hydrogen have been considered as basis of hydrogen batteries.

37.3  Chemical Hydrogen Storage In its simplest definition, chemical energy storage is a chemical reaction in which one chemical (the storage material, H2 acceptor A) is transformed into a product (AH2) of higher energy content (Eq. 37.1).

A + H2  AH2

(37.1)

821

822

37  Hydrogen Storage and Recovery with the Use of Chemical Batteries

In many cases, the product is used as a fuel (i.e. its energy content is liberated in a combustion process, and then the amount of stored energy is equal to the Gibbs free energy difference (∆∆Gr) of combustion of the product and the storage material). In favorable cases the process results in a product with physical properties which facilitate its use in already existing and proven devices, such as gas burners, fuel cells, or internal combustion engines. In open-end hydrogen storage and utilization processes, in general, the hydrogenated storage material is used as fuel; however, collection and reuse of the products are not a major concern. Examples are the methanation of carbon dioxide (Eq. 37.2) or the hydrogenation of CO2 to methanol (Eq. 37.3).

CO2 + 4H2 = CH4 + 2H2O

(37.2)



CO2 + 3H2 = CH3OH + H2O

(37.3)

Combustion of CH4 or CH3OH results in CO2 and H2O. Methanol can be utilized in direct methanol fuel cells, too. The resulting carbon dioxide can be released to the atmosphere, which in this case serves as a CO2 reservoir. Alternatively, CO2 can be directly captured from flue gases of power stations or other point sources with concentrated CO2 emission. What is important, though, is that the used CO2 originates from the atmosphere and finally returns to the atmosphere, so that in principle these processes are carbon neutral, i.e. they do not result in a net increase of the amount of atmospheric CO2 (except the quantities produced in the technical implementation). Methane and methanol can be efficiently transported in pipelines or tanker ships, so it is possible to synthesize these hydrogen carriers at large-capacity central facilities and distribute them geographically if required. A practical realization of this concept, hydrogenation of atmospheric carbon dioxide is carried out at the George Olah Renewable Methanol plant at Svartsengi, Iceland, where H2 production (water electrolysis) and general operation of the plant is made possible at the expense of energy from the Svartsengi geothermal power plant. In fact, methanol was suggested by Olah as the central platform chemical of the so-called Methanol Economy [9]. A very recent review of Onishi and Himeda describes the state-of-art of homogeneous catalysis for CO2 hydrogenation to methanol and methanol dehydrogenation to hydrogen generation [10]. Several homogeneous catalytic processes for aqueous reforming of methanol resulting in H2/CO2 mixtures have been developed recently with Ru(II)-complexes 1–5 (Figure 37.4) [11–14]. These processes were not applied in hydrogen batteries, although the reversibility of the hydrogenation/dehydrogenation steps were demonstrated in some cases [15]. An efficient homogeneous catalytic acceptorless dehydrogenation of methanol selectively to formaldehyde has not been disclosed, yet. Formaldehyde is a commodity chemical which can be dehydrogenated in aqueous systems with metal-complex catalysts under reasonably mild conditions, so it can be used as an important storage material for H2 [16]. Nevertheless, processes involving formaldehyde dehydrogenation have not been applied in hydrogen batteries, yet.

37.4  Liquid Organic Hydrogen Carriers Hydrogen is often stored chemically in liquid organic compounds (LOHCs) [17–23]. In principle, an LOHC is an organic liquid of high hydrogen content. From such a compound H2 can be made free by catalytic dehydrogenation. The liberated hydrogen is utilized for a given purpose. It is usually converted into electricity in a proton exchange membrane (PEM) fuel cell (FC), often described as a polymer electrolyte membrane fuel cell (PEMFC) [6]. The organic product of this

37.4  Liquid Organic Hydrogen Carriers

decomposition can be hydrogenated back to the original LOHC, exemplified here with the interconversion of toluene/methylcyclohexane (Figure 37.2, Table 37.1). In principle, organic hydrogen carriers could be solids or gases, as well. However, liquids are preferred, since they are easier to handle than solids, Figure 37.2  Reversible and offer higher hydrogen storage capacity than gases. Most hydrogen storage in the toluene often, both the hydrogenation and dehydrogenation processes – methylcyclohexane are catalyzed by heterogeneous catalysts, and require rather interconversion. harsh conditions (elevated temperatures, especially for dehydrogenation reactions) [18, 19]. The possibilities of H2-storage offered by homogeneous catalysis are also actively studied recently [17, 22]. Regeneration of the liquid organic hydrogen carriers (LOHCs; hydrogenation of the decomposition products) is usually done in separate devices and at different times than their actual application as fuel. This resembles the present utilization of fossil-derived fuels for vehicles or for other purposes. Typical liquid organic hydrogen carriers include aromatic hydrocarbons, N-heterocyclic or other N-containing compounds, methanol, formic acid and its salts (Table 37.1), and several others [18, 19]. The capacity of hydrogen carriers (storage materials) is characterized by their gravimetric and volumetric hydrogen densities (capacities). Gravimetric hydrogen density (capacity) is defined as the mass of available hydrogen in unit mass of the given hydrogen carrier (wt. %). Similarly, the volumetric hydrogen density means the amount of available hydrogen in the decomposition of a hydrogen storage material of unit volume. The amount of available H2 is often given as its volume under standard conditions, but can also be in units of mass, moles, or even as the energy (kJ) of its combustion. These definitions refer to single compounds, such as an alcohol or a hydrocarbon, and can be regarded as the theoretical maximum of hydrogen storage capacities. Nevertheless, solid hydrogen storage materials can also be used in solutions (these must be made with a chemically stable solvent). When working with solutions, the solubility of the hydrogenated storage material limits the theoretical amount of deliverable H2 (deliverable capacity). Conversely, when the solubility of the dehydrogenated storage material is lower than that of the hydrogenated form (this is the case with the MHCO3/MHCO2 pairs, M = alkali metal ions) exploitation of only a part (i.e. the usable capacity) of the theoretically deliverable hydrogen is possible under homogeneous conditions. In addition to its hydrogen capacity, the choice of a given LOHC depends on many other factors, such as chemical stability and side reactions, reversibility of hydrogenation/dehydrogenation, rate of charge/discharge, solvent (if any), operational temperature, corrosion, toxicity, investment and operational costs, and other technical considerations. Table 37.1  Hydrogen storage materials and their theoretical gravimetric hydrogen capacity (wt. %). CO2/Formic acid CsHCO3 /CsHCO2 (70% aq. soln) CO2/Methanol

4.4 0.8 12.6

Toluene/Methylcyclohexane

6.2

Benzyltoluene/Perhydrobenzyltoluene

6.1

Dibenzyltoluene/Perhydrodibenzyltoluene

6.2

Quinoline/1,2,3,4-Tetrahydroquinoline

3.0

N-Ethylcarbazole/Perhydro-N-ethylcarbazole

5.8

823

824

37  Hydrogen Storage and Recovery with the Use of Chemical Batteries

The use of formic acid (FA) as a LOHC deserves a special mention here. Formic acid has a gravimetric hydrogen density of 4.4 wt. %, whereas its volumetric H2 density is 53 g H2/L, and this makes it an attractive hydrogen storage material. Extensive research over the last 40 years has identified extremely active and stable homogeneous and heterogeneous catalysts for the decomposition of FA to H2 and CO2 (Eq. 37.4). In fact, a large number of homogeneous catalysts have been designed and studied in detail for FA decomposition, which resulted in an enormous contribution to the general field of chemical hydrogen storage.

HCO2H  H2 + CO2

(37.4)

It has been demonstrated unambiguously that this decomposition can supply H2 + CO2 gas mixtures suitable for use in PMEFCs with no prior separation of CO2 from H2. Because carbon monoxide is a poison for catalysts used in the fuel cells, it is important that such CO2/H2 mixtures contain less, than 10 ppm CO. The pressure of the generated gas mixture can reach as high as 120–140 bar [24–26], so that the fuel cells can operate under their optimal pressure with high efficiency. Furthermore, direct formic acid fuel cells may be available in the foreseeable future. One major drawback of using formic acid as an H2 storage material is that at temperatures, required for the practical delivery of the stored hydrogen, FA is highly corrosive. Besides, despite promising recent results, direct hydrogenation of CO2 to FA is not possible on an industrial scale without the stoichiometric use of suitable bases, such as amines or alkalies, although electrolysis of aqueous electrolytes under CO2 pressure may produce FA solutions suitable as H2 storage media.

37.5  Definitions and Fundamental Questions Because this chapter deals with hydrogen storage in so-called hydrogen batteries, it is time to define what this term covers. A hydrogen battery is a device in which H2 can be absorbed at a given temperature and H2 pressure and which is suitable for delivery of the stored hydrogen upon the sole modification of the temperature and/or H2 pressure. This definition requires that in case the hydrogen storage and delivery are catalytic processes with both directions efficiently catalyzed with the same catalytically active compound (heterogeneous or homogeneous catalyst). Furthermore, the cyclic H2 charge and discharge of a H2-battery do not require continuous or cyclic addition of any auxiliary chemicals (e.g. acids or bases). Altogether, the operation of a H2-battery is much similar to an electric battery. This is exemplified in Figure 37.3. This definition has rather strict requirements regarding the chemical process of hydrogen storage. First, the reaction (Eq. 37.1) should be fully reversible, with no degradation of the storage material during repeated cycles. Second, the eqilibrium between the hydrogenated and dehydrogenated form of the storage material must be mobile (kinetic requirement) and (under the technically feasible pressure and temperature conditions) must not be in extreme favor of one or the other (thermodynamic requirement). Any required catalyst, too, should be chemically stable during the repeated cycles of Figure 37.3  General scheme of hydrogen storage and delivery. These requirements highly limit the a hydrogen battery. Reproduced number of chemical reactions which can serve as basis for hydrogen with permission from Ref [27] / batteries which truly conform to this definition. John Wiley & Sons.

37.5  Definitions and Fundamental Questions

A hydrogen battery is not a closed system thermodynamically, because there is cyclic in/out mass transfer of H2. In the literature, though, one may find mention of so-called open hydrogen storage systems. This description is used for systems composed of two processes, in which the hydrogen storage (half cycle) occurs with one catalyst, and the hydrogen delivery (the other halfcycle) proceeds with another catalyst, usually after some kind of physical separation of the storage material. In several cases, the corresponding reactions of the hydrogen storage material have been demonstrated, but the two half cycles were not coupled to result in a H2-battery that could satisfy the definition. These considerations lead us to the question: What are H2-batteries good for? Their most obvious feature is simplicity, in that a single reactor serves both for the charge and for the discharge of H2 upon only reasonable changes in pressure and temperature. This may be advantageous at remote or hardly accessible locations, where renewable solar, wind, and other energy is available in abundance but with fluctuating intensity, which prevents direct supply of the generated electricity to the grid. In these cases, application of a charge/discharge H2-buffer device can be useful to achieve the required constant electrical parameters. Hydrogen batteries may also be practical, when the charged storage material produced by the individual wind, solar, or other generators can be collected in large reservoirs and processed (dehydrogenated) in central units for feeding H2 to fuel cells. With processes for sufficiently fast hydrogenation of the oxidized (depleted) storage materials, even mobile applications of hydrogen batteries can be envisaged, provided that H2 becomes available at “gas” stations of the future. Alternatively, freshly charged containers of H2-storage materials could simply replace the exhausted ones (which may be recharged at the filling station). Obviously, in the latter cases the theoretical mass or volumetric H2-density of the storage material is of paramount importance (together with the efficiency of the discharge process). These considerations are valid for all types of H2 storage, but of course, some particular properties of a storage process may contribute with different weight than others to the overall efficiency (and applicability) of the actual method of hydrogen storage. For energy requirements, the bottom line is that the energy consumed for H2 storage must be only a small fraction of the utilizable energy of the stored hydrogen. This is why storage of liquid hydrogen is not preferred (high costs of liquification and cooling in addition to safety concerns), and this is also the reason for which an ideally functioning storage device should make possible the delivery of the total (theoretically available) amount of stored H2. In chemical hydrogen storage, the storage step is the hydrogenation of a suitable substrate. This reaction is usually an exothermic process at the mild temperature and pressure conditions of H2 entry into storage. (Elevated hydrogen pressure may be used for speeding up the charging of the storage device.) Conversely, H2 removal from storage generally requires high temperature and low H2 pressure. In case both processes result in high conversions, the thermal efficiency (so the incurred costs) of the total charge/discharge process mostly depends on the thermodynamic parameters of the reaction partners (H2 and the storage material). The process parameters, such as the temperature difference between the charge and discharge steps (heating and cooling requirements), and the need for compression/decompression of H2, together with the general costs of operation (e,g, those of moving the liquid storage solution, automatization, continuous monitoring of the process, safety precautions) must also be taken into account. Consequently, for the practical utilization of chemical hydrogen storage devices, heat management is of utmost importance (e.g. use of proper insulation, heat-exchangers) [18, 28, 29]. It should also be mentioned that when the stored hydrogen is converted to electricity in fuel cells, the heat generated by H2 oxidation can be utilized to cover (at least in part) the heat requirement of the discharge process [29]. In general, these considerations on heat management are applicable to both stationary and mobile equipments which use fuel cells and stored H2.

825

826

37  Hydrogen Storage and Recovery with the Use of Chemical Batteries

37.6  Catalysts Applied in Hydrogen Batteries Reduction of oxidized (dehydrogenated) storage materials with H2 can be carried out both with heterogeneous and homogeneous catalysts and also with biocatalysts (with the use of isolated enzymes or whole-cell systems). The catalysts applied in hydrogen batteries can be chosen from any of the categories described in this section. Heterogeneous catalysts [30] include the well-known Pd/C [31], but may be much more complex (e.g. metal nanoparticles, Pd nanoclusters supported on graphitic carbon nitride [GCN]) [32, 33] or a bimetallic single atom catalyst supported on reduced graphite oxide (rGO) [34]. In hydrogenation of CO2 to formate in the presence of 1,8-diazabicyclo[5.4.0]undec-1-ene (DBU), Ir-complexes with polymerized cyclic (alkyl)(amino)carbene ligands (PCAAC-Ir), 6, were found more effective than their un-polymerized analogs [35]. Homogeneous catalysts are exemplified by well-known transition metal complexes, such as [RuCl2(PPh3)3] (PPh3 = triphenylphosphine) or the similar, but water-soluble, dimeric [{RuCl2(mtppms-Na)2}2] (mtppms-Na = meta-monosulfonated triphenylphosphine, sodium salt; 7). Mixed ligand phosphine-NHC (NHC = N-heterocyclic carbene) complexes, such as [Ir(cod)(emim) (mtppms)] (cod = 1,5-cyclooctadiene, emim = 1-ethyl-3-methyl-imidazole-2-ylidene) also proved excellent catalysts both for hydrogenation and dehydrogenation reactions. So far, only one hydrogen battery is known which is based on biocatalysis. When required, heterogeneous (contact) catalysts are easily removed from the liquid hydrogen storage medium (e.g. by stopping the flow through the catalyst bed). This is an obvious advantage when charged batteries are transported between locations, or when the supply of H2 as fuel has to be discontinued. Conversely, tailoring the properties of a heterogeneous catalyst to the specific chemical process and reaction parameters is usually more difficult than in the case of soluble metal complexes. Dissolved metal complex catalysts come with the advantage of all metal ions being available for catalysis in contrast to the availability of active catalytic centers (atoms or their assemblies) only on the surface of a heterogeneous catalyst. However, a homogeneous catalyst resides in the same solution as the storage material, consequently, triggering and stopping the hydrogen evolution must be controlled by means other than physical separation. In general, this can be achieved by setting the optimum temperature both for the storage and delivery steps. An universal strict requirement of applicability of a catalyst (either homogeneous or heterogeneous) is its high chemical stability under operational conditions of the battery. By definition, there is no possibility to replace or reactivate a spent catalyst within the battery without removing the liquid storage material (not even mentioning the costs of frequent catalyst regeneration). Oxidative deactivation of catalysts is usually not a real danger in the reductive atmosphere of a battery, however, other types of reactions leading to loss of activity are known from long-run industrial processes. For example, hydrogenolytic splitting of phenyl groups of tertiary phosphine ligands was detected in hydroformylation reactions [36]. However, similar stability studies under real life conditions and long time-on-stream periods are scarcely found in the literature on hydrogen batteries. High chemical stability of the storage material and auxiliaries (if any) is an important requirement per se, but also because the decomposition products may poison the catalysts. In addition to their effect on catalytic efficiency, multidentate ligands, such as (iPr)2PCH2CH2P(iPr)2 [37], P(CH2CH2PPh2)3 [38, 39] or chelating bis-NHC-s, 10, 11 [40, 41] provide increased stability to the complexes. For the same reason pincer-type complexes (several examples on Figure 37.4) are also among the preferred catalysts. Examples of these catalysts are shown in Figure 37.4. Chemical hydrogen storage, including the use of hydrogen batteries, too, is envisaged from small to very large scale applications. Consequently, the price and availability of the catalysts are of crucial concern. So far, the most investigated and best performing catalysts contain precious metals

37.7  Formic Acid and Formate Salts as Storage Materials in Hydrogen Batteries

Figure 37.4  Selected ligands and catalysts for chemical hydrogen storage.

(Pd, Ru, Rh, Ir). However, these metals are not only expensive, but all are among the “endangered elements” in the category of “rising threat from increased use” [42]. Intensive research is being devoted to development of catalysts based on more abundant and cheaper metals [43], such as Fe [38, 44], Mn 12, 13, 14 [37, 45, 46], and Co 16 [47]. Very recently, a Mn(I)-pincer complex, 13, has been described with appropriate catalytic properties for use in hydrogen batteries (Figure 37.4) [45].

37.7  Formic Acid and Formate Salts as Storage Materials in Hydrogen Batteries It is hard to distinguish precisely the chemical hydrogen batteries based on formate/bicarbonate equilibrium in aqueous solution from those operating on the principle of HCOOH decomposition to H2 + CO2. Although neat formic acid can be used for this purpose, in most cases the storage medium is an aqueous solution of FA, so the reactions take place in the presence of water. The rate of decomposition (gas evolution) as a function of the pH goes through a sharp maximum around the pKa of formic acid, so much, that several catalysts show greatly reduced activity in very acidic FA solutions. For this reason, some formate or bicarbonate salt or alkali hydroxide (e.g. KHCO2 or KOH, respectively) are given to the storage medium [39, 48, 49] (e.g. in a HCO2H:HCO2Na = 9:1 ratio) [48]. With the progress of formic acid decomposition, the reaction mixture becomes more and more basic due to the formate additive dissolved in the residual water content of the used FA, which again results in diminished activity of the applied catalyst. Correspondingly, in the known systems of catalytic FA decomposition with formate additives, the added formate salt is found unchanged after the complete dehydrogenation of formic acid. Such problems are not encountered in catalytic dehydrogenation of aqueous formate to bicarbonate, because the pH of the reaction

827

828

37  Hydrogen Storage and Recovery with the Use of Chemical Batteries

mixture can change only between the pH of the aqueous solutions of the formate and bicarbonate salts, respectively (e.g. between pH 8.9 and 10.9 in case of the HCO2Na/HCO3Na pair at 1 M, 25 °C).

37.7.1  Formic Acid as a Hydrogen Storage Material In the gas phase, hydrogenation of CO2 to formic acid is thermodynamically unfavored with a ∆Gr,298 K = +32.9 kJ/mol. In an aqueous solution, however, hydration of the reactant gases and the product formic acid makes the reaction thermodynamically favorable with ∆Gr,298 K = –4 kJ/mol (Eq. 37.5). Still, under a gas phase with partial pressures of P(H2) = P(CO2) = 1 atm, the calculated equilibrium concentration of formic acid is only [HCO2H]aq = 1.33×10–4 M [50]. Indeed, attempts of FA synthesis in pure water produced only very low amounts of formic acid even under elevated gas pressures [51–53]. Clearly, this reaction is not practical for large scale synthesis of formic acid.

CO2(aq) + H2(aq)  HCO2H(aq)

(37.5)

All processes described so far utilize some kind of additives, such as amines or other bases [4, 54– 57], or water-organic solvent mixtures that include co-solvents such as DMSO [28, 53], THF [55, 58, 59], toluene [60], and ionic liquids [61–63] that facilitate dissolution of H2 and CO2 and may stabilize the product formic acid by hydrogen bond formation. Nevertheless, the highest FA concentrations achieved in these systems were typically around 2 M (well below of the concentration of neat formic acid; [FA] = 26.5 M). Aqueous-organic biphasic mixtures were also applied for continuous separation of formic acid from the reaction mixture to prevent its decomposition [64]. In a flow system, separation of formic acid from the purely aqueous reaction mixture was also achieved by electrodialysis, resulting in FA concentrations up to 2 M [65]. Interestingly, in aqueous solutions, addition of some cations (most often applied as formate salts) also resulted in some increase of the final FA concentration, up to 0.13 M [66–68]. This list of research results on formic acid synthesis is far from being comprehensive, because the main focus of this chapter is on hydrogen batteries. For the reasons outlined in the previous paragraph, formic acid is not suitable for use in hydrogen batteries in the strictest sense. Conversely, it may serve as an important hydrogen storage material provided it is synthesized in separate chemical or electrochemical processes. Hence, a very brief outline of dehydrogenation processes is presented here. Again, the list of examples is non-exhaustive. Decomposition of formic acid catalyzed by heterogeneous catalysts [69] (e.g. Pd/C) has long been known [30, 31]. Recently, active research is directed to the use of supported Pd nanoparticles [30, 32], as well as to mono- or bimetallic single atom catalysts on graphitic supports [33, 34]. Soluble catalysts (Figure 37.4) were first studied by Coffey, with the use of FA solutions in acetic acid, and of various Pt(II)-, Ru(II)-, and Ir(III)-complexes; [IrH2Cl(PPh3)3] was found the most active [70]. In their landmark studies [48], Fellay, Laurenczy, and Dyson first applied Ru(II) complexes prepared from [Ru(H2O)6](tos)2 (tos = tosylate) and various water-soluble tertiary phosphines (such as mtppts-Na3 = meta-trisulfonated triphenylphosphine Na-salt, 8, or pta = 1,3,5-triaza-7-phosphaadamantane, 9) as catalysts. Hydrogen was generated from aqueous formic acid solutions containing approximagely 10% Na-formate. Beller et al. studied Ru(II)-based catalysts with several monodentate and bidentate tertiary phosphine ligands, but the most active and stable catalysts were pincer-type complexes [such as e.g. 2 or 13]. Louloudi et al employed a triglyme/water mixed solvent to study hydrogen storage with a Ru(II)–P(CH2CH2PPh2)3 (Ru-PP3) catalyst [39]. At the beginning of the process, 20% of the used formic acid was converted to formate with addition of KOH. After decomposition of FA (but not of HCO2K), the resulting solution could

37.7  Formic Acid and Formate Salts as Storage Materials in Hydrogen Batteries

be stored with no protection from air and light and applied again for decomposition of a new batch of FA the next day (hence the term Use-Store-Reuse) and altogether for 30 days. Himeda [71–73] and Li [74] introduced Ir(III)-complexes, such as 17, 18 and 19, respectively, with pH-sensitive multidentate N-donor ligands as catalysts of FA decomposition. Similarly, pH-responsive Ir(III) and Ru(II) complexes with pyridin-2-ol-based ligands, 20, were synthesized and studied by Papish et al [75]. Papp et al described a straightforward synthesis of cis,mer-[IrH2Cl(mtppms-Na)3], which proved to be an extremely active catalyst for decomposition of formic acid [26]. The latter catalyst generated virtually CO-free (< 10 ppm) hydrogen, and was characterized by a turnover number (TON) of 674,000 (mol H2) × (mol catalyst)–1 and turnover frequencies (TOF) up to 298 000 (mol H2) × (mol catalyst)–1 × h–1. Such catalytic activites may allow practical applications; however, other reqirements, such as catalyst stability, corrosion, availability of formic acid, should also be considered.

37.7.2  Formate Salts as Hydrogen Storage Materials As defined in Section 37.5, a hydrogen battery should involve a reversible and mobile hydrogenation/dehydrogenation reaction in which the chemical equilibrium is not in extreme favor of any of the reactants. The bicarbonate/formate equilibrium in aqueous solution (Eq. 37.6) is well suited for this purpose [27, 50–52, 76–78].

HCO3− (aq) + H2(aq)  HCO2− (aq)  + H2O(l)

(37.6)

Russo et al. made a rigorous thermodynamic analysis of this equilibrium (Eq. 37.6) in the presence of various cations (H+, Na+, K+, NH4+) [78]. They considered several parameters which may influence the position of the equilibrium, such as the nonlinear effects on the activity coefficients and operating conditions (concentration of solutes, pressure of hydrogen, temperature). Their paper gives a wealth of information on the thermodynamic properties of the compounds (ions) involved in the equilibria of the mentioned formate salts (and formic acid) Moreover, the values of equilibrium constants are tabulated for the 20–100 °C temperature range in 2.5 °C steps. This largely helps the reader in determining the available maximum conversions for both directions, depending on the operational mode of the battery. For example, for the dehydrogenation process this may be a continuous hydrogen discharge against a constant outside H2 pressure, or hydrogen evolution in a closed vessel with increasing inside H2 pressure until equilibrium is reached. Results of the thermodynamic calculations were compared to experimental findings from many sources (mostly observed with heterogeneous catalysts). It should be emphasized, that in lack of thermodynamic data for non-aqueous or mixed solvent systems, such calculations are presently possible only for fully aqueous reaction mixtures. Nevertheless, in those cases the calculated data can be used as benchmark for comparison of the various catalytic systems studied under widely different operational conditions [78]. One of the main conclusion of this work is that despite the relatively low hydrogen capacity of aqueous alkali formate salt solutions, development of such hydrogen storage systems is strongly advised due to the minimal safety risks in comparison to other methods of hydrogen storage. During the operation of a hydrogen battery based on the bicarbonate-formate interconversion, reaction of bicarbonate may yield CO2 according to Eq. 37.7.

2HCO3−  CO32− + CO2 + H2O

(37.7)

829

830

37  Hydrogen Storage and Recovery with the Use of Chemical Batteries

The extent of this side reaction strongly depends on the concentration and the cation of the bicarbonate salt, the catalyst used for hydrogen storage, and the reaction temperature. However, in unfavorable cases concentration of CO2 may reach as high as 30 vol % of the gas flow delivered by the battery [79]. This leads to loss of the storage material and also may hamper the use of the stored hydrogen in fuel cells. Obviously, generation of CO2 during discharge of the battery must be minimized (eliminated) with the proper choice of the catalyst, the formate salt and the operating conditions of the battery. Autrey et al. analyzed chemical and technological (e.g. heat management) requirements of using bicarbonate/formate mixtures in aqueous systems in much detail [77]. Dehydrogenation of formate salts in aqueous systems was first studied with the use of Pd/C by Wrighton et al [31], and later by Sasson et al [80]. Generation of hydrogen with negligible CO content was also observed to accompany the transfer hydrogenation of aldehydes from aqueous formate catalyzed by [{RuCl2(mtppms-Na)2}2] [81]. Similarly, the reversibility of hydrogenation of Ca(HCO3)2 was clearly demonstrated in aqueous solution with the use of [RhCl(mtppms-Na)3] as the catalyst [67]. The first deliberate studies on using water-soluble Rh(I)-, Ru(II)-, and Ir(I)complexes as catalysts for bicarbonate hydrogenation in aqueous solution (without additives such as e.g. amines) were published in 1999 [52]. Homogeneous catalysis of hydrogenation of a bicarbonate slurry to formate in a phase-transfer catalyzed toluene/water biphasic reaction system has also been described [82]. Very recently, Beller described a Mn(I)-pincer complex, 13, which, in the presence of the potassium salt of lysine (LysK), was suitable for both CO2 hydrogenation and formate dehydrogenation in H2O/THF solutions (see also Section 39.9) [45]. Large-scale use of aqueous solutions of KHCO2 as a hydrogen storage material was suggested by Sasson et al already in 1986 [83] and recent studies by Autrey et al [77], and Russo et al [78] clarified many of the requirements that should be fulfilled for a successful practical application of the bicarbonate/formate equilibrium in hydrogen storage. Excellent reviews of the field are also available [50, 76, 84]. An important aspect of the work with aqueous solutions of alkali bicarbonates and formates is the solubility of these salts (Table 37.2) [50, 76, 78, 85]. In general, with the same cation, at room temperature formates dissolve better in water than bicarbonates. Consequently, under fully homogeneous conditions, only a fraction of hydrogen present as aqueous formate can be utilized (deliverable capacity), since hydrogen availability (usable capacity) is limited by the lower solubility of the corresponding bicarbonate. It should be mentioned, though, that only H2 enters/leaves the battery in a bicarbonate-formate hydrogen storage system. Solubility of alkali formates and bicarbonates is also influenced by the temperatures at which the hydrogenation and dehydrogenation processes take place. Table 37.2  Solubility of formate and corresponding bicarbonate salts.

HCO2Na HCO3Na

Solubility, wt.%

T, °C

Solubility, mol/kg

44.8

22

6.59

20

1.04

HCO2K

76.8

8.72

18

9.13

HCO3K

25

20

2.5

HCO2Cs

81.7

21

4.59

HCO3Cs

67.8

20

3.5

Data from Linke, W.F. and Seidell, A. (1965). Solubilities: Inorganic and Metal-organic Compounds, 4th edition. Washington, D.C.: American Chemical Society.

37.8  Catalysts and Reaction Conditions Potentially Applicable in Hydrogen Batteries

37.8  Catalysts and Reaction Conditions Potentially Applicable in Hydrogen Batteries Based on the Formate-Bicarbonate Equilibrium Next, we describe a few systems in which the applicability of the same catalyst for both the hydrogenation of bicarbonate and dehydrogenation of formate was demonstrated; however, the two half-cycles were not coupled and operated as a hydrogen battery. Olah et al. used Ru(II)-PNP pincer complex 2 (Figure 37.4) for both hydrogenation of HCO3Na and dehydrogenation of HCO2Na in H2O/THF (50 vol.%) solutions [86]. Dehydrogenations proceeded relatively slowly, but with no siginificant loss of activity in time. The total turnover number (TON) after six cycles of hydrogenation/dehydrogenation reached 11,500. In principle, construction of a hydrogen battery with this catalyst is feasible; however, such attempts were not described. Beller et al. reported that a catalyst system comprising [{RuCl2(benzene)}2] and 1,2-bis­ (diphenylphosphino)methane (dppm) was able to catalyze both hydrogenation of various alkali and alkaline earth bicarbonates in H2O/THF (5/1 V/V) and dehydrogenation of the corresponding formates in DMF/H2O (4/1 V/V) mixed solvents [79, 87]. Hydrogenation reactions typically proceeded at 70 °C under 50–80 bar H2 pressure without or with CO2 in the gas phase (up to 30 bar) with turnover numbers in the 320–1,754 in two hours. Hydrogen generation from the corresponding formates with the same catalyst was studied at 60 °C, and in a three hour reaction time 889–2000 TON was observed. HCO2NH4 reacted sluggishly with a TON of only 93. Following dehydrogenation of HCO2Na with 90% conversion, the resulting HCO3Na/HCO2Na solution was evaporated to dryness and the solid residue was subjected to hydrogenation (with no further addition of catalyst), resulting in HCO2Na in 80% overall yield. The dehydrogenation–hydrogenation (and vice versa) reactions were succesfully carried out with commercially available HCO2Na and HCO3Na, too. Based on these proof-of-concept findings, the system was suggested for construction of hydrogen batteries, although in the two half-cycles different reaction conditions (solvent mixtures, temperature, reaction time) were employed. In an other publication [88], Xin et al disclosed important findings on the effect of Lewis-acid co-catalysts on decomposition of formic acid, and also on hydrogenation/dehydrogenation of bicarbonate/formate salts catalyzed by complexes formed in situ from [Ru(2-Me-allyl)2(cod)], triphos (21) and Al(OTf)3 in 50% aqueous dioxane. The reaction was investigated in a sealed NMR tube in three consequtive runs. At 90 °C and under 70 bar H2 pressure, in all three cycles HCO3Na was hydrogenated with approximately 30% conversion within 10–14 hours, whereas dehydrogenation of HCO2Na proceeded with about the same conversions but required longer reaction times (10–30 hours). These experiments demonstrated that the catalyst system may allow construction of a hydrogen battery without any other additives. Sasson made extensive studies on the decomposition of aqueous formate to H2 and bicarbonate both with heterogeneous and homogeneous catalysts. In fact, he was the first to realize the possible industrial importance of this reaction. Already in 1986, his seminal paper [83] gave an account of the methods of hydrogen storage already known; furthermore, the technological aspects of the use of the formate-bicarbonate equilibrium for this purpose were also analyzed. In addition, Wiener, Sasson, and Blum studied the kinetic and mechanistic details of formate dehydrogenation with 10% Pd/C catalyst [80] during which 610 was determined. With the aim of constructing a hydrogen battery, Treigerman and Sasson designed a separation method of formate from the reaction mixtures obtained by hydrogenation of bicarbonate as described previously [90]. The separation process used a strongly basic anion exchanger, Dowex-Cl, and approximately 70% of formate could be separated from dilute HCO2Na solutions (0.05 M–0.50 M) in two steps. The anion-exchanger was regenerated with the use of 1 M HCl, leaving behind formic acid in the filtered solution. Because several efficient methods are known for decomposition of formic acid to H2 and CO2, the separation of formate as HCO2H during acid regeneration of the anion-exchanger may be regarded as the final stage of a formic acid cycle. Such a cycle could start with CO2, converted with a base to CO32– and HCO3–, which are then hydrogenated to HCO2–, recovered as HCO2H by ion-exchange and catalytically decomposed to H2 and CO2. Although such a reaction sequence is certainly possible, unfortunately each cycle (together with regeneration of the ion-exchanger) consumes 1–1 mol of a base and acid, respectively, which does not make the process practical for large scale storage and regeneration of H2.

37.9  Functional Hydrogen Batteries

Figure 37.5  General scheme of a hydrogen battery based on the formate-bicarbonate equilibrium.

In the following sections, the term functional hydrogen battery is applied to systems or devices that were designed with the purpose of reversible H2 uptake and delivery, with the corresponding chemi­ cal reactions catalyzed by the same catalyst in both directions, and the charge and discharge of the battery regulated only by the hydrogen pressure and/or the temperature.

37.9  Functional Hydrogen Batteries

37.9.1  Hydrogen Batteries Based on CO2–Formic Acid Cycles Plietker et al. designed a hydrogen battery with the use of the [RuCl(PNNP)(acetonitrile)]PF6 complex (Figure 37.6) as the catalyst in the presence of DBU as a basic additive [91], In a typical experiment, 20 g dry ice, i.e. 455 mmol CO2 and 65.7 mmol DBU reacted under 70 bar H2 pressure and 100 °C to yield DBU formate salt with TON values around 2000 with 0.075 mol% catalyst. Addition of toluene ensured homogeneous conditions and increased the rate of CO2 hydrogenation to formate. When the reaction stopped (indicated by no further pressure change), the reactor was cooled to room temperature and flushed with N2. The resulting solution (DBU formate in toluene) could be stored at room temperature for days, and could be decomposed to yield a H2 + CO2 gas mixture by heating to 100 °C at ambient pressure. After cooling, and addition of fresh CO2 and H2 to the reactor, the hydrogenation and dehydrogenation steps could be repeated with no need of replenishing the catalyst, DBU, or toluene. This hydrogen storage/delivery cycle was repeated five times with no significant change of the rate of the discharge process and the amount of the delivered H2. Furthermore, it was demonstrated that decomposition of DBU formate could supply a H2 + CO2 mixture of 22 bar total pressure which is advantageous for efficient use of the stored H2 in PEM fuel cells. Gas evolution could also be regulated by changes in the reactor temperature, since no decomposition of DBU formate was detected at room temperature.

Figure 37.6  Hydrogen storage in a CO2-DBU system. Reproduced with permission from Ref [91] / John Wiley & Sons.

833

834

37  Hydrogen Storage and Recovery with the Use of Chemical Batteries

As described in the paper of Plietker et al., this hydrogen battery requires addition of not only H2 but also CO2 to start each cycle. Furthermore, due to the high CO2 excess, only about 15% of the used CO2 is converted to DBU formate and the rest is vented at the end of the hydrogenation step. Also, in a complete H2 storage/delivery cycle, the reaction mixture should be twice heated to 100 °C and cooled to room temperature to start a new cycle. Nevertheless, most of these problems may likely be eliminated by proper engineering solutions, and therefore the system has potential for practical applications. Beller et al. made a very important step forward in the quest to develop viable hydrogen batteries based on reversible hydrogenation of CO2 to formic acid. They have found that in water/THF mixtures, several [MnBr(CO)2(PNP)] complexes were active both in hydrogenation of CO2 and in decomposition of FA. The most promising results were obtained with the use of complex 13 (Figure 37.4). In the presence of the potassium salt of lysine, LysK, total turnover numbers of 2,000,000 for CO2 hydrogenation and 600,000 in FA dehydrogenation were observed in studies of catalyst stability through several cycles. Dehydrogenations were achieved at 90 °C under ambient pressure conditions, whereas hydrogenations were run at 85 °C under 80 bar initial H2 pressure. Cycling of the battery started with dehydrogenation of formic acid in the presence of equivalent amount of LysK. At the end of this step, the gas phase was vented and analyzed, while the solution phase was subjected to hydrogenation. LysK made possible the retention of >99.9% of the CO2 formed in the dehydrogenation of FA (internal carbon dioxide capture); i.e. the evolved gas contained only less than 0.01% CO2. At the same time, the CO content was less than 10 ppm (the detection limit with the GC method used). Consequently, in contrast to Plietker’s system, the hydrogen storage and delivery steps coud be repeated without replenishing CO2 (or other components of the battery) between the cycles. The operation of this battery requires equivalent amounts of lysine to the FA used in the first dehydrogenation step (or to the desired FA concentration to be obtained in the hydrogenation step). However, lysine is an industrial product produced in large quantities via microbial fermentation so its availability may not limit the practical application of the H2 storage process. The battery was operated for 10 charge/discharge cycles on a 90.0 mmol scale, which shows the process to be suitable for scale-up and gives good chances for large-scale applications [45].

37.9.2  Hydrogen Batteries Based on Formate–Bicarbonate Cycles Cao et al. developed a fully aqueous hydrogen battery based on heterogeneous catalysis of HCO2K/ HCO3K interconversion under pressurized H2 and pressure-free conditions, respectively [92]. They synthesized a catalyst comprised of Pd nanoparticles supported on reduced graphene oxide (rGO). The strong metal-support interaction resulted in distortion of the crystal structure of the Pd-NP-s (leading to so-called lattice microstrain), which, in turn, largely increased the catalytic activity of the rGO-suported catalyst relative to Pd catalysts obtained with the use of other carbonbased supports. This heterogeneous catalyst proved to be an active and durable catalyst for bicarbonate hydrogenation as well as of formate dehydrogenation. Among the studied formate salts, HCO2K was by far superior to Na-, Li- or NH4-formate with regard to its reactivity and stability. In hydrogenation of HCO3K (storage of H2) of 5 mL aqueous 4.8 M HCO3K solution containing 9.6 μmol Pd under P(H2) = 40 bar, a turnover number TON = 7088 was achieved (94.5% bicarbonate conversion). After decompression at 25 °C, the resulting formate solution was subjected to dehydrogenation (H2 delivery) by heating to 80 °C, yielding H2 gas by nearly complete conversion of formate in 40 minutes. This charge/discharge cycle was repeated six times with no impairment of the battery performance. Furthermore, when a charged hydrogen storage solution

37.9  Functional Hydrogen Batteries

was stored for four months under ambient conditions, subsequent heating to 80 °C led to hydrogen evolution at the same rate and amount than in the previous cycles. These observations unambiguously showed the stability of the catalyst. It is important to note, that even at high HCO2K concentrations (>8 M) and temperatures (>150 °C) only a very small amount of CO2 (