Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications [1 ed.] 3527344977, 9783527344970

Learn more about foundational and advanced topics in polymer thin films and coatings besides species with this powerful

328 79 23MB

English Pages 768 [724] Year 2021

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications [1 ed.]
 3527344977, 9783527344970

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Inorganic and Organic Thin Films

Inorganic and Organic Thin Films Fundamentals, Fabrication, and Applications

Edited by Yujun Song

Volume 1

Inorganic and Organic Thin Films Fundamentals, Fabrication, and Applications

Edited by Yujun Song

Volume 2

Editor Yujun Song

University of Science and Technology Beijing School of Mathematics and Physics 30 Xueyuan Road Haidian District 100083 Beijing China

All books published by WILEY-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate. Library of Congress Card No.:

applied for Cover British Library Cataloguing-in-Publication Data

Cover Image: © SanerG/iStock/Getty Images

A catalogue record for this book is available from the British Library. Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at . © 2021 WILEY-VCH GmbH, Boschstr. 12, 69469 Weinheim, Germany All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form – by photoprinting, microfilm, or any other means – nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law. Print ISBN: 978-3-527-34497-0 ePDF ISBN: 978-3-527-34501-4 ePub ISBN: 978-3-527-34499-4 oBook ISBN: 978-3-527-34498-7 Typesetting SPi Global, Chennai, India Printing and Binding

Printed on acid-free paper 10 9 8 7 6 5 4 3 2 1

v

Contents

Volume 1 Biography xv Preface xvii Acknowledgments 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7

2 2.1 2.2

xxi

Introduction: Progress of Thin Films and Coatings 1 Yujun Song Introduction 1 Thin Films for the Innovation of Information Technology 2 Thin Films for Ultrasensitive Sensing Devices 7 Thin Films for Sustainable Energy Application 9 Thin Films and Coatings for Key Sources and Ecological Environment of Earth 28 Thin Films and Coatings for Biomedical Engineering and Life Science 32 Thin Films and Coatings for National Defense and Homeland Security 38 Acknowledgments 41 List of Abbreviations 42 References 44 Fundamental in Functional Thin Films and Coatings 59 Weiwei Zhang and Yujun Song Introduction 59 Theory of Magneto-electric Coupling in Magnetic Thin Films 59

vi

Contents

2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.9.1 2.9.2 2.10 2.11 2.12 2.13 2.14 2.15

3

3.1 3.2

3.2.1 3.2.2 3.2.2.1 3.2.2.2 3.2.3

Theory of Electronic Thin Films: Electronic Percolation and Spintronic Theory on the Semiconductor Thin Film 60 Theory of Metal Structural Thin Films: Metamaterials and the Negative Permeability Theory and Maxwell Theory 62 Theory of Surface Plasmon Resonance and Magnetoplasmonic Thin Films 66 Heterojunction Theory 73 Topological Insulator, Topological Semi-metal, and Perovskite 74 Acoustic Theory 77 Theory of Magnetoacoustic and Photoacoustic Coupling 79 The Mechanism of the Sound Pressure in the Presence of the Pulse Magnetic Field 80 The Mechanism of the Sound Pressure in the Presence of the Pulsed Magnetic Field and Static Magnetic Field 80 Theory of Acoustooptic Effect 82 Magnetothermal Thin Films: Phonon Thermal Theory 83 Theory of Thermoelectric Effect 84 Thermal Barrier Insulation Theory for TBC Coating 86 Permeability Theory: Fick First Diffusion Theory and Fick Second Diffusion Theory 87 Multi-physical Field Coupling Theory and Simulation Software Introduction 88 Acknowledgments 90 List of Abbreviation 91 References 91

Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing: From GMR, CMR, TMR to Quantum Anomalous Holzer Effect 95 Weiwei Zhang and Yujun Song Introduction 95 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing Based on Magnetoresistance (GMR) Effect 96 Introduction of GMR 96 Fabrication of GMR Multilayered Thin Films 97 MBE Method for the Fabrication of the GMR Devices 99 Magnetron Sputtering Method for the Fabrication of GMR Devices 99 GMR Applications for Sensors 100

Contents

3.3

3.3.1 3.3.2 3.3.3 3.4

3.4.1 3.4.2 3.4.3 3.5 3.6

4

4.1 4.2 4.2.1 4.2.2 4.2.3 4.3 4.3.1 4.3.2 4.3.3

Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing Based on Colossal Magnetoresistance (CMR) Effect 102 Introduction of CMR 102 Fabrication of Multilayered Thin Films Based on CMR Effect 103 CMR Applications 105 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing Based on Colossal Tunneling Magnetoresistance (TMR) Effect 106 Introduction of TMR 106 Fabrication of Multilayered Thin Films of the TMR Effect 107 TMR Applications 110 The Multilayered Magnetic Thin Film Based on Quantum Anomalous Holzer Effect (QAHE) 111 Summary and Perspectives 112 Acknowledgments 113 List of Abbreviation and Symbol 114 References 114

Bismuth-Substituted Iron Garnet Films for Magnetophotonics: Part A – Fabrication Methods and Microstructure Property Study 125 Andrey A. Voronov, T. Mikhailova, Olga V. Borovkova, Alexander N. Shaposhnikov, Vladimir N. Berzhansky, and Vladimir I. Belotelov Introduction 125 Fabrication Methods 126 Synthesis Technology and Conditions of Bismuth-substituted Iron Garnet Films 126 Fabrication of Fabry–Perot 1D-MPC with BiIG Bilayer 135 Fabrication of Tamm 1D-MPC with BiIG Bilayer 136 Properties of the Structures 139 Magneto-optical Properties of FP-1D-MPCs 139 Magneto-optical Properties of T-1D-MPCs with BiIG Bilayer 143 An increase of the Magneto-optical Response in the Ultrathin Films 145 Acknowledgment 155 List of Abbreviations and Symbols 156 References 156

vii

viii

Contents

5

5.1 5.2 5.3 5.3.1 5.3.2 5.3.3 5.3.4 5.3.5 5.3.5.1 5.3.5.2 5.3.5.3 5.3.5.4 5.3.6 5.3.7 5.3.7.1 5.3.7.2 5.3.8 5.4

6

6.1 6.2 6.2.1 6.2.2 6.2.3

Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications 161 Andrey A. Voronov, Daria O. Ignatyeva, Nikolay A. Gusev, Petr M. Vetoshko, Nazar V. Lugovskoy, Yujun Song, Vladimir N. Berzhansky, and Vladimir I. Belotelov Device Assemble and Application of Iron Garnet Films for Ultrasensitive Magnetometry 161 Devices Assemble and Application of BiIG Films for Biosensing 174 Devices Assemble and Application of Iron Garnet Films for Magneto-optical Eddy Current Flaw Detection 178 Introduction 178 Experimental Part 179 Introscope 180 Physical Properties of MO Sensors 181 The Sensory Properties of the EA Films 184 The Effect of Alternating Field Amplitude 184 The Effect of Alternating Field Frequency 184 The Effect of Bias Magnetic Field 185 Dynamic Domains in the Garnet Film Sensor Element 186 The Sensory Properties of the EP Films 188 Applications of MOEC: Imaging of Welds 188 Nondefective Welds 188 Defective Welds 190 Simulation of EC Magnetic Fields in Samples with Defects 191 Conclusions and Perspectives 193 Acknowledgments 193 List of Abbreviation and Symbol 194 References 194

MEMS, NEMS, AEMS, and Quantum Films for the Next Generation of Computing and Information Technology 199 Haishuai Chai, Junmei Wang, and Yujun Song Introduction 199 Typical Fabrication Methods for MEMS, NEMS, and AEMS 200 Fabrication of Microstructures 200 Fabrication Process of Complementary Metal Oxide Semiconductor (CMOS) 203 Fabrication Process of Field Emission Transistors (FET) 203

Contents

6.2.4 6.3 6.3.1 6.3.2 6.3.3 6.4 6.4.1 6.4.2 6.5 6.6

7

7.1 7.2 7.2.1 7.2.2 7.2.3 7.2.4 7.3 7.3.1 7.3.2 7.3.3 7.3.4 7.4 7.4.1 7.4.2 7.4.2.1 7.4.2.2 7.4.3

Giant Magnetoresistance (GMR) Sensor and Its Fabrication Method 204 From MEMS to NEMS and then to Quantum Films and AEMS for the Next Generation of Information Technology 205 The Trend of Microsystem Integration Technology 205 The Development Trend of Microsystem Packaging Technology 207 Challenges in the Development of Microsystems Technology 207 NEMS and AEM 210 NEMS 210 AEMS 211 Quantum Films for Information Technology 212 Summary and Perspectives 214 Acknowledgments 214 List of Abbreviations 214 References 215 Metamaterial or Metastructural Thin Films for EM Wave Control 221 Menglin L.N. Chen, Luzhou Chen, Xunwang Dang, Maokun Li, Li Jun Jiang, and Wei E.I. Sha Introduction 221 Modeling and Synthesis Methods of Metasurfaces 222 Jones Vector and Jones Matrix 223 Polarizability Model 224 Susceptibility Model 225 Equivalent Impedance Model 226 Simulation Algorithms of Quasi-periodic Electromagnetic Surfaces 227 Introduction to EM Surfaces 227 Design of Quasi-periodic EM Surfaces 228 Simulation Algorithms of Quasi-periodic EM Surfaces 229 Review of Simulation Algorithms of Quasi-periodic EM Surfaces 230 Orbital Angular Momentum of Electromagnetic Waves: Generation and Detection 233 Introduction 233 Generation of Orbital Angular Momentum 234 Geometric-phase Metasurfaces 234 Photonic Crystals 237 Detection of Orbital Angular Momentum 238

ix

x

Contents

7.5 7.5.1 7.5.2 7.5.3 7.5.4 7.5.5 7.6

8

8.1 8.2 8.2.1 8.2.2 8.2.3 8.3 8.3.1 8.3.2 8.3.3 8.3.4 8.4 8.4.1 8.4.2 8.4.3 8.4.4 8.4.5 8.5

9

9.1

Application in Spontaneous Emission Modification 241 Spontaneous Emission in Inhomogeneous Electromagnetic Environment 241 Calculation of Spontaneous Emission Rate 242 Metamaterials Enhance Spontaneous Emission 242 Metasurfaces Enhance Spontaneous Emission 243 Other Potential Application in Quantum Optics 245 Conclusion and Perspectives 245 Acknowledgments 245 List of Abbreviations 246 References 247 Semiconductor Thin Films for Information Technology 257 Na Chen Introduction 257 Fabrication of Semiconductor Thin Films 258 Molecular Beam Epitaxy (MBE) 259 Magnetron Sputtering 259 Metal–Organic Chemical Vapor Deposition (MOCVD) 260 Nonmagnetic Semiconductor Thin Films and Typical Applications 261 Semiconductor Thin Films for Light-emitting Devices 261 Thin Film Transistors for Displays 263 Phase-change Semiconductor Thin Films 264 Semiconductor Thin Films for Sensors 268 Magnetic Semiconductor Thin Films 269 Diluted Magnetic Semiconductors 270 Amorphous Magnetic Semiconductors 272 Phase-change Amorphous Magnetic Semiconductor Thin Films 275 Magnetic Semiconductor Thin Film-based Spintronic Devices 277 Prospective for Magnetic Semiconductors 279 Conclusion and Outlook 280 List of Abbreviations 280 References 280 Glass Transition in Organic Semiconductor Thin Films 285 Han-Nan Yang and Zheng-Hong Lu Introduction 285

Contents

9.2 9.3 9.4 9.5

10

10.1 10.2 10.3 10.3.1 10.3.2 10.3.3 10.4 10.4.1 10.4.1.1 10.4.1.2 10.4.1.3 10.4.1.4 10.4.2 10.4.2.1 10.4.2.2 10.4.2.3 10.4.2.4 10.4.3 10.4.3.1 10.4.3.2 10.5 10.5.1 10.5.2 10.5.3

Determination of Glass Transition Temperature in Organic Thin Films 287 Model for Predicting Glass Transition Temperature of Organic–Organic Composites 291 Model for Predicting Glass Transition Temperature of Nano-organic Composites 292 Summary 295 Acknowledgments 296 List of Abbreviations 296 References 296

Thermoelectric Films for Electricity Generation 299 Metin Yurddaskal, Melis Yurddaskal, Ozan Yilmaz, and Serdar Gultekin Introduction 299 Thermoelectricity 300 Overview of Inorganic and Organic Thermoelectrics for Thin Films 301 The Seebeck Effect 301 The Peltier Effect 307 The Thomson Effect 309 Classification of Thin Film Thermoelectric (TE) Materials 311 Inorganic Thermoelectric Thin Films 311 Bi–Te-Based Superlattices 311 Cobalt Oxide-Based Thin Films 311 Zn-Based Thin Films 312 Cu-Based Thin Films 312 Organic-based Thin Film TE Materials 313 Polyacetylene and Polyaniline 313 Poly(3,4-ethylenedioxythiophene) 313 Polypyrrole and Polythiophene 314 Other n-Type Polymers 314 Inorganic–Organic Composite Thermoelectric Thin Film Materials 314 Metal–Organic Frameworks 315 Carbon Nanotube–Polymer Composites 315 Applications of Thermoelectric Materials 315 Thermoelectric Cooling 316 Thermoelectric Power Generation 316 Organic Inverter Circuit 316

xi

xii

Contents

10.5.4 10.5.5 10.5.6 10.5.7 10.6 10.6.1 10.6.2 10.6.3 10.6.4 10.6.5 10.6.6 10.6.7 10.7

Organic Light-Emitting Diode (OLED) 318 Organic Radio Frequency Identification Tags 319 Organic DNA Sensors 319 Limitations 320 Techniques of Thin Film Deposition for Thermoelectric Device 320 Sputtering 320 Molecular Beam Epitaxy (MBE) 321 Metal–Organic Chemical Vapor Deposition (MOCVD) 321 Electrochemical Deposition (ECD) 323 Flash Evaporation (FE) 323 Thermal Evaporation 324 Pulsed Laser Deposition (PLD) 324 Conclusion and Future Trends 326 List of Abbreviations and Symbols 327 References 328 Volume 2 Biography xv Preface xvii Acknowledgments

xxi

11

Template-assisted Fabrication of Nanostructure Thin Films for Ultrasensitive Detection of Chemicals and Biomolecules: Part A – Template-assisted Nanoimprinting Technology for Functional Thin Films 339 Xiaomin Zhu, Xinhua Chen, Andrey A. Voronov, Vladimir I. Belotelov, and Yujun Song

12

Template-assisted Fabrication of Nanostructured Thin Films for Ultrasensitive Detection of Chemicals and Biomolecules: Part B – Detection of Chemicals and Biomolecules Based on Nanostructured Thin Films 381 Xiaomin Zhu, Xinhua Chen, and Yujun Song

13

Polymer-based Films for Artificial Intelligence 411 Ran Liu, Junmei Wang, and Yujun Song

14

Selective Permeable Thin Films and Membranes 447 Qiong Wu, Xiaoxiong Zhao, Lifan Peng, and Yujun Song

Contents

15

Biomass-Derived Functional Films and Coatings 489 Gao Xiao

16

Polymer Composite Coating for Anti-marine and Related Organism Corrosion 511 Kaifeng Chen, Zhipeng Xie, Yu Liang, Jingjing Wang, and Haiyan Zhuang

17

Anechoic Coating for Underwater Vehicles 549 Weiwei Zhang and Yujun Song

18

Thin Films and/or Coating for Electromagnetic Interference and Stealth 587 Junmei Wang and Yujun Song

19

Thermal Barrier Coating for Aerial and Aerospace Engine 615 Zaidao Li and Yujun Song

20

Perspectives for Thin Films and Coatings 647 Yujun Song Index

681

xiii

v

Contents

Volume 1 Biography xv Preface xvii Acknowledgments

xxi 1

1

Introduction: Progress of Thin Films and Coatings Yujun Song

2

Fundamental in Functional Thin Films and Coatings 59 Weiwei Zhang and Yujun Song

3

Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing: From GMR, CMR, TMR to Quantum Anomalous Holzer Effect 95 Weiwei Zhang and Yujun Song

4

Bismuth-Substituted Iron Garnet Films for Magnetophotonics: Part A – Fabrication Methods and Microstructure Property Study 125 Andrey A. Voronov, T. Mikhailova, Olga V. Borovkova, Alexander N. Shaposhnikov, Vladimir N. Berzhansky, and Vladimir I. Belotelov

5

Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications 161 Andrey A. Voronov, Daria O. Ignatyeva, Nikolay A. Gusev, Petr M. Vetoshko, Nazar V. Lugovskoy, Yujun Song, Vladimir N. Berzhansky, and Vladimir I. Belotelov

vi

Contents

6

MEMS, NEMS, AEMS, and Quantum Films for the Next Generation of Computing and Information Technology 199 Haishuai Chai, Junmei Wang, and Yujun Song

7

Metamaterial or Metastructural Thin Films for EM Wave Control 221 Menglin L.N. Chen, Luzhou Chen, Xunwang Dang, Maokun Li, Li Jun Jiang, and Wei E.I. Sha

8

Semiconductor Thin Films for Information Technology 257 Na Chen

9

Glass Transition in Organic Semiconductor Thin Films 285 Han-Nan Yang and Zheng-Hong Lu

10

Thermoelectric Films for Electricity Generation 299 Metin Yurddaskal, Melis Yurddaskal, Ozan Yilmaz, and Serdar Gultekin Volume 2 Biography xv Preface xvii Acknowledgments

11

11.1 11.2 11.2.1 11.2.2

xxi

Template-assisted Fabrication of Nanostructure Thin Films for Ultrasensitive Detection of Chemicals and Biomolecules: Part A – Template-assisted Nanoimprinting Technology for Functional Thin Films 339 Xiaomin Zhu, Xinhua Chen, Andrey A. Voronov, Vladimir I. Belotelov, and Yujun Song Development of Template-assisted Nanoimprinting Technology 339 Nanosphere Lithography (NSL) 340 Size and Shape Controlled Fabrication of Nanomaterials via NSL 340 Multi-hierarchy Micro Windows (MHMW) for Single Nanostructure and/or Array Identification 344

Contents

11.2.3

11.2.4 11.2.4.1 11.2.4.2 11.2.4.3 11.3 11.3.1 11.3.2 11.3.2.1 11.3.2.2 11.3.3 11.3.3.1 11.3.3.2 11.3.3.3 11.3.4 11.3.5 11.4

12

12.1 12.2

Identification of Localized Surface Plasmon Resonance (LSPR) of Single Noble Metal Nanoparticles and/or Nanoarrays by MHMW-assisted NSL 346 Development of NSL for Nanomaterials Synthesis 351 Aqueous Phase Ag Nanoparticles with Controlled Shapes Fabricated by NSL 351 Ultrathin Nanopore Arrays with Uniform Size 356 Fabrication of Periodic Uniform Nanoporous Films with Controlled Layers 357 Anodic Aluminumoxide (AAO) Template-assisted Nanoimprinting 359 Introduce of AAO Template 359 Preparation Methods 361 Preparation of Anodic Alumina Template 361 Preparation of Thin Film Substrates for Surface-enhanced Raman Scattering 362 Preparation and Characterization of Nanoporous Gold Thin Films Based on Anodic Alumina Template (AAO) 363 Substrate Transfer and Characterization of Nanoporous Gold Thin Films 364 Preparation and Characterization of Aqueous Spherical Gold Nanoparticles 367 Composite of Nanoporous Gold Thin Films and Aqueous Gold Nanoparticles 367 AAO Template-assisted Fabrication of Ultra-dense Nanoparticle Arrayed Thin Films 367 AAO Template-assisted Fabrication of Semiconductive Nanowire Thin Films 371 Summary and Perspective 376 Acknowledgments 377 List of Abbreviations and Symbols 378 References 378

Template-assisted Fabrication of Nanostructured Thin Films for Ultrasensitive Detection of Chemicals and Biomolecules: Part B – Detection of Chemicals and Biomolecules Based on Nanostructured Thin Films 381 Xiaomin Zhu, Xinhua Chen, and Yujun Song Introduction 381 Nanostructured Thin Films with Enhanced Magneto-optical Kerr Effect 381

vii

viii

Contents

12.3 12.4 12.4.1 12.4.2

12.4.3

12.5 12.5.1 12.5.2 12.5.3 12.5.4 12.5.4.1 12.5.4.2

12.5.5 12.5.5.1 12.5.5.2 12.5.5.3 12.5.5.4 12.6

13 13.1 13.2 13.2.1 13.2.2 13.3 13.3.1

Surface Plasmon-enhanced Magneto-optical Kerr Effect for Chemical Analysis 386 The Application of AAO in the Detection of Biological Cells 388 Difficulties and Bottlenecks in Cancer Treatment 388 Label-free Reflectometric Interference Microchip Biosensor Based on Nanoporous Alumina for Detection of Circulating Tumor Cells 389 Surface and Interface Engineering Multilayered Nanopore Films for Enhanced Fabry−Pérot Interferences for Biological Application 393 Nanostructured Thin Films for SERS 394 Introduction of SERS 394 Development Nanostructured Thin Films for SERS 396 Preparation of SERS Thin Film Substrates 397 Characterizations of SERS 398 Properties of Porous Gold Nanofilms with Composite Aqueous Gold Nanoparticles in SERS Based on PMMA Substrate 398 Surface-enhanced Raman Properties of Porous Gold Nanofilms and Their Composite Aqueous Gold Nanoparticles Based on PDMS Substrate 399 Chemical and Biological Application of SERS Based on Nanostructured Thin Films 401 Application of SERS in Ion Detection 401 Application of SERS in Environmental Pollution 401 SERS Detection of R6G on Magneto-optical Thin Films with Nanoholes 402 Application of SERS in Cancer Cell Detection 402 Summary and Perspective 405 Acknowledgments 406 List of Abbreviations and Symbols 406 References 407 Polymer-based Films for Artificial Intelligence 411 Ran Liu, Junmei Wang, and Yujun Song Introduction 411 Preparation and Integration Methods of Smart Polymer Thin Films and Coatings for AI 412 Preparation of Multifunctional Polymer Thin Films and Coatings 412 Integration of Smart Devices for AI 413 Thin Films or Coatings for AI+ Biological Application 415 Tactile Sensor (for Artificial Skin) 415

Contents

13.3.2 13.3.3 13.3.4 13.3.5 13.4 13.4.1 13.4.1.1 13.4.1.2 13.4.1.3 13.4.2 13.4.2.1

Thermal Sensor (for Temperature Monitoring) 417 Human Thermoregulation 418 Biosensor Film Substrate 421 Bio-Integrated Wearable Sensors 422 Thin Films or Coatings for AI+ Environmental Protection 425 Thin Films or Coatings for AI+ Environmental Monitoring 425 Ultraviolet Radiation Monitoring 425 Explosives Detection 426 Ammonia Sensor 428 Thin Films or Coatings for AI+ Wastewater Treatment 428 Efficient Degradation of Complex Phthalocyanine Dye Wastewater 428 13.4.2.2 Efficient Adsorption of Patulin 429 13.4.3 Thin Films or Coatings for AI+ Seawater Desalination 430 13.5 Thin Films or Coatings for AI+ Energy 431 13.5.1 Thin Films for AI+ Solar Energy 431 13.5.2 Thin Films for AI+ Capacitor 432 13.5.3 Thin-Film Thermoelectric Devices 434 13.6 Artificial Intelligence for Information Technology 435 13.6.1 Artificial Intelligence for Optical Polymer-based Films 435 13.6.2 AI+ Smart Display (Plastic Liquid Crystal Display) 436 13.6.3 High-efficiency THz-Wave Modulators 438 13.6.4 Thermal Management for Electronic Equipment 439 13.7 Summary and Perspectives of Thin Films and Coatings for AI+ 439 Acknowledgments 440 List of Abbreviations 441 References 442 Selective Permeable Thin Films and Membranes 447 Qiong Wu, Xiaoxiong Zhao, Lifan Peng, and Yujun Song 14.1 Introduction 447 14.2 The Principle of Membrane Separation 449 14.3 Types of Selective Permeable Membranes 452 14.3.1 Types and Characteristics of Membranes 452 14.3.2 The Rise of New Membranes 453 14.3.2.1 Polymer Separation Membrane 453 14.3.2.2 Polymer Functional Membrane 454 14.3.2.3 The Plasma Polymer Film 454 14.4 Preparation Methods of Varieties of Selective Permeable Membranes 456 14.4.1 Phase Inversion 456

14

ix

x

Contents

14.4.2 14.4.3 14.5 14.5.1 14.5.2 14.5.2.1 14.5.2.2 14.5.3 14.5.3.1 14.5.3.2 14.5.3.3 14.5.3.4 14.5.3.5 14.6

15 15.1 15.2 15.2.1 15.2.2 15.2.3 15.2.4 15.2.5 15.2.6 15.2.7 15.3 15.3.1 15.3.2 15.3.3 15.4 15.4.1 15.4.2 15.4.3 15.4.4

Interfacial Polymerization 456 Chemical Modification 456 Application of the Selective Permeable Membranes 457 Purification of Water 457 Application in Gas Separation 469 Oxygen Separation 470 CO2 Separation 472 Application in the Separation and Purification of Low Molecular Substances 476 Isolation and Purification of Oligosaccharides 476 Fructooligosaccharides 477 Galactooligosaccharides 477 Isolation and Purification of Amino Acids 477 Application in the Isolation and Purification of Antibiotics 478 Current Status and Recent Progress and Perspectives 479 Acknowledgments 481 List of Abbreviations and Symbols 482 References 483

Biomass-Derived Functional Films and Coatings 489 Gao Xiao Introduction 489 Biomass-derived Polymers 490 Collagen 491 Cellulose Nanofibrils 492 Pectin 494 Starch 495 Chitosan 496 Xylan 497 Lignin 498 Coating Technologies of Biomass Thin Films 500 Sol–Gel Coating Method 500 Atomic Layer Deposition (ALD) 501 Multilayers Coating Method 502 Degradable Biomass-derived Functionalized Films 502 Polysaccharide-based Films 502 Protein-, Pig Skin Gelatin-, Lipid-based Films 503 Biomass Plastics Films 503 Cellulose-based Films 504 Acknowledgments 505 References 505

Contents

16

16.1 16.2 16.2.1 16.2.2 16.3 16.3.1 16.3.2 16.3.3 16.3.4 16.3.5 16.4 16.4.1 16.4.2 16.4.2.1 16.4.2.2 16.4.2.3 16.4.3 16.4.3.1 16.4.3.2 16.4.3.3 16.4.3.4 16.5

17 17.1 17.2 17.2.1 17.2.2 17.2.3 17.3 17.4 17.4.1

Polymer Composite Coating for Anti-marine and Related Organism Corrosion 511 Kaifeng Chen, Zhipeng Xie, Yu Liang, Jingjing Wang, and Haiyan Zhuang Introduction 511 Microbial Corrosion and Protection 512 Primary Mucosa-forming Organisms and Their Attachment 512 Overview of Fouling Biological Control 513 Anti-mold Coating 516 Silicone Anti-mold Coating 517 Acrylic Mold-proof Coating 517 Polyurethane Anti-mold Coating 517 Epoxy-based Mold-proof Coating 517 Development Trend of Anti-mold Coatings 518 Antifouling Coatings 518 Main Types and Selection of Antifouling Agents 518 Main Types and Selection of Antifouling Coatings 520 Abrasion-Resistant Antifouling Coating 520 Self-polishing Antifouling 523 Fouling Release Type Antifouling Paint 531 Development Trend of Antifouling Technology 535 Antifouling Agent Design 535 Gel Mixed Fouling Release Type Antifouling Paint 538 Surface Microstructure Bionic Antifouling Paint 539 UV Antifouling Technology 540 Summary and Prospects 540 List of Abbreviations 541 References 541 Anechoic Coating for Underwater Vehicles 549 Weiwei Zhang and Yujun Song Introduction 549 Fundamental Physics for Sonar System: Echolocation and Doppler Effect 552 Types and Construction of Sonar Systems 553 Introduction of Some Typical Sonar Systems 555 Threat from Advanced Hostile Sonar System 565 Basic Physics of Anechoic Coatings 573 Development of Structure and Materials Design for Anechoic Coatings 574 Perspective for the Sonar System and Anechoic Coating Development 582

xi

xii

Contents

Acknowledgments 582 List of Abbreviations 583 References 583 18

18.1 18.2 18.2.1 18.2.2 18.2.2.1 18.2.2.2 18.2.2.3 18.2.2.4 18.3 18.3.1 18.3.1.1 18.3.1.2 18.3.1.3 18.3.2 18.3.2.1 18.3.2.2 18.3.2.3 18.4

19

19.1 19.2 19.3 19.4 19.5 19.5.1 19.5.2 19.5.3 19.5.4

Thin Films and/or Coating for Electromagnetic Interference and Stealth 587 Junmei Wang and Yujun Song Introduction 587 EMI Shielding Materials 589 Basic Theory of Electromagnetic Shielding 589 EMI Shielding Materials 591 Surface Coating Film Shielding Material 591 Ferromagnetic Material and Good Metallic Conductor Material 591 Conductive Coating Shielding Materials 592 Composite Shielding Materials 592 Stealth Coatings 599 Radar Stealth Coatings 600 Radar-Absorbing Materials 601 Radioisotope RAM 604 Carbonaceous-Based RAM 604 The Development of Stealth Coating 606 Negative Refractive Stealth 606 Multiband Absorbing Material 607 Plasmon Active Stealth 607 Summary and Prospects 608 Acknowledgments 608 List of Abbreviations and Symbols 608 References 609 Thermal Barrier Coating for Aerial and Aerospace Engine 615 Zaidao Li and Yujun Song Introduction 615 Superalloy Substrates for TBCs 618 TBC System Compositions 619 Applications of TBCs 620 Processing Techniques for TBCs 621 Electron Beam-Physical Vapor Deposition (EB-PVD) 621 Plasma Spraying (PS) Process 624 Solution Precursor Plasma Spraying (SPPS) Process 627 Suspension Plasma Spraying (SPS) 629

Contents

19.6 19.7

Thermal Transport in TBCs 633 Summary and Perspectives 635 Acknowledgments 637 List of Abbreviations 637 References 637

20

Perspectives for Thin Films and Coatings 647 Yujun Song Introduction 647 Development of the Subversive Novel Concepts for the Theoretical and Technological Breakthrough and New Findings of Thin Films and Coatings 647 Development of Highly Precise Fabrication Techniques for Thin Films and Coatings 650 Perspective in the Development of High-Spatiotemporal-Resolution Characterization Methods for Microstructures and Properties 655 Perspectives in the Further Study of Relationship Between Microstructure and Property 661 Perspectives of Thin Films and Coatings Promoting the Modern Technological Innovation and the Society Progress 668 Acknowledgments 671 List of Abbreviations 671 References 672

20.1 20.2

20.3 20.4

20.5 20.6

Index

681

xiii

xv

Biography Dr. Yujun Song is currently a full-time Professor in Physics and Applied Physics at University of Science and Technology Beijing (USTB), Deputy Director of the Center for Modern Physics Research of USTB, and Director of the Zhejiang Key Laboratory for Pulsed Power Technology Translational Medicine. After obtaining his PhD degree in Materials Science and Engineering in 2000, Dr. Song has successively studied/worked at the Center for Advanced Microstructures and Devices of Louisiana State University, Old Dominion University, Beihang University, University of Toronto, and Harvard University, etc. His research areas are focusing in the integrated innovation of modern physics fusing into biomedicine, information technology, and energy/catalysis mediated by novel nanomaterial and high precise fabrication technology, particularly the nanomedicines mediating ultra-strong pulsed field activating immunogenicity for novel therapy of intractable diseases. Dr. Song has been responsible for more than 30 funds as PI or co-PI, such as the National S&T Major Project, the NSFC–BRICS STI Framework Program, the National Natural Science Foundation of China. He also took part in many special funds as one key scientist, such as the major project on bio-nanoscience of NIH: the study of biomolecular transport mechanism by nanobiomolecular probes funded, the major project on nanoscience of NSF: The design and preparation of nano optical biomolecule probe for biomolecular transport study, the major project on biosensor of DARPA: design and fabrication of GMR biosensor, the NSF – EPSCoR project Microfluidic synthesis of magnetic nano particles and their application in cancer diagnosis and therapy.

xvii

Preface During the past decades, great progress has been achieved in promoting thin films and coatings toward new technological innovation and the development of basic theories such as the magnetoplasmon technology and the quantum anomalous Holzer effect. As told by Confucius, “if a man keeps cherishing his old knowledge, so as continually to be acquiring new, he may be a teacher of others.” With engineers and scientists serving as both thinkers and practitioners, it is only fitting that the recent progress in the fundamental physics and chemistry of the microstructure-dependent property of thin films and coatings be summarized, including their successful applications that will feedback to promote the deeper fundamental study on the microstructure-property relationship of this vivid field toward greater achievements in their applications. Thin films are microscopically thin layers of materials that are deposited onto a metal, ceramic, semiconductor, polymer substrate or a layer on a supporting liquid. These have been developed with thickness ranging from one atom/molecule layer to millions of atoms/molecule layers, i.e. from hundreds of micrometers (μm: 10−6 m) to nanometers (nm: 10−9 m) or even atomic layer thin (10−10 m; e.g. atomic electronic mechanics). Thin films can be conductive or dielectric (non-conductive), which are used for advanced electronic and/or optoelectronic components (e.g. capacitor, resistor, coil, cryotron, transistors), devices, or circuits for various applications (e.g. computing, sensing, signal/energy/mass transmitting or transporting or exchanging). On the other hand, coatings on the surface of substrates create or improve corrosion protection, heat and radiation resistance, thermal management, electromagnetic responses (e.g. stealth or detecting), acoustic responses (sensing or anechoic), water and ice protection, friction reduction, antifouling and antibacterial properties, and self-cleaning and other specific physical or chemical functions. Their thickness usually ranges from micrometer (10−6 m) to millimeters (10−3 m) or thicker, which can be expanded to centimeters (cm) or above (e.g. EM stealth coating or anechoic coating) and even thinner (e.g. nanocoating). Thin films and coatings are sometimes combined to achieve multifunctionality (e.g. magneto-optical, thermoelectric). Thin films and coatings are intrinsically related to surface and interface science and technology, but becoming more and more multidisciplinary. Research and application both involve basic physics, chemistry, materials science, biology and life, and engineering in materials, aerospace, marine, energy, and information technology.

xviii

Preface

Great academic and industrial achievements in information technology, computer science, new energy, aeronautics, astronautics, and ocean engineering have been made in the past decades owing to the fundamental breakthrough in surface and interface science and the corresponding precise fabrication technology. These fundamental quantum theory in thin films, construction of thin film devices precisely in atomic scale, orientation growth and microstructure control of coating, and bonding theory according to their unique applications. At present, thin films and coatings have been upgraded as a brand-new technology, not only for their own functions but also for their role as an indispensable bridge between fields of multidisciplinary science and technology leading to innovations. Quoting Herbert Kroemer, “The interface is the device.” However, a surface is only a kind of special interfaces between the gas phase and the solid phase or the liquid phase. Therefore, the fundamentals of the unique surface/interface effects on the thin films and coatings will be firstly summarized and primarily described: interfacial magneto-electric coupling, electronic percolation, metasurface and metastructure-related complex permittivity and Maxwell equation, surface plasmon resonance, magnetoacoustic and photoacoustic coupling, interfacial heterojunction, thermal barrier insulation, and the interface diffusion theory. In addition, the recently developed multiphysical field coupling theory and the related simulation software will be briefly introduced for the readers’ further study. Subsequently, the key fabrication technology, the structure–property relationship, and typical applications of some typical thin films and coatings of special functions will be summarized and discussed in detail, chapter by chapter, based on these surface/interface effects. This book can be a fundamental tool for the current researchers and commercial users of thin films and coatings who investigate the underlying physics and chemistry theory and the manual of their current and future fantastic applications. In addition, this would be a useful and powerful reference for the newcomers to enrich their knowledge and enlighten their own strategies in the development of new theories and extended applications of thin films and coatings. There is an abundance of available literature related to thin films and coatings. We have compiled them into two volumes for the readers’ convenience. Volume 1 covers topics closely related to inorganic (e.g. metallic, dielectric, semiconductor) thin films with unique magnetic, optical, electronic, and/or thermal properties will be mainly summarized together. We will summarize and discuss the recent progress on thin films and coatings in Chapter 1 and then their unique functions in detail in Chapter 2. The key thin film technologies will be presented in Chapters 3–8 mainly for the current and next generation of computing and information technology (5G and 6G). Specifically, we will discuss in Chapter 3 multilayered magnetic thin films for electron transport control and signal sensing, from GMR to CMR, then to TMR, and finally to quantum anomalous Holzer effect. We will tap on the recently developed magnetophotonic thin films including their basic physics and typical fabrication methods in Chapter 4 and their structure–property relationship and applications in Chapter 5, for example, using bismuth-substituted iron garnet thin films. The recent progress on the semiconductor thin films and related MEMs, NEMs,

Preface

and AEMs will be summarized and discussed in Chapter 6. We will then review comprehensively in Chapter 7 the recently developed metamaterials and metastructural thin films for the electromagnetic wave control, including their theoretical foundations, design routes, numerical methods, and engineering applications of the metastructural films (metasurfaces). The semiconductor thin films for information technology and one recent key issue on the organic semiconductors (or their glass transition) will be discussed in Chapters 8 and 9, respectively. In the final chapter of Volume 1, by comparing with conventional technologies, we will focus on one emerging energy technology, a thermoelectric thin film, whose main principle for the electricity generation and advantages. In Volume 2, the first three chapters are about the organic–inorganic composite thin films: the recent progress in the magnetoplasmonic thin films via template-assisted nanoimprinting lithography and their surface modification for some chemical or biological applications (e.g. surface-enhanced Raman scattering, ultrasensitive biosensor) in Chapters 1 and 2 and the organic–inorganic composite films for new energy technology in Chapter 3. Following this, the basic physicochemical theory and the recent progress of some key organic thin films will summarized and discussed: polymer functional thin films for artificial intelligence in Chapter 4; the selective permeable thin films or membranes for the mass separation, concentration and purification in Chapter 5; and biomass-derived functional films and coatings in Chapter 6. Finally, some advanced coatings for marine engineering and aero-engineering will be summarized and discussed: polymer composite coatings for anti-marine microbial coating in Chapter 7, anechoic coatings for the underwater vehicles in Chapter 8, thin films and coatings of electromagnetic compatibility and/or stealth in in Chapter 9, and thermal barrier coatings in Chapter 10, which are of important applications for national key industrial sectors and the national long-term scientific development strategies. Finally, we will share some prospects for the development of inorganic and organic thin films and coatings in the breakthrough in basic theory and concepts, design principle, experimental testing methods, the in-depth study on the relationship between microstructures and properties and their future impact on the society and economy. “Actions are done after thorough consideration rather than casual decision,” as quoted from the book “Quan Xue Jie (Learning Persuasion)” by Yu Han in Tang Dynasty. We hope that this book could be an effective tool or reference that would benefit the readers’ critical thinking to master the reasonable direction of this field. This book shall contribute to the research and teaching in this field as a valuable addition to the literature and also encourage more readers to pay attention to this rapidly developing field. It is impossible to include all the reported progress and principles with regard to thin films and coatings because both are inherently related that they significantly cover several interdisciplinary topics that bind almost all academics and industrials. Therefore, we will feel honored if only this monograph can serve as a useful handbook that can provide readers support on learning and critical thinking in order to thrust new scientific and technological developments.

xix

xx

Preface

Finally, I am much grateful to all authors of each chapter, my students who took part in writing the chapters, and the editors from Wiley during editing. I dedicated this book to my family for their utmost support and encouragement (particularly from my father, Mr Sigan Song, and my mother, Ms Xiuyun Meng) while I was editing this book during the COVID-19 pandemic. Much interestingly, my lovely daughter, Xinran Song, influenced by the endless writing and reading for this book, starts writing her own fiction novel. 2 February 2020 Beijing, China

Yujun Song

xxi

Acknowledgments This book received support from the NSFC-BRICS STI Framework Program (No. 51861145309), the National S&T Major Project (No. 2018ZX10301201), the National Natural Science Foundation of China (No. 51971029), the all-English Teaching Demonstration Course Construction Project of University of Science and Technology Beijing (No. KC2015QYW06, 2016), the Joint Research Project of University of Science and Technology Beijing and the University of Science and Technology Taipei (Grant No. TW2018007), the “100 Talent Plan” fund of Fujian province (No. 39080067), and the “1125” Zhihui Zhengzhou Talent Project of Henan province (No. 39080070 in USTB).

1

1 Introduction: Progress of Thin Films and Coatings Yujun Song 1,2 1 University of Science and Technology Beijing, Center for Modern Physics Technology, Applied Physics Department, School of Mathematics and Physics, 30 Xueyuan Road, Beijing 100083, China 2 Zhejiang Key Laboratory for Pulsed Power Technology Translational Medicine, Hangzhou Ruidi Biotechnology Company, Hangzhou 310000, China

1.1 Introduction As one cornerstone of advanced modern technologies, thin films and coatings have been expanding its scope into varieties of emerging research areas and developing rapidly to match the requirement from academics and industrials in the past decades. As elucidated in this book, thin films and coatings that are never dispensable for the development of modern science and technology will continuously play the key innovation driving force in the next generation of computing and information technology, new energy, biology and life science, new medicines, astronautics and aeronautics, geology and ocean engineering, military science, etc. Eventually pushed by the requirement of industrial sectors of countries, the successful application of this field depends intrinsically on the fundamental progress of the surface/interface science and the precision of the related mass fabrication technology. This may be the main reason for “the interface is the device” stated by Herbert Kroemer, the Nobel Prize winner in Physics in 2000 [1, 2] and for so many Nobel Prize winners in the fields relating to surface/interface science (e.g. 1930 Nobel Prize winner in Physics Chandrasekhara Venkata Raman; 1981 Nobel Prize winner in Physics Kai Siegbahn; 2007 Nobel Prize winners in Physics Peter Greenberger and Albert Fert; 2007 Nobel Prize winner in Chemistry Gerhard Eitel; 2010 Nobel Prize winners in Physics Konstantin Novoselov and Andre Geim; 2016 Nobel Prize winners in Physics David Thouless, Duncan Haldane, and Michael Kosterlitz; 2018 Nobel Prize winners in Physics Arthur Ashkin, Gerard Mourou, and Donna Strickland; from www.nobelprize.org). Thus, it is no doubt that each innovation in this field generates much profit mainly from the theoretical breakthrough and the subversive novel concepts in basic physics and chemistry, particularly the condensed matter physics and interface chemistry, and the related measurement and characterization technology, besides the invention of the advanced fabrication and synthesis technology. These innovations will enable Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications, First Edition. Edited by Yujun Song. © 2021 WILEY-VCH GmbH. Published 2021 by WILEY-VCH GmbH.

2

1 Introduction: Progress of Thin Films and Coatings

the discovery of several singular physical phenomena and materials (e.g. topological semimetals: Dirac semimetal, Weyl semimetal, nodal-line semimetals, and triple degenerate semimetals [3]; one-dimensional van der Waals [4]). Sequentially, these innovations promote the application of the finally constructed devices, equipment, and instruments using thin films and coatings as building blocks extending into novel fields (e.g. 2D transistor [5], extremely huge magnetoresistance [3]), and/or upgrading the conventional fields (e.g. rollable optoelectronic devices, such as flexible and foldable organic light-emitting diode [OLED], and graphene enforced rubber composites for industrials [6–9]). (https://china.Huanqiu.com/ article/9CaKrnJY80x; http://en.tireworld.com.cn/cnews/201937/6540.html; http:// www.gianroitire.cn/gianroitire/vip_doc/12760379.html). These will lead to the revolution of modern science and technology and then to societal reform toward a high-level civilization. Thus, thin films and coatings have been instrumental in bridging the academic studies and their industrial applications and the society. Sources, energy, environment, and sustainable development are four fundamental themes of the society that we use to provide the basic necessities of life: food, clothing, shelter, and transportation. These themes depend on the continuous technology innovation in information, transportation, communication, and public health and safety supported by the basic research in physics, chemistry, biology, and life science. Now, several issues and crises arise related to source-, energy-, and ecology-related sustainable development as well as ultrafast and high-throughput communication related to security, public health, national defense and homeland security, etc. Technologies derived from thin films and coatings have been used to develop tools.

1.2 Thin Films for the Innovation of Information Technology In the development of information technology, in order to satisfy the precise navigation for the exploration of outer space and the highly efficient nanosatellites, rapid processing of huge data and long-distance communication are needed, and the related devices should be as small as possible. With the progress of data processing and information technology, this requirement is desired not only in astronautics and aeronautics but also in many other military and society fields, leading to the formation of subversive concepts in miniaturized devices and the planar fabrication technology. Finally, silicon-chip-based thin-film fabrication via photolithography and the subsequent e-beam lithography appears formatting a brand-new industrial field: microelectronics. This requirement also intensively promoted the basic research progress in condensed matter physics (e.g. Si-based semiconductors; complementary metal-oxide semiconductors [CMOS]), chip processing technology for chip-based thin-film devices (e.g. field emission transistor [FET]; vertical transistors [10]; laser emission diodes [LED]), and microelectronic device assembly (e.g. integrated circuit), and various data and imaging processing technologies (e.g. graphics processing unit [GPU] derived software) for information technology. Particularly, several novel physical phenomena or findings related to the surface/interface effects

1.2 Thin Films for the Innovation of Information Technology

are discovered in the thin films constructed by multilayers of magnetic, electronic, and optical materials, such as magneto-optical effects, giant magnetoresistance (GMR) effects, colossal magnetoresistance (CMR) effects, and tunneling magnetoresistance (TMR) effects. A more intensive study was carried out to investigate the effects of the new physics of these thin films (see Chapters 1, 2, 5, 9, and 10), which further boosted the fundamental physics progress and simultaneously widened the range of applications of computer science and communication technology in industrials. As the related large-scale economical fabrication/processing technology was developed along with the progress and interconnection of the physical technology, chemistry synthesis, and materials science and engineering, these concepts and technologies were realized commercially, thus enabling profit generation by the primarily funded enterprises, which can invest more in basic research and advanced technology, forming a benign circling of this field. With the information technology development, more application concepts have been advanced (e.g. 3D vision and artificial intelligence [AI] technology), and ultrafast high-throughput data and image processing technology and the related information security are needed. The current Si-based chip information technology can no longer satisfy this demand since the integration limited by the feaure size of the Si-based chip will reach the ceiling against Moore’s law in the near future. Therefore, more and more novel technical concepts or terms emerge with the rapid technology evolution, such as quantum computer, quantum regulation, quantum information, magnetron light quantum, quantum entanglement communication, and long-distance low-loss optical communication. These concepts have initiated new findings, technologies, and related theoretical studies, such as spintronics, spin-orbital coupling, single photo control, the quantum anomalous hall effect (QAHE), quantum key, room-temperature superconductivity, and high k-space lasing in a dual-wavelength quantum cascade laser technology. In the literature, thin films have been reported as important media that can be used in technologies that need new materials to be further designed. As a result, many new materials for thin-film devices have been explored, such as two-dimensional (2D) materials (e.g. graphene [Gp]; black phosphorus: P-Hg; hexagonal boron nitride [hBN]; transition metal dichalcogenides [TMDs]: WSe2 , MoS2 ), topological insulators (e.g. (Hg, Cd)Te 2D materials [11] Bi2 Se3 nanoribbon [12]), magnetically doped topological insulators enable the QAHE [13], van der Waals heterojunctions (more than one single-atom layer of 2D materials organized using van der Waals force, such as magic-angle graphene [14], TMD/Gp [15], graphene/hBN [16, 17]), nanostructural hybrid multilayered thin films of high magneto-optical effects [18–22], metasurface [23–27] and metastructures [28], etc. Clearly, the development of large-scale thin-film fabrication methods at low cost is primarily required for high-performance devices fulfilling their industrial applications. Besides the conventional fabrication methods (e.g. physical vapor deposition, ion implanting, chemical vapor deposition [CVD], epitaxial growth, nanoimprinting, mechanical exfoliation from bulk crystals), large-scale high-resolution reliable synthesis and fabrication technology have been developed in the past decades. The fabrication limit has reached sub-10 nm line precision and then atomic resolution

3

4

1 Introduction: Progress of Thin Films and Coatings

fabrication or even subatomic resolution, particularly in the two-dimensional (2D) materials based thin films [10, 13, 29] or sensing devices based on single molecular assembing thin films [30, 31] or single molecular sensor-assembled thin-film devices. Two-dimensional materials contain electrons that can only move freely in non-nanoscale (1–100 nm) objects, such as nanoscale films, superlattices, and quantum wells. Two-dimensional heterojunctions can combine individual layers of different properties in the atom resolution, which can be used to study to study and understand the exhibited novel physical phenomena that can be further employed for potential applications. The availability of numerous different 2D materials – with different band structures, from semimetals to semiconductors to insulators – also makes it possible to assemble unique materials with well-designed band alignments [32]. Particularly, 2D materials based on thin films, consisting of 2D heterojunctions/heterostructures formed by atomically thin crystal layers bound by the van der Waals force, have attracted much interest because of their potential in diverse technologies, including electronics, optoelectronics, and catalysis. Therefore, 2D thin films are used as an example to exhibit this prerequisite and their development. The self-assembly approach to materials synthesis is generally related to the strong chemical bond to hold different components together (e.g. epitaxial growth via biaxially textured technology, molecular beam epitaxial and metal organic CVD), which is usually limited in those materials of highly matched crystal structures and compatible process. Now, this strategy can be extended to the weak van der Waals interaction to pre-assemble the chemical building blocks physically to realize the fabrication without chemical bonds (e.g. van der Waals heterojunctions), which can avoid the limitations of crystal lattices and manufacturing process [33]. Thus, there are two important kinds of 2D heterostructural materials that can be developed conveniently for the fabrication of microstructures for more tunable physicochemical properties of thin films and coatings. One is the 2D heterostructures stacked by strong bonding (e.g. GaN/NbN epitaxial semiconductor/superconductor heterostructures [34]), and the other is the very popular 2D heterojunctions stacked by van der Waals force (e.g. magic-angle graphene). The function of the 2D heterojunction or heterostructures can be flexibly modified according to the “embedded” design theory. They can be used to assess and reduce the difficulty encountered in the bandgap regulation of a single component [16] and the low on/off current ratio caused by the semimetal nature of single component graphene-based devices [10]. The “embedded” design theory for 2D heterojunction/heterostructures can be conveniently used to fabricate novel highly efficient electronic devices (e.g. vertical transistors formed by graphene/molybdenum ditelluride [MoTe2 ]) [10] and optoelectronic devices [13]. Therefore, the “embedded” design theory for 2D heterojunctions can have fewer components to realize the high-performance analog signal modulation, avoiding the limit of miniaturization with increasing demand on energy in the conventional metal oxide semiconductor field-effect transistor (MOSFET). A wide perspective in the telecommunication field would be developed due to their vast application in the high-performance analog circuits. There are currently two kinds of methods that have to be developed here for the fabrication of the emerging research field of thin films, such as the 2D

1.2 Thin Films for the Innovation of Information Technology

heterojunctions, the topological insulators [35] and the topological semimetals [10]: One type is the top-down physical methods, such as the intercalation/stripping method [36] and the liquid-phase [37] or liquid–air interface assemble [38], usually relating to the progress in physical fabrication. Another type is the “bottom-up” methods, usually relating to chemistry synthesis, such as the gradually modified on-surface synthesis technology. These approaches may require the conventional epitaxial growth (e.g. molecular beam epitaxy [MBE]) process [34] or doping steps [13] to prepare some of the 2D layers. Of course, the first prerequisite for the two types of fabrication methods is the 2D materials, many of which have to use chemistry methods firstly. Among 2D materials, the 2D semiconductors, such as MoS2 or WSe2 , have great applications in the large-scale thin-film electronics. However, the synthesis of high-quality soluble-processing 2D semiconductor nanosheets, nanoribbons, or nanoflakes remains challenging. A general intercalation/stripping approach was recently developed by Duan and Huang group [36] for the preparation of highly uniform, solution-processable, phase-pure semiconducting nanosheets. This process involves the electrochemical intercalation of quaternary ammonium molecules (such as tetraheptylammonium bromide) into 2D crystals, followed by a mild sonication and exfoliation process. Phase-pure, semiconducting 2H-MoS2 nanosheets with a narrow thickness distribution can be obtained, which can be further processed into high-performance thin-film transistors. The structure and composition of the exfoliated MoS2 nanosheets obtained was characterized by this approach [36]. The scalable fabrication of large-area arrays of thin film transistors can be further fabricated, enabling the construction of functional logic gates and computational circuits, including inverter gates, NAND (neither agree nor disagree [surveys]) gates, NOR gate, AND (agree nor disagree) gates, and XOR (exclusive-OR) gates, and a logic half-adder [36]. This kind of solution-processable method can be extended to fabricate many other high-quality 2D nanosheets for large-scale electronics, such as WSe2 , Bi2 Se3 , NbSe2 , In2 Se3 , Sb2 Te3 , and black phosphorus. In addition, even the traditional method can be further developed into sophisticated technologies by combining them with some advanced instruments in thin-film fabrication. For example, the “tear and stack” technique for graphene fabrication [39] can be further developed by combination with layer exfoliation, electron-beam lithography, and reactive ion etching (RIE) to a modified “tear and stack” technique for the precise angle control in the fabrication of the stacking 2D superlattices [14]. The success of this method has made it possible to synthesize two different layers with controlled stacking angles, such as the magic-angle graphene superlattices, which provide the essential materials to study their correlated insulator behavior at half-filling in magic-angle graphene superlattices for exotic quantum phenomena [14] and to reveal their unconventional superconductivity [35]. Even though the epitaxial growth (particularly molecular beam epitaxial growth) approach is still the main “bottom-up” technology for the 2D films including semiconductors and their heterojunction and van der Waals heterojunctions [34], a new bottom-up method, or the on-surface synthesis technology, has been gradually

5

6

1 Introduction: Progress of Thin Films and Coatings

developed for large-scale synthesis of 2D heterojunction thin films at atomic precision (e.g. heterostructural nanoribbon assembled thin films) [29, 40]. This method entails using the precursors to synthesize ribbons with controlled width and edge microstructures on the oriented crystal surfaces of the metal substrates. The basic fabrication steps of the on-surface synthesis process using graphene nanoribbons (GNRs) as models to produce the prototypical armchair graphene ribbon of width N = 7 obtained using 10,10′ -dibromo-9,9′ -bianthryl as precursor are schemed [40]. The metal surfaces are usually very clean Au(111) and Ag(111) single crystals as well as 200 nm Au(111) thin films that are epitaxially grown on mica for GNR growth. Thermal sublimation of the monomers onto a solid substrate surface is firstly conducted to remove their halogen substituents, yielding the molecular building blocks of the targeted graphene ribbon in the form of surface-stabilized biradical species. During the first thermal activation step, the biradical species diffuse across the surface and undergo radical addition reactions to form linear polymer chains as imprinted by the specific chemical functionality pattern of the monomers. Then, a surface-assisted cyclodehydrogenation is performed in the second thermal activation step establishing an extended fully aromatic system. For the fabrication of straight N = 7 armchair GNRs, the substrate is maintained at 200 ∘ C during monomer deposition to induce dehalogenation and radical addition. After deposition, the sample is post-annealed at 400 ∘ C for 10 minutes to cyclodehydrogenate the polymers and form GNRs. For the chevron-type GNRs, the preparation is identical using another precursor monomer 2: 6,11-dibromo-1,2,3,4-tetraphenyltriphenylene monomer, except for the fact that slightly higher substrate temperatures of 250 and 440 ∘ C are used during monomer 2 deposition and for post-annealing, respectively. In both cases, essentially all deposited monomers are attached to the GNRs, spontaneously transforming into the desired GNR structures. This approach is versatile for the on-surface synthesis of GNRs of different widths and edges, controlled by the monomers and reaction conditions. Ultrasmall transistors – and thus the next step in the further miniaturization of electronic circuits – are the obvious application possibilities. Although they are technically challenging, electronics based on nano-transistors actually work fundamentally the same as today’s microelectronics. The Empa researchers would produce transistors with a channel cross section 1000 times smaller than those typically manufactured today from semiconducting nanoribbons. However, further possibilities are also feasible, for example, in the fields of spintronics or quantum informatics [41]. These nanoribbons can self-assemble to dense monolayered thin films via the well-established self-assembly for the further precise study of their microstructure dependent properties that some of them may be only theoretically studied years ago. Thus the on-surface synthesis technology makes it possible to fabricate the nanostructured thin films at the atomic resolution to realize the experimental study for some theoretically predicted physical effects or phenomena in atomic-scale thin films. The typical examples can refer to the different origins of energy gaps for GNRs with armchair-shaped edges and with zig-zag-shaped edges calculated by ab initio calculations according to the first-principles approach [42], and the existence of symmetry-protected topological phases, junction states, and

1.3 Thin Films for Ultrasensitive Sensing Devices

spin centers in armchair-edge graphene ribbon systems, and the interesting topological phases protected by the spatial symmetries in the chevron GNRs and cove edged GNRs [43]. The topological nontrivial and trivial study of GNRs [29] can be further extended due to the success in the precise synthesis of the graphene nanoribbons of different edge structures and controlled widths (e.g. zig-zag, armchair). Steven G. Louie advanced the armchair topological band engineering of GNRs, which classify them as topological nontrivial and topological trivial according to the difference in the width and the edge microstructure [29]. As the two GNRs of different types (e.g. nontrivial and trivial) are intersected, the interface state of topology protection (the zero-energy state) will be formed. The GNR with armchair edge preserves the width-dependent semiconducting properties, and that with zig-zag edge preserves the magnetic side. The simulated result suggests that the edge state at the same side has the same spin states and preserves the ferromagnetic property, while the spin state of the other edge state is contrary to this edge, leading to the antiferromagnetic property in the whole ribbon [42]. Besides the abovementioned progress in the fabrication of the 2D heterojunctions or heterojunctions, the liquid–air self-assemble process has been developed by Christopher B. Murray group 10 years ago [38]. The liquid–air self-assemble approach can be treated as a cost-efficient and convenient approach by arranging varieties of monodisperse nanomaterials to form heterogeneous thin films by adjusting the interface interaction between the nanomaterial solution and the air. This method straightly expanded the precision of the nanosphere Lithographie, Galvanoformung, Abformung (LIGA) process [44, 45] from several hundred nanometers to 10 nm or higher. Clearly, this approach significantly depends on the synthesis of monodisperse nanoparticles (NPs) and the matched interface between the liquid solution and air, which can possibly reach sub-10 nm resolution to assemble different nanoparticles to varieties of heterogeneous nanoparticle arrayed thin films.

1.3 Thin Films for Ultrasensitive Sensing Devices One of the main goals to develop large-scale thin-film fabrication methods and study their structure–property relationship is to manufacture ultrasensitive sensing devices. Besides the integrated circuit design and manufacture process for their application in high-performance computers, the assembling and integrating methods (e.g. microelectronic mechanics [MEMs] approach) are needed to realize their application in other fields, which can process thin films into active devices and couple with necessary outer instruments (e.g. display, data processing devices) even though “interfaces themselves are devices.” With the progress of thin-film fabrication, processing thin-film based devices into ultrasensitive sensors have evolved from the traditional MEMs to the advanced nanoelectronic mechanics (NEMs) and then to atom-scale electronic mechanics (AEMs) or single molecule electronic mechanics (details can be referred to Chapter 6). In addition, the continuously developing template-assisted nanoimprinting methods, template transfer nanoimprinting molding process [18, 20, 22, 46–49], can be conveniently

7

8

1 Introduction: Progress of Thin Films and Coatings

coupled with other advanced nanoimprinting approaches (e.g. e-beam LIGA, super-resolution lithography) due to their flexibility in some special nanostructure (e.g. nanopore, nanolace) fabrication at large area economically, which can be used in the sensing film fabrication economically for MEMs, NEMs, and AEMs for ultrasensitive electronics, optoelectronics, magnetic detectors, and other kinds of flexible sensors. One of the important applications in thin films is the flat screen for video information display. OLED are still the fastest growing technology in the display industry even though only one chapter in this book discusses one topic related to this field in this book (referring to Chapter 11) [50]. OLED is an emerging display technology that enables beautiful and efficient displays and lighting panels. OLEDs are already being used in many mobile devices and TVs, and the next generation of these panels will be flexible and bendable. One of the many benefits of OLEDs includes its manufacturability on rigid (glass) or flexible substrates. A number of technologies are required for the fabrication of OLEDs on flexible substrates or foldable OLEDs (FOLEDs). FOLEDs are OLEDs built on nonrigid substrates such as plastic or metal foil. This enhances the durability and enables conformation to certain shapes and even repeated bending, rolling, or flexing. FOLEDs, still in their infancy, will usher in a range of new design possibilities for the display and lighting industries. Imagine having a mobile phone that looks like a pen but has a bright, full-color display that rolls in and out for use, which is called the Universal Communication DeviceTM (UCD). In the future, it may be very common to open a foldable smartphone from your pocket into a tablet display, to roll up a television into your pocket, and to use conformable transparent interior lighting panels that could be unbreakable. These ideas offer, you should believe, a mere glimpse into the wonders and possibilities that FOLEDs pffer. The market can reach US$ 63 billion in 2030 predicted by IDTechEx company (https://www.idtechex.com/en/research-report/ flexible-printed-oled-displays-2020-2030-forecasts-markets-technologies/693). Simultaneously, the flexible and wearable devices [51] have also been developed rapidly together with the progress of flexible OLED technology [50, 52, 53], providing much convenience in dreaming light and acoustic tasting of new technology making new life. If we can fabricate these abovementioned 2D materials under the progress of these thin-film fabrication and device assemble technology, the invention of ultrathin flexible and foldable sensing devices and more efficient OLED displays will be of interest. In addition, thin films are always associated with nanotechnology and precise atom manipulation based on interface physics and chemistry. Introducing nanostructures or atomic structures into thin films can even sprout an old academic field. For example, nanoporous magnetoplasmonic thin films developed by Song’s group can enhance the amplitude of the Fabry–Perrot interference and tune their optical oscillating into infrared (IR) range with wavelength at least from optical range to 2600 nm due to the surface plasmon resonance-enhanced Fabry–Perrot interference in the nanoporous structures [18, 47]. They also discovered the optical cavity-enhanced magneto-optical Kerr effect with a distinct reversed magneto-optical Kerr signal [22] and further improve the Raman scattering signals

1.4 Thin Films for Sustainable Energy Application

for biosensors by careful design of pore size, depth, interspacing, and components of the multilayered nanoporous thin films [54]. Acousto-optical devices, such as modulators, filters, or deflectors, implement a simple and effective way of light modulation and signal processing techniques. However, their operation wavelengths are restricted to visible and near-infrared (NIR) frequency region due to a quadratic decrease of the efficiency of acousto-optical interaction with the wavelength increase. At the same time, almost all materials with a high value of acousto-optical figure of merit are nontransparent at wavelengths larger than 5 μm, while the transparent materials possess significantly lower acousto-optical figure of merit. Daria O. Ignatyeva’s group designed a hybrid nanostructural Otto structural thin film to overcome this limit [21]. Their results indicated that the acousto-optical light can be modulated to the mid-infrared spectral range (more than 5 μm) by the planar semiconductor structures supporting guided modes at low loss using this kind of Otto-type multilayered nanostructures (i.e. semiconductor prism/noble metal/acoustic piezoelectric). Belotelov et al. also demonstrated that magnetic field sensors based on magnetoplasmonic effects can be theoretically up to a detecting resolution of fT/Hz−1/2 if designing nanostructural thin films with a high-quality Q factor [19]. We can expect that the detecting resolution for a weak magnetic field can be further increased, possibly up to aT/Hz−1/2 if other physics fields (Fabry–Perrot interference, electric field, localized surface plasmon resonance induced near field) or effects (inversed Hall effect, spin-orbital coupling effect), can be coupled into the magneto-optical effects. Clearly, success of these devices is the result of the strength to leverage multiphysics coupling by designing suitable multilayered nanostructural thin films. In the future, the combination of nanotechnology with the multiphysics coupling may address many other limitations in the ultrasensitive sensor development by precisely constructing suitable nanostructured thin films.

1.4 Thin Films for Sustainable Energy Application The rapid development of human society enforces consumption of more and more energy and resources. It is difficult for our current science and technology to achieve zero pollution and perfect recycling, leading to eco-environment deterioration. Now, people still face resource crisis, energy crisis and ecocrisis. Thin-film technologies have been impinging into these fields to address some issues in these crises. Facing the energy crisis, the sustainable and/or zero pollution energy and the related technology including energy-saving and recovery technology are desired. Solar energy, geothermal energy, and ocean energy are promising sustainable primary energy, and hydrogen energy is one of the real zero-pollution potable energy. These technologies can simultaneously address the eco-crisis because less wastes and environmentally harmful gases will be discharged into our ecosystem. The escalating energy demand for efficiently running the modern society requires the exploration of clean alternative or sustainable energy and the development of new and effective materials for energy conversion, storage, recovery, and

9

10

1 Introduction: Progress of Thin Films and Coatings

transfer. In the past decades, new kinds of solar cells based on thin films and coatings (e.g. multilayer heterostructure semiconductor coating cells, thin-film cells, and perovskite coating cells), thermoelectric (TE) thin films, and the related proton-exchange membranes (PEMs) for hydrogen fuel cells (FCs) have been developed intensively. In addition, lots of 2D materials (e.g. graphene) have been used in the fabrication of high-performance film-based electrodes for new batteries, such as the scaffold or interfacial layer for lithium-metal anodes and the cathode for Li/Na–O2 batteries in the past decades [55]. The thermoelectric (TE) process is fundamentally a microscopic one involving the transport and exchange of energy by and between electrons and lattice vibrations or phonons, in a solid (i.e. TE materials). The TE process is generally based on the oldest basic thermoelectric phenomena of the Peltier and Seebeck effects [56, 57]. TE generators are fabricated based on the Seebeck effect or an electrical potential being generated when there is a temperature difference across a piece of TE materials (e.g. Bi2 Te3 [58], PbTe [59], and SiGe [60]). Thermoelectric power generation has been significantly pioneered by NASA to overcome the energy issue faced by some special missions, such as the interstellar mission of Voyager-1 launched in 1977. Solar cells and other energy sources are infeasible in this kind of interstellar mission because of the extended distance from the sun or operating on densely shielded planets. Thus, the radioisotope thermoelectric power generator (RTG) was invented, by which the heat generated by the radioactive decay of a 238 Pu source can be converted to the electrical power by a thermoelectric generator (http://voyager .jpl.nasa.gov/index.html). Since then, many other NASA missions have been powered by RTGs. The most famous example is the Mars Science Laboratory, “Curiosity” rover (http://mars.nasa.gov/msl). Thermoelectric generators have been currently much actively being pursued and used as the main power supply by the National Aeronautics and Space Administration of many countries in the aeronautic project and the astronautics exploration. Due to the capability of directly converting heat into electricity by the utilization of small temperature difference, TE materials have also shown great potential to generate electricity particularly from the industrial waste heat, geothermal source, ocean thermal gradient, and body thermal in the past decades. These heat sources are previously much difficult to be utilized by traditional energy transfer technology. They are promising materials for power generation from environmentally friendly sources, thus reducing our dependence on fossil fuels and the associated risk of a future energy crisis. Inversely, the Peltier effect can provide all-solid-state heat pumping under electrical activation for climate control, such as Peltier coolers and portable refrigerators. Peltier coolers are often implemented in a charge-coupled device (CCD) detectors or in the heat sinks of microprocessors for reliability and precise temperature control. Due to small size and the absence of moving parts and hence no vibrations, TE niches and thin films can be used to make portable picnic refrigerators and flexible thermal or electric devices in cooling microelectronics (e.g. CPU chips) and optoelectronics (e.g. infrared detectors, laser diodes) [61]. Most microelectronics and optoelectronics devices require responsive small-scale or

1.4 Thin Films for Sustainable Energy Application

localized spot cooling that does not impose a large heat load, which is best satisfied by the thermoelectric refrigeration or coolers [62]. Terrestrial applications of thermoelectric power generation and all-solid-state heat pumpers have become more pertinent. Their unique features of TE materials have led to surging interest in the basic research in TE materials and structure design of high efficiency in the past decade, such as the band structure engineering and the nanoscale effect in the electron crystal and phonon glass [63–67]. The study has promoted the mature of the TE materials design and found lots of novel TE materials, such as varieties of nanocomposites (e.g. lead-antimony-silver-telluride family: PbTe–AgSbTe2 ; Pb-chalcogenide nanocomposites), PbTe1−x Sex , YB66 , CeFe4 Sb12 , tetrahedrites (Cu3 SbSe4 ), MgSb2 , La, and Fe- or Co-doped Ca–Co–O misfit-layered cobaltites, and filled skutterudite antimonides [67–70]. Recent progress in new thermoelectric materials has paved the thin-film power generators of TE materials, which can be referred to in Chapter 8 of this book for details. Up to now, TE applications cover a wide spectrum of heat sources, from utilizing environmentally friendly heat sources to generate electricity, to using body heat for portable electronics, and to using thin-film cooler/generator. The stringent requirements on the Seebeck coefficient, electrical conductivity, and thermal conductivity that are necessary to produce good thermoelectric behavior continue to offer challenges and inspiration to myriad researchers. Thermoelectricity has become a proving ground for numerous new and innovative concepts in physics, chemistry, and materials science, including quantum size effects, nanostructuring, phonon-glass–electron crystal behavior, and large anharmonicity. New knowledge uncovered in these and other areas is driving improvements in material performance, and the future holds the promise that highly efficient thermoelectric devices for both power generation from waste heat and solid-state climate control will become a reality [71]. Given the progress in the understanding of TE, we still want to mention that no single technology can meet the world’s energy needs in the twenty-first century. One needs a combination of many technologies and novel materials into thin films, such as the solar-thermoelectric hybrid power generator of full band range by using TE materials and semiconductors together [72]. Solar energy may be the most popular sustainable clean energy on the Earth. Photovoltaic (PV) technology, such as solar cells harvesting solar energy directly into electricity, has been developed rapidly in the past decades. Besides the semiconductor multilayered thin-film solar cells, the perovskite thin-film solar cells are another viable competitor to the commercially available silicon-based solar cells. Apart from low-cost, simple device processing and manufacturability combined compatibility with roll-to-roll processing and fabrication on flexible substrates add to the merits of the perovskite PV technology by comparing with III–V semiconductor [73–79]. The term “perovskite” was attributed to the crystal structure of calcium titanate (CaTiO3 ), which was discovered by the German mineralogist Gustav Rose in 1839 and named in honor of the Russian mineralogist Lev Perovski [80]. The perovskite solar cells (PSCs) present numerous advantages include unique electronic structure, bandgap tunability, superior charge transport properties, facile processing, and low cost [73, 79]. PSCs have demonstrated unprecedented progress in efficiency and its

11

12

1 Introduction: Progress of Thin Films and Coatings

architecture evolved over the period of the last decade since 2009, leading to the advent of new low-cost PV technology. They can be easily fabricated on the flexible substrates (conductive polymers) by the traditional ink-coating technology and conveniently coupled with other technologies and materials (e.g. graphene, semiconductors, ferroelectrics, Si chip) for enhanced performance. Their power conversion efficiency (PCE) has achieved a high PCE of about 22% in 2016 [79], serving as a promising candidate with the potential to replace the existing commercial PV technologies. This breakthrough led to the so-called “perovskite fever” [81], attracting much research interest in the following years, eventually increasing the efficiency to a record 22.1% (National Renewable Energy Laboratory [NREL]) in early 2016 [79]. PV technology is a multidisciplinary and versatile field in which lots of novel advanced technologies can be coupled with varieties of the related targeting materials, the progress of fundamental research, and even the traditional technologies very well, finding their own strong points and/or creating novel pinpoints to address the issues in solar cells. Much more advanced design methods (e.g. tandem solar cells [TSCs], multi-junction cells such as that formed by semiconductors and perovskites) and fabrication technologies have been invented for the development of more and more types of perovskites (e.g. metal halide perovskite, perovskite–copper indium gallium selenide [CIGS]) of high photoelectric effects after 2016. In March 2018, Jinsong Huang group [82] developed a molecular (i.e. 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquino dimethane: F4TCNQ) additive-assisted strategy for p-type molecular doping of solution-bladed perovskite films (i.e. methylammonium lead iodide: MAPbI3 ) to address the issue of the mismatched energy band between the indium tin oxide (ITO) and the perovskite (Figure 1.1a). Their final fabricated F4TCNQ-doped films exhibiting increased electrical conductivity, especially at grain boundary regions, and increased charge carrier concentrations (Figure 1.1b–j). Molecular doping of perovskite film by F4TCNQ led to the considerable enhancement of PV performance from 11.0% to 20.2% [82]. The simple but effective approach enabled the scalable fabrication of HTL-free PSC devices with a simplified device geometry using the convenient and economical Doctor blading and doping fabrication technique (Figure 1.1). As one emerging PV technology rooted from the Earth-rich element compounds and the economical and simple solution casting fabrication techniques, PSCs are expected to realize the economic electricity generation. It plays one of the main roles to settle the energy crisis and the ecological deterioration once and for all by fully utilizing solar energy. However, their stability and large-scale fabrication are two obstacles to commercialization. In September 2018, Hongwei Han’s and Edward H. Sargent’s group summarized the progress of PSCs again and pinpointed the challenges for their commercializing [83]. The power transfer efficiency and stability have been gradually increased after developing various device configurations (including mesoscopic, planar, triple mesoscopic, and tandem structures) and lots of highly efficient materials systems in the past few years [83]. The highest laboratory photoelectric conversion efficiency of PSCs notarized by a third party reached 23.3% in Sept. 2018, which has exceeded the commercial polycrystalline silicon solar cells, the CdTe and CIGS thin-film solar cells, exhibiting the commercialization

1.4 Thin Films for Sustainable Energy Application

F4TCNQ

Blade

Perovskite film

Perovskite ink ITO

N

F

F

N

N

F

F

N

(b)

Height

(a)

500 nm

(c)

2 μm

nm 200

nm 200

nm 200

0.00

0.00

0.00

–200

(d)

8 μm

mV 40.0

–200

(e)

10 μm

–200

mV 40.0

mV 40.0

CPD

Diffused region 0.00

(f)

2 μm

–40.0

0.00

0.00

(g)

8 μm

–40.0

(h)

10 μm

–40.0

–4.0 2500

MAPbl 3 F4TCNQ solid-diffused MAPbl3 F4TCNQ-doped MAPbl3

2000 –5.2

1500 1000

F4TCNQ

500 0 –0.10

(i)

MAPbl3

Number of points

3000

–0.05 0.00 Potential (V)

0.05

(j)

–5.5

–8.3

Figure 1.1 Doctor blading and doping of perovskite films by F4TCNQ. (a) Schematic illustration of a doctor-bladed perovskite film and the chemical structure of the F4TCNQ dopant. (b) Cross-sectional SEM image of the MAPbI3 film deposited directly onto ITO glass via bladed coating at 150 ∘ C, showing the film thickness of around 500 nm. Topography KPFM images of (c, f) MAPbI3 , (d, g) F4TCNQ-doped MAPbI3 , and (e, h) F4TCNQ solid-diffused MAPbI3 films. CPD represents the contact potential difference between the tip and the sample’s surface. (i) Surface potential profiles of different perovskite films as indicated. (j) Schematic illustration of the energy diagram and electron transfer process for MAPbI3 :F4TCNQ blends. F4TCNQ: 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquino dimethane; MAPbI3 : methylammonium lead iodide; ITO: indium tin oxide. Source: Wu et al. [82]. © 2018, Springer Nature Limited.

perspective and great potential market value. As for the stability of devices, there are varieties of PSCs showing no distinct efficiency damping after the 10 000 hours’ standard testing under the required testing conditions (e.g. high temperature, continuous illumination, high humidity) using the simulated sunlight. A 110 m2 perovskite PV system with printable triple mesoscopic PSC modules (3600 cm2 for each) has been launched by Wonder Solar in China. According to their configurations and coating components for light harvesting, these solar cells can be classified as single-junction GaAs cells, multi-junction cells

13

1 Introduction: Progress of Thin Films and Coatings

(two-terminal, monolithic), crystalline silicon cells, thin-film PVs technologies, and emerging PVs as reported from the National Center for Photovoltaics (NCPV) at NREL of the USA (https://www.nrel.gov/pv/cell-efficiency.html). In addition, there are two key top layers in the multi-junction cells: III–V semiconductor top coatings and perovskite top coatings. Particularly, the marriage to other advanced materials (e.g. 2D materials) and the optimized structure design theory have greatly improved the solar energy transfer efficiency and long lifetime after hastening varieties of thin-film solar cells, such as the double-junction (noncondensing) thin-film solar cells and perovskite–silicon stack cells, perovskite-based TSCs, and organic–inorganic solar cells [73, 74, 79]. After three years’ accumulation in the theoretical and experimental study and the application transformation since 2016, the PV technology experienced a blowout development after 2019 that is indeed a great year for solar cells. There are so many achievements that deserve our consideration, which can encourage us to have more interest in this vivid field. Some of them will be briefly described as following. Wei Huang group highlighted the current status and recent advances in perovskite-based TSCs, including perovskite–silicon, perovskite–perovskite, and perovskite–CIGS integrations. Figure 1.2 describes the solar efficiency growth of the three kinds of perovskite-based solar cells in the past 4 years [73], among which the perovskite–silicon can reach up to 27.1% near to the theoretical limit or the Shockley–Quiesser limit of 33.16%. Thereby, more attention has to be paid to TSCs that are suggested as an alternative to beat the efficiency limit since a maximum efficiency of 42% can be reached if using two subcells with bandgaps of 1.9 eV/1.0 eV, opening up a great potential to

33

Perovskite single-junction theoretical limit 33.16% 27.1% Solliance

28 26 23.9% ANU

25.5% HZB

24 Efficiency (%)

14

23.1% UT

22

22.43% UCLA 21.0% UT

20 18 2T Perovskite–Si

16

4T Perovskite–Si 2T Perovskite–perovskite

14

4T Perovskite–perovskite 2T Perovskite–CIGS

12

4T Perovskite–CIGS

10

2015

2016

2017 Year

2018

2019

Figure 1.2 Progress in the efficiency of the perovskite single-junction solar cells in the past five years. CIGS: copper indium gallium selenide. Source: Hu et al. [73].

1.4 Thin Films for Sustainable Energy Application

develop perovskite-based TSCs for commercialization [73]. This kind of TSCs can serve as a promising candidate with the potential to replace the existing commercial PV technologies in the near future [73], as shown in the progress in the early of the following year (2020) (https://www.nrel.gov/pv/cell-efficiency.html). Dion Jacobson (DJ) two-dimensional halide perovskites have been attracting much attention. However, the barrier thickness of the related quantum well (QW) is not well-known even though it is key in the design of highly efficient DJ PSCs. Wei Huang and Yonghua Chen’s group designed the efficient DJ thickness by regulating the orientation and uniform dispersion of DJ perovskites via the chain length of large-volume organic ammonium spacers [84]. It was found that the DJ perovskite could have the suitable QW barrier thickness, excellent orientation, and more uniformly dispersed QW if using 1.3-propanediamine (PDA) and 1.4-butanediamine (BDA), leading to smooth bandgap transition, longer carrier diffusion length, higher charger mobility, and lower defect density [84]. Finally, they successfully fabricated DJ-type PV cells with efficiency up to 16.38% using BDA [84]. It is a substantial challenge to prevent the degradation of metal PSCs by humid air to improve their stability and lifetime for their future commercialization. Michael Grätzel’s group invented an intercalation method to improve their stability in humid air by forming the 3D/2D bilayer perovskite. This method entails introducing a 2D A2 PbI4 perovskite layer by employing pentafluorophenylethyl ammonium (FEA) as a fluoroarene cation inserted between the three-dimensional (3D) light-harvesting perovskite film and the hole-transporting material (HTM) [85]. The perfluorinated benzene moiety confers an ultrahydrophobic character to the spacer layer, protecting the perovskite light-harvesting material from ambient moisture and reducing the mitigating ionic diffusion in the device. The 2D layer simultaneously enhances interfacial hole extraction to suppress nonradiative carrier recombination, enabling a PCE up to 22.09%. Surprisingly, their unsealed 3D/2D PSCs can retain 90% of their efficiency during PV operation for 1000 hours in humid air under simulated sunlight. Rosei and coworkers [86] reported the integrated effects of carbon quantum dots (C-dots) in the fabrication of high-efficiency inverted plane heterojunctions (PHJ) PSCs by using C-dots to modify the hole-transporting layer in the plane PSCs. The PSC efficiency can be up to 16.2% as introducing C-dots onto oxide graphene layers as the hole-transporting layer, whose efficiency can reach 16.8% under the UV range if using C-dots as the downshift layer. The introduction of C-dots can also extract the hole and transfer it to the conductive substrate and delay the charge recombination, leading to the enhanced stability of PSCs. Energy loss in the hybrid lead halide perovskite cells is interrelated with the non-radiation combination in the interface and the perovskite layer. Liao and coworkers [87] developed a simple but efficient strategy to reduce this loss via the coupling of the external electric field with the intrinsic doping of ferroelectric polymer as the interface polar layer. This strategy entails doping a series of polar ferroelectric (PFE) polymers into the methylammonium lead iodide (MAPbI3 ) layer and/or inserting them between the perovskite and the hole-transporting layers to modify and/or enhance the build-in field (BIF), improve the crystallization

15

16

1 Introduction: Progress of Thin Films and Coatings

of MAPbI3 , and regulate the non-radiative recombination in PSCs. Doped PFE polymers can enable MA+ orderly arranged to reduce trap states and optimize the oriented growth of the perovskite layer. At the same time, inserting PFE into the gap between the perovskite layer and the hole-transporting layer can enhance the BIF via the widened depletion region in the cell. The consequently assembled cells exhibit an open-circuit voltage of 1.14 V and a PCE of 21.38% [87]. Cost is always one main issue for the commercialization of PSCs. For reducing the materials cost, Mohammad Khaja Nazeeruddin’s group [88] developed the synthesis of three enamine HTMs based on TrçgerQs base scaffold. The best performing material of HTM3 demonstrated 18.62% PCE in PSCs and presented a markedly superior long-term stability in nonencapsulated devices [88]. Moreover, the high glass-transition temperature (176 ∘ C) of HTM3 also suggests promising perspectives in high-temperature device applications of PSCs. An organic solar cell (OSC) is another promising PV device, particularly for the flexible potable power providers. Ternary OSCs also show great potential to enhance the PV property of single-junction OSCs. Liu et al. [89] have prepared a series of ternary OSCs using the previously developed PM7: ITC-2Cl [89–91] as the main system, the ultralow-bandgap receptor of IXIC-4Cl as the ternary component. The active absorbance layer can be up to 1000 nm and the PCE of 15.37% with only 0.42 eV energy loss [89]. It is an efficient method to improve the power transfer efficiency for OSCs by intensifying the light absorbance through enhancing the intramolecular push–pull effect of PV materials. However, as for the electron acceptors, the design strategy for halide molecular usually decreases the molecule energy level, leading to the reduced open-circuit voltage. Yao et al. [92] designed and synthesized a kind of chloride non-fullerene acceptor, showing extended optical absorbance and higher voltage than the fluoride compounds. The chloride non-fullerene acceptor can modify the short-circuit photocurrent density and the open-circuit voltage, realizing the PCE of 16.5%. This result indicated that reducing the band gad voltage offset can dramatically improve the power transfer efficiency by precisely tuning the organic PV materials, suggesting the promise of fullerene-free OSCs in practical applications. The III–V semiconductor solar cells are also one promising type of PV devices because of their high photoelectric efficiency, power density, and stability. However, their expensive manufacturing cost severely impedes their commercialization, partially due to the expensive Ge-based substrates for the epitaxial growth. Oh and coworkers [93] proposed a germanium-on-nothing (GON) technology to fabricate ultrathin Ge films for lightweight and thin GaAs solar cells. As shown in Figure 1.3, the ultrathin epitaxial single crystalline Ge membrane can be formed as the reusable substrate by utilizing the reorganization of cylindrical pores of porous Ge films during hydrogen annealing enable the growth and transfer of GaAs cells. Compared with previous porous Ge studies, the surface quality of reformed Ge can be significantly improved by engineering the initial pore morphology and surface passivation before annealing. The GaAs cells growing on the reformed Ge can have an efficiency of 14.44%, much near to GaAs cells growing on the bulk Ge substrates (16.53%). Their open-circuit voltage is almost the same as those GaAs cells growing

1.4 Thin Films for Sustainable Energy Application

Growth and liftoff of a GaAs cell

Ge film Void Ge bulk 4𝞵m Germanium on nothing

Current densuty (mA/cm–2)

Pore transformation

30 25

GaAs on Ge bulk ~16.53%

GaAs on 20 Ge on nothing ~14.44% 15 10 5 0 0.0

GaAs cell

Ge on nothing 0.2

0.4

0.6

0.8

1.0

Voltage(V)

Figure 1.3 The germanium-on-nothing (GON) technology to fabricate ultrathin Ge films and the GaAs cell, and the voltage-dependent current density curve. Source: Park et al. [93]. Copyright from Elsevier.

on the bulk Ge substrates. This GON technology can reduce the materials cost distinctly of III–V solar cells. The requirement for flexible electronic and optoelectronic devices provokes the development of flexible energy sources. The flexible composite solar cells have been paid more and more attention recently. With the application of graphene, CVD graphene and the sponge-like 3D reduced graphene oxide (rGO) have been used as a sole HTM in the PSCs, exhibiting excellent performance due to their good deformation features. Cai and Yu in their article on the graphene application summarized this kind of graphene-based flexible PSCs [94]. Figure 1.4 gives the layered structure of this kind of graphene-based flexible PSCs (a,b), V–C curves at different bending radius, and their PCE% and short-circuit currents change with the bending times, showing good cell performance under various bending angles and cycles. These Gp-FPSCs shall have great applications in the flexible folded-potable or wearable electronic and optoelectronic devices [95]. This exciting rapid development of thin-film PSCs continues with the coming of 2020; more and more amazing signs of progress have been achieved. Sargent’s group precursor complexes to prepare the metal halide perovskite [96] designed PbI2−x x nanoflakes (PNPLs) with more uniform multi-quantum well distribution based on the idea that iodine-based PNPL permitted wide-band absorbance of sunlight, leading to large Stokes displacement to overcome the bandgap limited absorbance range of the conventional Br based perovskites. Using these I-based nanoflakes,

17

1 Introduction: Progress of Thin Films and Coatings

Ag PCBM ite Perovsk

r = 0.670 cm

r = 0.365 cm

r = 0.175 cm

r = 0.130 cm

P3HT TM OAT

ZEOC

PET

(a)

(c)

(b) 12

4

–4 –8 –12 –16 –20 0.0

0.2

0.4 0.6 Voltage (V)

12

9

Bending radius: Flat 0.670 cm 0.365 cm 0.175 cm 0.130 cm

PCE (%)

0

0.8

Degradation: 14%

6

6

3

3

0

1.0

0 (d)

9

100

200 300 400 500 Bending times

JSC (mA/cm2)

Current density (mA/cm2)

18

0

Figure 1.4 (a) Scheme of layered structures of graphene-based flexible PSCs; (b) pictures of the flexible solar cell units at different bending radius; (c) the C–V curves of the flexible solar cells at different bending radius; (d) the power conversion efficiency (PCE%) change with the bending times (black line) and the short-circuit current (red line). Source: Cai and Yu [94].

they fabricated the luminous solar concentrator (LSC) to realize the highly efficient luminescence. The photoluminescence quantum yield (PLQY) of this kind of thin films can reach up to 56%, and their light scattering efficiency is only 2.0%, which is 1.3 times of the best 10 × 10 cm LSC fabricated under room temperature. Cost is forever one of the key issues for large-scale application of PSCs. Efficient electron transport layer-free perovskite solar cells (ETL-free PSCs) due to their cost-effective and simplified design, highe efficiency, and potential compatibility greatly demonstrates the large-area flexible application of PSCs. However, the absence of ETL usually results in the mismatched interface energy level between the ITO layer and perovskite layer, limiting the charge transfer and collection, which leads to significant energy loss and low device performance [97]. Ge’s group invented a method to lower the work function of ITO and optimize the interface energy level alignment by virtue of an inherent dipole by introducing a polar nonconjugated small-molecule modifier to address this issue. The formed barrier-free ITO/perovskite interface is a benefit for the efficient charge transfer and restraining nonradiative recombination, which endows the device with enhanced open-circuit voltage, short-circuit current density, and fill factor. Consequently, the PCE of the modified device can reach 20.55%, much higher than 12.81% of

1.4 Thin Films for Sustainable Energy Application

the ITO-based device, and comparable to state-of-the-art PSCs with an ETL [97]. Moreover, the stability is enhanced with decreased hysteresis effect due to interface defect passivation and inhibited interface charge accumulation. The key advantages of the present device design are the high conversion efficiency potential with simple device structure and the fact that the whole device production process can be carried out at economical and energy-efficient temperatures. This work facilitates the further development of highly efficient, flexible, and recyclable ETL-free PSCs with simplified design and low cost by interface electronic structure engineering through facile electrode modification. After 60 years of research, the PCE of Si solar cells is approaching the Auger recombination-constrained Shockley−Queisser limit of 29.8% [98, 99]. To further increase the PCE while simultaneously reducing the cost per kWh, new strategies such as tandem configurations have been developed in the past decades. Organometal-halide perovskite/Si TSCs were proposed as one promising candidate to surpass Si efficiency records. A TSC consists of two or more cells that are optically coupled by absorbing different parts of the incident spectrum. This allows for a more efficient conversion of the broad-band solar spectrum into electric power. In a two-cell configuration, the high-energy region of the spectrum is absorbed by the top cell, whereas the transmitted low-energy light is further absorbed by the bottom cell. Hybrid organic–inorganic perovskite-based cells are especially well suited as a top cell for Si-based TSCs due to their high charge carrier mobility, high quantum yield, long diffusion length, sharp absorption edge, and large tunable bandgap covering almost the entire solar spectrum. Several analyses on the limiting efficiency of TSCs have been performed using detailed-balance calculations showing efficiencies up to 69.9% for an infinite number of subcells under 1 sun illumination [100–102]. However, the efficiency of perovskite/Si TSCs is strongly affected by spectral and temperature changes. Consequently, weather conditions at the specific site of deployment should be taken into account when designing perovskite/Si TSCs. Futscher and Ehrler have theoretically suggested in 2016 that perovskite/Si TSCs with PCE limits above 41% are possible for all three tandem configuration even at nonideal climate conditions by using a perovskite top cell with the ideal bandgap for the respective tandem configuration [102]. In addition, it is challenging to monolithically process PSCs directly onto the micrometer-sized texturing on the front surface of record-high-efficiency amorphous/crystalline silicon heterojunction (SC) cells, which limits both high temperature and solution processing of the top cells. The challenge for solar cell design is that both perovskite and Si are finicky. It is well-known that graphene is a 2D material with unique and powerful electronic properties. Graphene is notoriously difficult to work with, and perovskites have a durability issue that needs to be factored into the design. Nevertheless, researchers have been tinkering around with the graphene–perovskite combo to address these issues by utilization of their advantages and suppressing their shortcoming by all in one strategy to fabricate graphene/perovskite/silicon heterojunctions (SCs). To tackle these hurdles, Di Carlo and coworkers developed a mechanically stacked two-terminal perovskite/silicon TSC, with the subcells independently fabricated, optimized, and subsequently coupled by contacting the back electrode of the

19

20

1 Introduction: Progress of Thin Films and Coatings

Glass/FTO cTiO2 mTiO2 Perovskite Spiro-OMeTAD or PTAA ITO

Mechanically stacked perovskite/silicon tandemsolar cell

Perovskite top cell Silicon bottom cell

Front contact +metal grid HJT Si cell Stabilized PCE = 25.9%

Figure 1.5 Scheme of the microstructure and the graphene-doped mesoporous electron selective layer mechanically stacked into the perovskite top cell and the silicon bottom solar cell forming the perovskite/graphene/silicon heterojunction tandem SCs. Source: Lamana et al. [103].

mesoscopic perovskite top cell with the texturized and metalized front contact of the silicon bottom cell. The structure and the graphene layer bonding to the heterojunction are schemed in Figure 1.5 [103]. By minimizing optical losses, as achieved by engineering the hole selective layer/rear contact structure, and using a graphene-doped mesoporous electron selective layer (the middle layer between the two layers as shown in the right image of Figure 1.5), the perovskite top cell reaches better electrical performance by graphene doping of the electron selective layer [103]. This heterojunction microstructure design endows a solar cell around those twin challenges and achieved an impressive solar conversion efficiency of 26.3% (25.9% stabilized) over an active area of 1.43 cm2 , the best solar cell ever of great potential industrialization. More amazing progress in solar cells is coming in early 2020. On Jan 27, 2020, NREL of the USA announced that the PCE% of single-junction perovskite–silicon stack solar cell invented by Berlin Institute of materials, at Helmholtz (HZB, Germany), has reached 29.15%, exceeding the record of 28% previously reported by Oxford PV company (https://www.nrel.gov/pv/cell-efficiency.html). At the same time, groups from Stanford University and Arizona State University declared that the PCE% of their single-junction PSCs was 25.3%, up to the world record kept by Massachusetts Institute of Technology and Korea Institute of Chemical Technology. Finally, and soon, NREL obtained a PCE% of 32.9% in their Double junction (noncondensing) thin-film solar cells. The PCE% of the single-junction perovskite–silicon stack solar cells by HZB was further confirmed and authorized by Fraunhofer Institute of Solar Systems (Germany), updated the record in the PCE% chart of varieties of solar cells (Figure 1.6) by NREL (https://www.nrel.gov/ pv/cell-efficiency.html; https://www.nrel.gov/pv/assets/pdfs/best-research-cellefficiencies.20200128.pdf; https://www.helmholtz-berlin.de/pubbin/news_seite? nid=21020;sprache=en;seitenid=1). In this design, a special electrode contact layer

1.4 Thin Films for Sustainable Energy Application

52

Best research-cell efficiencies

48 44 40

Cell efficiency (%)

36 32 28 24 20 16 12 8 4 0

1975

1980

1985

1990

1995

2000

2005

2010

2015

2020

Figure 1.6 Updated solar cell efficiency by NREL in February 2020 for multi-junction cells, single-junction GaAs cells, crystalline Si cells, cells by thin-film technologies, and cells by emerging photovoltaic technologies. Safe and stable energy storage devices of high energy density at low cost are also essential for our routine transportation and other industrial applications. Varieties of rechargeable supercapacitors and high-performance and high-volume batteries have been developed, such as Li-S batteries, AIBs, and graphene-based supercapacitors [55, 104–107]. The recently developed Li/Na-S batteries and supercapacitors may be the most promising energy storages satisfying these requirements for the next-generation rechargeable energy storage with the rapid coupling of 2D thin-film materials into this field [55, 104]. Metal–sulfur batteries hold practical promise for next-generation batteries because of high energy density and low cost. Development is impeded at present, however, because of unsatisfied discharge capacity and stability in long cycling. Combination of experimental and theoretical approaches can be used to develop insight into the relationship between electrochemical behavior of sulfur redox and metal stripping-plating and the structural properties of electrode materials. With metal–sulfur batteries, two-dimensional (2D) thin-film nanomaterials are a suitable model with which to connect and test experimental results with theoretical predictions and to explore structure–property relationships. Through the view of combining experimental and theoretical approaches, sulfur redox conversion on 2D nanomaterials in various reaction stages was explored, and crucial factors affecting 2D nanomaterials as artificial solid electrolyte interfaces (SEIs) and host materials in protecting Li and Na metal anodes were critically unveiled by Shi-Zhang Qiao [55]. It is indicated that Li/Na-sulfur batteries hold practical promise for next-generation batteries because of high energy density and low cost. Significant progress has been made in understanding mechanisms of sulfur redox and metal stripping/plating with a judicious combination of experimental and theoretical approaches. Two-dimensional (2D) nanomaterials offer a suitable model to correlate experimental results with theoretical predictions and, importantly, with which to explore structure–property relationships. Future research effort should focus on the establishment of correlations between macroscopic conversion kinetics and the electronic structure of the electrode materials with agreed standards and advanced combined experiments and theory. In addition, fabrication of three-dimensional (3D) electrodes from 2D materials might be a promising approach to promote the energy and power densities of the Li/Na-sulfur batteries and other metal–sulfur batteries. Source: https://www.nrel.gov/pv/cell-efficiency .html; https://www.nrel.gov/pv/assets/pdfs/ best-research-cell-efficiencies.20200128.pdf.

21

22

1 Introduction: Progress of Thin Films and Coatings

was developed for the TSCs, and then, the interface layer was modified. Thus, the top perovskite layer can utilize the optical range of sunlight, and the bottom Si layer of a special SiO2 interlayer can transfer most of the IR-NIR light into electric energy, leading to the enhanced PCE% comparing to each single cells. This kind of perovskite/graphene/silicon heterojunction tandem SCs have been fabricated into solar cell modules using the industrial scale fabrication and sealing process for the three kinds of pre-commercialization life testing. They have successfully survived the light resistance test, the damp heat resistance test, and the thermal cycling test on 23 January 2020, which is the first time for the PSCs to realize this goal for commercialization. The PSC modules can be either rigid or flexible and either transparent or semitransparent, which makes them suitable for varieties of application scenarios, such as assembling in windows, roof tiles, external wall faces, roads, sound baffles, and automobile roofs. They can also be colorful via imitation coloring by varieties of structure colors made from the nano/microfabrication. As it is well-known that PSCs and their modules preserve high-efficiency, low-cost processing ability and cheap and rich sources. These marvelous and exciting achievements make PSCs full of perspective for sooner commercialization and serving renewable energy for our routine life. Rechargeable aluminum-ion batteries (AIBs) are regarded as one promising candidate for post-lithium energy storage systems (ESSs) [107]. For addressing the critical issues in the current liquid AIB systems, here, a flexible solid-state AIB is established using a gel-polymer electrolyte (a kind of electric thin films) for achieving robust electrode–electrolyte interfaces. As shown in Figure 1.7, employment of polymeric electrolytes mainly focuses on addressing the essential problems in the liquid AIBs to transfer into the solid AIBs, including unstable internal interfaces induced by mechanical deformation and production of gases as well as unfavorable separators, which is much different from the utilization of solid-state systems for alleviating the safety issues and enhancing energy density in lithium-ion batteries [107]. Particularly, such gel electrolyte enables the solid-state AIBs to present an ultrafast charge capability within 10 seconds at a current density of 600 mA/g. Meanwhile, an impressive specific capacity ≈120 mA/h/g is obtained at a current density of 60 mA/g, approaching the theoretical limit of graphite-based AIBs. In addition to the well-retained electrochemical performance below the ice point, the solid-state AIBs also hold great stability and safety under various critical conditions. The results suggest that such a new prototype of solid-state AIBs with robust electrode–electrolyte interfaces promises a novel strategy for fabricating stable and safe flexible ESSs. Li storage is one key issue for the fabrication of high-performance Li batteries with high volumetric energy density. The carbon allotropes are mainly used as the host materials for reversible lithium uptake in Li-ion batteries [108, 109], thereby laying the foundations for existing and future electrochemical energy storage. Recently, the Li storage in layered 2D materials, such as graphene, has been developed for enhanced Li storage density. However, insight into how lithium is arranged within these hosts is difficult to obtain from a working system because

1.4 Thin Films for Sustainable Energy Application

Figure 1.7 Overall comparison of liquid-state aluminum-ion batteries (AIBs) and solid-state AIBs: (a) Schemes for the configurations of these two AIBs. (b) Schemes for demonstrating the electrode–electrolyte interfaces: an unstable interface based on porous separator in the liquid-state AIBs and robust interface based on the GPE electrolyte in the solid-state AIBs. (c) Schemes for illustrating the production of gases in the two prototypes of batteries. Source: Yu et al. [107].

(a)



v

+

– AlCl4-

v

+ GPE

Al2Cl7Cation Anion

Liquid-state AlBs

(b) Unstable interface

Porous textile separator

(c) Pronounced bubbles

Solid-state AlBs Robust interface

Compacted gel electrolyte Suppressed bubbles

the in situ high-resolution transmission electron microscopy (HR-TEM) [110–112] is not suitable to probe those light element, especially lithium (Li) with atomic number less than carbon due to their low scattering cross section for impinging electrons and their susceptibility to knock-on damage [113–115]. This unknown structure–property relation hinders the design and fabrication of more dense Li storage for high-performance Li batteries and the related supercapacitors. Jurgen H. Smet’s and Ute Kaiser’s groups [105] realize the study of the reversible intercalation of lithium into bilayer graphene by in situ low-voltage transmission electron microscopy, using both spherical and chromatic aberration correction to enhance contrast and resolution to the required levels [116]. It is found that lithium atoms can form a multilayered close-packed order between the two carbon sheets [105]. Consequently, the lithium storage capacity associated with this superdense phase far exceeds that expected from the formation of LiC6 , which is the densest configuration known under normal conditions for lithium intercalation within bulk graphitic carbon [117]. This finding shall insight the design and fabrication of more dense Li storage for future high-performance Li batteries. Supercapacitors are another energy storage mode besides the batteries. Thanks to their ultrafast rechargeable ability, high power, long-term lifetime, wide working temperature, and safety, supercapacitors have shown extraordinary promise for potable miniaturized electronics, hybrid electric vehicles, standby power, and field power. However, the power density (i.e. power in unit volume) of the electrochemical supercapacitors is relatively lower than the batteries, which restricts their broader applications, particularly in the potable smart equipment. The power density of supercapacitors needs to be further improved but is usually limited by

23

24

1 Introduction: Progress of Thin Films and Coatings

electrodes with rather low volumetric performance, which is largely due to the inefficient utilization of pores in charge storage. Benefited from the progress of the magic materials: graphene and some graphene-based technologies have been invented to address this issue. Recently, a freestanding graphene laminate film electrode with highly efficient pore utilization was developed for the compact capacitive energy storage by precisely tuning the interlayer spacing of graphene laminate films for optimized porosity [118]. The preparation process is as follows [118]: The mixture made from graphene oxide (GO) and thermal expansion reduced graphene (EG) of different ratios were firstly prepared and then vacuum filtrated to form the composite graphene thin film of tunable interlayer spacing, whose layered microstructures are characterized by scanning electronic microscope (SEM). The porosity of the electrode materials can be optimized by regulating the interlayer spacing. By systematically tailoring the pore size matching for the electrolyte ions, utilization of pores and their interspacing can be optimized, and thereby, the volumetric capacitance will be maximized. Consequently, flexible all-solid-state supercapacitors can be fabricated, which can deliver delivers a stack volumetric energy density of 88.1 Wh/l in an ionic liquid electrolyte, representing a critical breakthrough for optimizing the porosity toward compact energy storage [118]. Moreover, the optimized film electrodes exhibit excellent bending ability due to the intrinsic flexibility of graphene [118]. They can further be assembled into ionogel-based, all-solid-state, flexible smart devices with multiple optional outputs and superior stability, demonstrating enormous potential as a portable assembled to smart devices, realizing varieties of output results by designing the suitable circuits [118]. Hierarchically ordered structures with low tortuosity, excellent mechanical flexibility, high optical transparency, and outstanding electrical conductivity are critically important in developing flexible transparent supercapacitor electrodes for innovative applications in electronics and displays. Bionic technology has been recently immersed in the design and fabrication of supercapacitors with the abovementioned microstructures and properties [106]. The leaf-skeleton inspired electrodes have been successfully fabricated by a CVD process, which are reticulated monolithic networks consisting of carbon nanostructures serving as a 3D spongy core and graphene-based films as a protective/conductive shell [106]. The network electrodes show optical transmittance of 85–88%, an electrical sheet resistance of ∼1.8 Ω/sq, and an areal capacitance of 7.06 mF/cm2 (at 0.78 mA/cm2 in a three-electrode cell) in Na2 SO4 aqueous electrolyte. Flexible transparent and symmetric supercapacitors, based on poly(vinyl alcohol) (PVA)/H3 PO4 gel and the network electrodes, possess a stable working voltage of 1.6 V, energy and power density of 0.068 μWh/cm2 and 47.08 μW/cm2 at an optical transparency of ∼80%, and no capacitance loss over 30 000 flat-bend-release cycles. Similar to energy storage and utilization of renewable energy source, energy conversion technology is also intimately dependent on thin films and coatings. Energy conversion using electrochemical reactions for fuel cells has attracted increasing attention because of its advantages over traditional fossil energy sources, such as renewability, eco-friendliness, and high efficiency, one of whose key components is

1.4 Thin Films for Sustainable Energy Application

catalytic thin-film based electrodes or catalytic functional PEMs [119, 120]. Among the fuel cell technology, the PEM fuel cells (FCs) may be one of the promising models to determine the factors that influence the commercialization for common transportation, which is based on the highly efficient catalyst filling PEM (i.e. Pt catalyst filling porous Nafion thin films) as the cathodes [121]. PEMs of high electrochemical catalytic performance are particularly desired for the transportation of routine life using zero-waste-releasing potable fuels with high energy density (i.e. hydrogen) if considering the ecological issues as burning fuels of different kinds [122, 123]. Hydrogen fuel cell automobiles have become one popular routine transporter in Japan as Japan has started the FH2R (Fukushima Hydrogen Energy Research Field) model project since 2018, which is just completed at the end of February 2020. The FH2R intends to produce 1200 Nm3 /hr via a 10 MW water electrolysis system powered by a 20 MW PV device equipped on the 180 000 m2 field reported from the New Energy and Industrial Technology Development Organization of Japan in March 2020. This model project realizes the perfect coupling between the potable hydrogen fuel cell technology and renewable solar cell technology. The oxygen reduction reaction (ORR) is a critical factor associated with electrochemical energy conversion in fuel cells. It is an important cathode reaction in many electrochemical energy conversion devices, including hydrogen fuel cells and direct methanol fuel cells [122, 123]. The main difficulty associated with the ORR is the sluggish multiple-electron transfer process, which has to be catalyzed typically by those precious metals such as Pt, Pd, Ru, or their alloys [124–127]. This issue impedes the future common worldwide application. Therefore, the reduction of Pt or precious metal load in the cathode for PEM fuel cells but still keeping high power density is highly needed, particularly reducing the Pt content in the cathode thin films [128–133]. However, the low Pt content in the cathodes will result in the high voltage losses that come from the mass transport resistance of O2 through the platinum–ionomer interface in the PEM, the location of the Pt particle with respect to the carbon support and the supports’ structures. Strasser and Orfanidi group recently proposed a new Pt catalyst/support design that substantially reduces local oxygen-related mass transport resistance [130]. This new design includes the use of chemically modified carbon supports with tailored porosity enabling controlled deposition of Pt nanoparticles on the outer and inner surface of the support particles, resulting in unprecedented uniform coverage of the ionomer over the high-surface-area carbon support thin films, especially under dry operating conditions. Consequently, the present catalyst design exhibits previously unachieved fuel cell power densities in addition to high stability under voltage cycling. Owing to the Coulombic interaction between the ionomer and N groups on the carbon support thin films, homogeneous ionomer distribution and reproducibility during the ink manufacturing process for the catalytic polymer thin films can be ensured. Another alternative is to develop non-precious catalysts, such as some transition metal (TM) complexes, single-atom catalysts (SACs), or single-site catalysts (SSCs) [121, 134]. TM complexes have been widely used in physical and biological science, particularly playing essential roles in catalysis, chemical synthesis, materials science, photophysics, and bioinorganic chemistry [135–137]. Since 1964, N4

25

26

1 Introduction: Progress of Thin Films and Coatings

macrocycles of non-noble metals (e.g. cobalt) based on the organometallic complexes and bionics have been developed as fuel cell cathode catalysts [138]. Various non-precious earth-abundant metal catalysts, especially those based on transition metal–nitrogen–carbon (MNC) compounds, have been continuously developed to address the issues related to the cost and earth-abundant resources, and their catalytic performance has been gradually enhanced [139–142]. However, these no precious catalyst-based fuel cells are currently suffering from limited activity, poor stability, poor durability, limited ability for a large-scale yield of high-performance catalysts, impeding their sustainable commercial application by replacing those precious metal catalyst-based fuel cells [128, 143]. Recently, Song’s group presents a simple sequenced ultrasonic atomization microreaction, pyrolysis, acid leaching, and calcination process for the mass synthesis of FeNC SSCs with excellent ORR catalytic performance, as shown in Figure 1.8A [121]. The ultrasonic atomization process provides a huge number of microreactors to produce the highly dispersed iron precursors. The subsequent pyrolysis and calcination processes ensure that the iron atoms are conjugated to nitrogen ligands, anchoring them onto carbon black. This synthesis system is eco-friendly by only using water and ethanol as solvents and does not generate pollutant emissions. A combination of various microstructure and composition characterization of these catalysts and the related ORR performance based on these catalysts filling proton-exchange Nafion membrane suggest that the active centers root in the single-atom Fe sites chelating to the fourfold pyridinic N atoms or calling as SSCs. Their electrochemical catalytic ORR performance outperforms the commercial Pt/C catalysts, having much enhanced halfwave potential and kinetic current density (Figure 1.8B) and the substantially enhanced long-term stability and outstanding tolerance to methanol (Figure 1.8B). This synthetic strategy provides a new general method for the eco-friendly large-scale synthesis of high-performance single-atom catalysts for fuel cells. In addition, even though the large-scale production of hydrogen has recently been realized, the electrocatalysis for water splitting reaction, including hydrogen evolution reaction (HER) and oxygen evolution reaction (OER), is still depending on the catalytic proton-exchange Nafion thin films filled with Pt or other precious-metal-based catalysts. Therefore, similar to those catalysts for the catalytic thin films in ORR, reducing the precious metal loading in catalytic thin films or developing non-precious-metal-based catalysts for HER and OER are desired to further reduce the price of hydrogen as the economic portable fuel [127, 144, 145]. The late strategy is most desired since those precious metals particularly for Pt are nonrecyclable Earth-deficient sources. In particular, for both the OER and HER in electrolysis, it is necessary to develop non-precious, efficient, and durable catalysts. Considering the cost and yield efficiency, it is greatly desired for highly efficient bifunctional catalysts for overall water splitting in an alkaline medium. The transition metal oxides, metal sulfides, metal phosphides, layered bimetallic hydroxide (LDH), and metal alloys have been investigated as the catalysts for OER and HER, whose catalytic performance can be improved through morphology engineering, composite, and doping technology. Particularly, FeNi alloy and their

1.4 Thin Films for Sustainable Energy Application

(B) (b) FeNC FeNC SACs Pt/C

–2

30 mV

–3 –4 –5

Carbon black

Pyrolysis

8

0.86

5.13

0.85

6 4

0.84

0

0.82 FeNC

FeNC FeNC SACs Pt/C

60 40

100 80 60 Methanol injection 40 20

FeNC SACs Pt/C

0 0

30

60

90 120 150 180 Z' (Ω)

69 mV/dec 0.90

0.88

(f)

0 0

100

200 Time (s)

400

63 mV/dec

76 mV/dec

–0.4

–0.2

0.0

0.2

0.

Log[j(mA/cm2)] 100

93.0%

90 80

71.1%

70 60 50 40

300

FeNC FeNC SACs Pt/C

0.92

0.86 –0.6

Pt/C

FeNC SACs

(e)

20

FeNC

0.87

(c)0.94

10

2 0.8 0.4 0.6 1.0 Potential versus. RHE (V)

(d) 100

–Z" (Ω)

Calcination FeNC SACs

E12 J1

–6

80

HCI leaching

7.60

0.84

–7 0.2

FeCI2

12 10.44

0.88

Jk (mA/cm2) Potential versus. RHE (V)

0 –1

Relative current (%)

KOH

(a)

E1/2 (V)

FeNC precursor

Curremt density (mA/cm2)

Phenanthroline

Relative current (%)

(A) Ultrasonic nebulizer

30 0

61.0% FeNC FeNC SACs Pt/C 10 000

20 000

30 000

Time (s)

Figure 1.8 (A) Schematic of the preparation of FeNC SACs by ultrasonic atomization in conjunction with pyrolysis and calcination. (B) Evaluation of electrochemical performance on FeNC, FeNC SACs, and 20% Pt/C catalysts. (a) Linear sweep voltammetry (LSV) curves, (b) comparison of halfwave potential (E1/2 ) and kinetic current density (Jk ), (c) Tafel plots, (d) Nyquist plots obtained by electrochemical impedance spectroscopy (EIS), (e) relative current density–time curves of FeNC SACs and 20 wt% Pt/C catalysts at 0.6 V with methanol injection, and (f) relative current density–time curves at 0.6 V for 36 000 seconds (10 hours). Source: Ma et al. [121].

LDH preserve OER and HER bifunctions in the same reaction medium, having been studied intensively. It was also found that CoFe LDH, CoMn LDH, CuCoO nanowires, FeCoOH, [email protected], and NiMo nanorods have OER and HER bifunction. Recently, Inamdar et al. developed another robust non-precious copper–iron (CuFe) bimetallic composite that can be used as a highly efficient bifunctional catalyst for overall water splitting in an alkaline medium. Their catalyst exhibits outstanding OER and HER activity, and very low OER and HER overpotentials (218 and 158 mV, respectively) are necessary to attain a current density of 10 mA/cm2 [144]. When used in a two-electrode water electrolyzer system for overall water splitting, it not only achieves high durability (even at a very high current density of 100 mA/cm2 ) but also reduces the potential required to split water into oxygen and hydrogen at 10 mA/cm2 to 1.64 V for 100 hours of continuous operation. Many 2D materials, such as MoS2 , have been studied for catalyzing the HER of water, showing great promise as a cost-effective alternative to Pt even though the current catalytic efficiency is still worse than that of Pt. [146, 147] Cao and coworkers recently report a strategy to enable the catalytic activity of monolayer MoS2 films that are even better than that of Pt via engineering the interface interaction of the monolayer with supporting substrates [146]. The monolayer films were grown with CVD processes and controlled to have an optimal density (7–10%) of sulfur vacancies. They found that the catalytic activity of MoS2 could be affected by substrates in two ways: forming an interfacial tunneling barrier with MoS2 and modifying the chemical nature of MoS2 via charge transfer. Thus, excellent catalytic activities at the monolayer MoS2 films can be obtained by using substrates that can provide n-doping to MoS2 and form low interfacial tunneling barriers with MoS2 (e.g. Ti). The catalytic performance may be further boosted to be even better than Pt by crumpling the films on Ti-coated flexible polymer substrates, as the Tafel slope of the film is substantially decreased with the presence of crumpling-induced compressive strain.

27

28

1 Introduction: Progress of Thin Films and Coatings

The monolayer MoS2 films show no degradation in catalytic performance after being continuously tested for over two months. Another kind of thin-film electrode for water splitting, which can be a potential candidate to replace the precious metal catalyst-based electrode is based on TM-based nanocatalysts. They have been attracting great attention due to their low cost and abundance as compared with those precious-metal-based catalysts, but their low performance (particularly insufficient activity) still remains a challenge. Many strategies (e.g. size, shape, hierarchical structure, and composition control) have been developed to improve their performance in the past few years. One example of the progress in these TM-based nanocatalysts is the 4 nm Mn3 O4 nanoparticles (NPs) developed recently by Nam and coworkers [148]. It is confirmed that the overall increase in the active surface area is distinctly related to the superb catalytic activity of the 4 nm Mn3 O4 NPs by comparing them with those 8 nm species. To further enhance the OER performance, the interface interaction between the catalysts and the support should be optimized. As expected, introducing of Ni foam substrate can indeed maximize the entire number of the NPs participating in OER. An outstanding electrocatalytic activity for OER was obtained using the 4 nm Mn3 O4 /Ni foam electrode, with an overpotential of 395 mV at a current density of 10 mA/cm2 under neutral conditions (0.5 M phosphate buffer saline (PBS), pH 7).

1.5 Thin Films and Coatings for Key Sources and Ecological Environment of Earth Water is the first key source for all organisms on Earth [149, 150]. Even though 71% of Earth’s surface is covered by ocean, only 2.53% of water is fresh water that is not the uniform distribution in the earth. Many areas are very deficient of fresh water, such as the Sahara and many countries in Africa, the Middle-East areas, California and Nevada of the USA, North and Northwest District of China, and most isolated islands. Water has played a key role in social and national security in the history of human beings [149, 150]. In order to resolve the freshwater shortage, governments have invested lots in many huge projects, such as the partly-finished South-to-North Water Diversion Project in China. While the USA has paid attention to desalination technology, to obtain fresh water from the ocean or brackish water has been on the watch by the USA government to address the water shortage in California since the 1950s [151–153]. The reverse osmosis technology based on highly efficient permeability membranes (e.g. reverse osmosis thin-film composite membrane) emerges as the times require insight from the semipermeability of animal bladders [153]. Dr. Song has systemically summarized the progress of permeable membranes in his dissertation in 2000 [153]. With the rapid development of new separation concept, novel membrane materials science, advanced fabrication technology, and automatic control engineering in the past two decades, great progress has been achieved in membrane science and separation technology [151, 154–162]. Because of varieties of advantages of select osmotic membrane technologies (e.g. the high selectivity [based on molecule weight, geometry, affinity, and configuration] and

1.5 Thin Films and Coatings for Key Sources and Ecological Environment of Earth

permeability, the possibility non-phase transformation or controlled phase change during separation, noncontact and temperature difference between products and feeds, the possibility for nonthermal exchange process, and low operation cost), lots of permeable membranes or related separation technologies have been invented for the separation, purification, and concentration of desired products in almost all industry involving the fluids (gas or liquid), particularly foods, pharmacy, gas separation, chemical engineering, petroleum refining, chemistry synthesis, environment protection, and waste recycling and reutilization [149, 152, 154–165]. When using a membrane to separate materials, the efficiency of the separation is limited by how fast the gas or liquid passes through the membrane and by how selective it is. Thinner membranes usually allow for faster flow rates but are usually less selective and strength. In order to increase the flux without loss of selectivity, many technologies and membrane materials have been invented in the past decade. The most impressive progress may be the incorporation of the magic two-dimensional (2D) materials [166–171] and/or varieties of carbon allotropes (i.e. porous carbon, graphene, carbon nanotube) into the thin-film composite membranes that have greatly addressed the trade-off between selectivity and permeability, one of the main issues in the membrane materials development [149, 154, 156, 164, 172, 173]. Particularly, graphene – with great mechanical strength, chemical stability, and inherent impermeability – offers a unique 2D system with which to realize this membrane and study the mass transport. One sophisticated perforating strategy to maintain the selectivity without losing permeability was successfully developed by Celebi et al. [174], which entails the precisely drilling holes of controlled diameter in a graphene sheet about two layers thick, up to a few million pores with narrowly distributed diameters between less than 10 nm and 1 μm. For such a thin membrane, the primary barriers to separation come from entrance and exit from the holes and not from the motion through the membrane, which can have highly efficient mass transfer across physically perforated double-layer graphene but still maintain a high selectivity. The measured transport rates are in agreement with predictions of 2D transport theories. Attributed to its atomic thicknesses, these porous graphene membranes show higher permeance of gas, liquid, and water vapor far in excess of those shown by finite-thickness membranes, highlighting the ultimate permeation these 2D membranes can provide. Varieties of graphene-based (including GO: partially oxidized and stacked sheets of graphene [175]) thin-film composite membranes have been developed, which can possibly address this issue and has reached the separation resolution of molecule/ion sieving ability in the selectivity [149, 156, 164, 172, 173]. Two-dimensional materials such as graphene and graphene oxide membranes (GOMs) [175], MoS2 , can provide ultrathin, high-flux, and energy-efficient membranes possible with ångström-scale channels with atomically flat walls for precise ionic and/or molecular sieving in aqueous solutions [166, 169–171, 176, 177] and gas-phase (e.g. H2 or He) separation [167, 168, 178]. These materials have been made into varieties of thin films and coating of different microstructures and layers, showing potential in a variety of applications, including water desalination and purification [179–181] and gas and ion separation [160, 167, 168, 178, 182–184]

29

30

1 Introduction: Progress of Thin Films and Coatings

However, there are still lots of work to do in the controlled modification of 2D materials for highly efficient permeable membranes to overcome their intrinsic limitation, for example, graphene, unlike the pores of carbon nanotube membranes that have fixed sizes, while the pores of GOMs and the interlayer spacing between GO sheets are of variable size [185], and it is difficult to reduce the interlayer spacing sufficiently to exclude small ions and to maintain this spacing against the tendency of GOMs to swell when immersed into aqueous solution [186]. These challenges hinder the potential ion filtration applications of GOMs. In 2017, Fang and coworkers [160] developed a cationic control strategy to tune the interlayer spacing of GOMs with ångström (Å) precision using K+ , Na+ , Ca2+ , Li+ , or Mg2+ ions. Membrane spacings controlled by one type of cation can efficiently and selectively exclude other cations that have larger hydrated volumes. Using this strategy, they fabricated centimeter-scale GOMs on ceramic supports experimentally, achieving facile and precise control of the interlayer spacing, with a precision of down to 1 Å, and corresponding ion rejection, through the addition of one kind of cation (e.g. K+ ) [160]. This method is based on the understanding of the strong noncovalent hydrated cation–π interactions between hydrated cations and the aromatic ring, indicating that other ions could be used to produce a wider range of interlayer spacings [160]. Generally, this method provokes a step toward graphene-oxide-based thin-film applications, such as water desalination and gas purification, solvent dehydration, lithium-based batteries and supercapacitors, and molecular sieving. Freshwater flux and energy consumption are two important benchmarks for the membrane desalination process. Nanoporous carbon composite membranes have been developed by Sheng’s group in April 2018 [156], which comprise a layer of porous carbon fiber structures grown on a porous ceramic substrate, showing 100% desalination and a freshwater flux of 3–20 times higher than existing polymeric membranes. Thermal accounting experiments demonstrated that the carbon composite membrane could save over 80% of the latent heat consumption. Conventional technology for the purification of organic solvents requires massive energy consumption, as well as to reduce such expending calls for efficient filtration membranes capable of high retention of large molecular solutes and high permeance for solvents. In September 2019, Tang’s group [161] reported a surface-initiated polymerization strategy through C–C coupling reactions for preparing conjugated microporous polymer (CMP) membranes of high resistance to organic solvents due to the all-rigid conjugated systems in the backbone of the membranes. The prepared 42-nm-thick CMP membranes supported on polyacrylonitrile substrates exhibited excellent retention of solutes and broad-spectrum nanofiltration in both nonpolar hexane and polar methanol, the permeance for which reaches 32 and 22 l/m2 /h/bar, respectively. Besides this issue, antifouling may forever be one of the main issues during the long-term operation, particularly biofouling, which will result in the flux reduction, the deterioration of selectivity, and/or the broken of membranes, finally leading to a short lifetime of membranes [163, 187]. Many novel methods have been invented in resolving the membrane fouling problem [163, 187].

1.5 Thin Films and Coatings for Key Sources and Ecological Environment of Earth

Another interesting innovative technology in clean water production using membrane separation technology should be the bionic-plant-leaf-inspired sunlight-driven purifier for high-efficiency clean water production [155]. It is well-known that the transpiration and guttation process of natural vascular plant leaves can produce tons of clean water via osmotic pressure differences powered by sunlight. Inspired by this transpiration and guttation process, a sunlight-driven purifier is designed for high-efficiency water purification and production. This sunlight driven purifier is constructed by a negative temperature response poly(N-isopropylacrylamide) (PNIPAm) (PN) hydrogel anchored onto a superhydrophilic melamine foam skeleton, and a layer of PNIPAm modified graphene (PG) filter membrane coated outside. Molecular dynamics simulation and experimental results show that the superhydrophilicity of the relatively rigid melamine skeleton significantly accelerates the swelling/deswelling rate of the poly (isopropylacrylamide) (PNIPAm) (PN) and PNIPAm modified graphene (PG) -filter (PNPG-F) purifier. Under one sun, this rational engineered structure offers a collection of 4.2 kg/m2 /h and an ionic rejection of >99% for a single poly (N-isopropylacrylamide) (PNIPAm) (PN) and PNIPAm modified graphene (PG) (PNPG) filter from brine feed via the cooperation of transpiration and guttation. It is envisioned that such a high-efficiency sunlight-driven system could have great potential applications in diverse water treatments. Besides the freshwater resource crisis, another great challenge before people may be the deterioration of our ecology and environment, particularly the atmospheric pollution that is mainly resulted from more and more toxic gas releasing and greenhouse gas emissions. One strategy to reduce the pollution of these gases is to resource them, such as reducing CO2 into ethylene or CO into methanol. Lots of resource technologies have been developed in the past decades. The electrochemical reduction reaction of CO2 (CO2 RR) may be the most promising method to fulfill this issue by transferring CO2 into chemical resources (e.g. urea, merlon, ethylene) or synfuels (e.g. methane, methanol) [188–191]. The key to this technology is to find the catalysts of high efficiency and low cost to replace the expensive precious-metal-based species. Carbon-based solid-state catalyst materials containing small amounts of nitrogen and TM have emerged as a selective and cost-efficient alternative to noble metal catalysts for the direct electrochemical reduction of CO2 into CO, formic acid, or methane. Recently, Ana Sofia Varela and Peter Strasser summarized the recent progress of MNC catalysts for the CO2 RR [190]. There is a growing interest in MNC materials as catalysts for the CO2 RR, given their remarkably high activities and selectivity toward CO formation, their ability to form “beyond CO” hydrocarbons, and their affordable synthesis methods. Studies have shown that the metal center plays a crucial role in determining the catalytic performance of the material. For instance, Fe has been shown to produce CO selectively at low overpotentials and to have the ability to reduce CO further to traces of CH4 [192], whereas Ni-containing materials have been reported as highly selective toward CO production. They have mentioned that the density functional theory-guided experimental studies could provide the elucidation of key experimental parameters and molecular descriptors. The catalytic performance in

31

32

1 Introduction: Progress of Thin Films and Coatings

the presence of different metals can be rationalized based on the binding energy between the metal center and the reaction intermediates, which might promote the activity and selectivity of the MNC catalysts for future highly efficient nonprecious catalyst development. In this review, it is further pinpointed that there are still several key scientific issues that deserved to be deeply studied in this emerging field. One is the nature of the active sites. Understanding the role of all the possible active sites will be crucial for designing optimal catalysts containing highly active and selective catalytic sites. One needs new instruments for this active site study. The second issue is how to optimize the reaction conditions since they play a major role in the catalytic process besides the catalyst structure and composition. For instance, working in acidic pH favors the competing process of the HER reducing the selectivity toward the CO2 RR. Therefore, it would be desired to work in neutral or alkaline pH values to have a highly selective process. Another important factor to consider is the accessibility of the active sites. When working with a highly active catalyst, the CO2 RR can be limited by the transport of CO2 to the active sites. In this regard, the structure of carbon support plays a crucial role in the transport of CO2 through the catalyst layer. The low CO2 solubility, however, is also a major reason for mass transport limitations. Therefore, in general, we have to determine the reaction conditions that are practical for the catalytic screening based on their industrial application. For an instance, the use of gas diffusion electrodes (GDEs) can help to overcome the pH limitations by having the gas stream and electrolyte stream separated by the electrode, which preserve a more realistic assessment of the catalytic performance of a given material for the CO2 RR at a large scale if the screening is carried out using GDE electrolyzers.

1.6 Thin Films and Coatings for Biomedical Engineering and Life Science Polymers and biomass form an integral part of our existence and everything that surrounds us – from the basic building blocks of life constituting of proteins, nucleic acids, and polysaccharides to the commercial products obtained from automobile, construction and transportation industries, plastic toys and tools, reading glasses, etc. Most of these materials are composed of a combination of one or more materials to form polymer composites. And a film made of a polymer as a matrix or a carrier is referred to as a polymer-based film. Due to its excellent composite properties and processing diversity, polymer films have been increasingly employed in production and are gradually applied to defense, transportation, aerospace, marine engineering, and other fields (e.g. thin films for artificial intelligence). Many functional polymers, particularly biodegradable species, have been fabricated into thin films and coatings of special surface patterns as biosensors or other kinds of functions (antibacterial, antifouling) for biomedical engineering and life science [187, 193–196]. In this book, the polymer-based films for artificial intelligence, selective permeable thin films and their applications for water purification and wastewater treatment, biomass-derived functional films, and anti-marine corrosion coatings

1.6 Thin Films and Coatings for Biomedical Engineering and Life Science

have been summarized in details in Chapters 13, 14, 15, and 19, respectively. Here, we will briefly introduce some other interesting progress of polymeric thin films and coatings for biomedical engineering and life science. One of them is the thin films formed via self-assembly of biocompatible or biodegradable hyperbranched polymers (HBPs), which have lots of cytomimetic applications, summarized by Jin et al. [197] The HBPs have demonstrated great potential to be used as model membranes to mimic cellular behaviors, such as fusion, fission, and cell aggregation via self-assembly into varieties of sizes, shapes, and structures, such as honeycomb structures. Natural honeycomb structures are usually observed in plants or beehives with a columnar and hexagonal array of hollow cells formed between thin vertical walls. Now, many synthetic honeycomb films can be prepared through the self-assembly of homopolymers, linear block copolymers, and star copolymers according to a breath-figure technique. The amphiphilic HBPs could self-assemble into honeycomb-like microporous films by the slow evaporation of a chloroform solution of the precursors in a humid atmosphere. The pore diameter could be controlled easily by adjusting the casting volume, polymer concentration, molecular weight, and so forth. Another interesting progress is the relation between the polymers and nanoparticles for biomedicines, such as the antibacterial dressing [193], the formation the surface patterning of nanoparticles with polymer patches as biosensors [198], and artificial electronic skins [199]. Patterning of colloidal particles with chemically or topographically distinct surface domains (patches) has attracted intense research interest [198]. Surface-patterned particles act as colloidal analogs of atoms and molecules, serve as model systems in studies of phase transitions in liquid systems, behave as “colloidal surfactants,” and function as templates for the synthesis of hybrid particles. Although the fabrication of micrometer- and submicrometer-sized patchy colloids has been matured, it is still difficult to prepare the patched surface patterns from inorganic nanoparticles of tens of nanometers. These inorganic nanoparticles exhibit size- and shape-dependent optical, electronic, and magnetic properties, and their assemblies show new collective properties [200]. Nanoparticle patterning is usually limited to the generation of two-patch nanoparticles [201–203] and nanoparticles with surface ripples [204] or a “raspberry” [205]surface morphology. Choueiri et al. invented a method to precisely prepare nanoparticle surface patterning utilizing thermodynamically driven segregation of polymer ligands from a uniform polymer brush into surface-pinned micelles following a change in solvent quality [198]. Patch formation is reversible but can be permanently preserved using a photocrosslinking step, which would suppress nanoparticle assembly and enable the utilization of solutions with a higher nanoparticle concentration, thereby increasing the yield of patchy nanoparticles. This methodology offers the ability to control the dimensions of patches, their spatial distribution, and the number of patches per nanoparticle. These patchy nanocolloids have potential applications in fundamental research, the self-assembly of nanomaterials, diagnostics, sensing, and colloidal stabilization [198]. Patterning of multicomponent nanoparticles and the self-assembly of patterned nanoparticles into complex, hierarchical structures can be further explored by some other kinds of surface modification methods, such as

33

34

1 Introduction: Progress of Thin Films and Coatings

surface grafting. Furthermore, due to the progress in the synthesis of nanoparticles of different sizes, shapes, and structures, the proposed strategy should enable fundamental studies of polymer segregation on surfaces with large curvatures or surfaces with multiple curvatures. Silk protein is one of the promising biodegradable and biocompatible materials as substrates for electronic devices in artificial intelligence, such as on-skin and implantable electronic devices. However, its intrinsic brittleness and poor thermal stability limit its applications. Recently, robust and heat-resistant silk fibroin composite membranes (SFCMs) are synthesized by mesoscopic doping of regenerated silk fibroin via the strong interactions between silk fibroin (SF) and polyurethane invented by Guo and coworkers [199]. The schemed process to fabricate the integrated protein-based electronic skin (PBES) via SFCMs is illustrated in Figure 1.9 [199]. Some traditional micromachining techniques, such as inkjet printing, can be used to print flexible circuits on such protein substrates (Figure 1.9a). The obtained SFCMs can endure the tensile test more than 200% and the thermal resistance up to 160 ∘ C. Based on this substrate, Ag nanofibers (NFs) and Pt NFs networks have been successfully embedded onto both sides of the SFCMs as heaters and

PU

SF solution Silkworm cocoon

Degumming

Dialysis

Modification

Film formation

(a) Ag

Ag NFs networks Pt Network transferring Tearing tap

(b)

Pt NFs networks Temperature sensor (Pt NFs) Printing electrode (Ag) SFCM

PBES

Heater (Ag NFs)

(c)

Figure 1.9 Fabrication of PBES based on SFCM. (a) Schematic illustration of the preparation of flexible transparent SFCM; (b) schematic illustration of the fabrication of PBES realizing heating and temperature detection; (c) photographic images of the PBES attaching to the human neck and hand closely. Source: Huang et al. [199].

1.6 Thin Films and Coatings for Biomedical Engineering and Life Science

temperature sensors, respectively (Figure 1.9b,c). The integrated PBES exhibits high thermal stability and temperature sensitivity (0.205%/∘ C). Heating and temperature distribution detection are realized by array-type PBES, contributing to potential applications in dredging the blood vessel for alleviating arthritis (Figure 1.9c). This PBES is both inflammation-free and air-permeable, which can directly be laminated onto human skin for long-term thermal management. Single-molecule detection is vital for basic research and practical applications in nanobiotechnology and nanomedicine [206]. Fabrication techniques that yield sensors with repeatable performance are critical to the sensitive detection and precise control of biomolecules [194, 207]. One way is to fabricate bio-wells as small as possible to only save one or several biomolecules [194]; another way to exert control on a biomolecule is to use guiding structures to impose a change in shape or behavior on the molecule of interest [194, 206, 207]. Nanopore or nanocavity thin films provide nanoscale channels or femtoliter volume, which are now well established as possible single-biomolecule sensors label-free or not, which hold great promise as sensing elements in diagnostic and gene sequencing applications [194, 206, 207]. Nanopore technology is particularly attractive for DNA sequencing because of its potential advantages, which include long reads and high speed [206]. Single nanopores can be used for highly sensitive detection of DNA by threading the polymer through the pore, blocking some of the electrical current that normally would pass through the pore. Usually, the nanopore or cavity thin films can be fabricated by traditional e-beam and RIE processes [194, 206, 207]. However, this promise has been limited by the expensive, labor-intensive, and low-yield methods used to fabricate low-noise and precisely sized pores or cavities [207]. Nanoscale preconfinement of DNA is shown to reduce the variation of passage times through solid-state nanopores. Preconfinement is previously achieved by forming a femtoliter-sized cavity capped with a highly porous layer of nanoporous silicon nitride (NPN). This cavity is formed by sealing an NPN nanofilter membrane against a substrate chip using water vapor delamination. However, this method of fabrication cannot keep a consistent spacing between the filter and solid-state nanopore due to thermal fluctuations and wrinkles in the membrane, nor can it be fabricated on thousands of individual devices reliably. To overcome these issues, McGrath and coworkers advanced a new process to fabricate the femtoliter cavity monolithically using a selective xenon difluoride (XeF2 ) etch to hollow out a polysilicon (poly-Si) spacer sandwiched between silicon nitride (SiNx ) layers, as schemed in Figure 1.10 [194]. These monolithically fabricated cavities behave identically to their counterparts formed by vapor delamination, exhibiting similar translocation passage time variation reduction and folding suppression of DNA without requiring extensive manual assembly. The ability to form nanocavity sensors with nanometer-scale precision and to reliably manufacture them at scale using batch wafer processing techniques will find numerous applications, including motion control of polymers for single-molecule detection applications, filtering of dirty samples prior to nanopore detection, and simple fabrication of single-molecule nanobioreactors.

35

36

1 Introduction: Progress of Thin Films and Coatings RTP (b)

(a)

Lithography (c)

Wet etch (d)

NPN

RIE

Photoresist

SiO2 pnc-Si SiNx polySi Si (e)

(f)

XeF2 (g)

Sensing membrane

EDP

Nanopore drilling

Figure 1.10 Monolithically created femtoliter cavity process flow. (a) pnc-Si is formed as a hard mask over the thin-film stack with RTP and (b) lithographically patterned. (c) The exposed capping SiO2 layer is removed using an ammonium bifluoride wet etch. (d) The nanopore pattern is transferred into the prefilter layer using an RIE process (SF6 + O2 ), forming NPN in the desired cavity geometry. (e) After an EDP etch to create access to the backside of the device, (f) the femtoliter cavity is formed through the use of XeF2 , which enters through the NPN and hollows out the poly-Si layer between the NPN and the sensing SiNx membrane (15 nm). The remaining pnc-Si and SiO2 are also concurrently removed. (g) Finally, the cavity is wetted, and then, a sensing nanopore is formed in the sensing membrane layer via CBD. Source: Madejski et al. [194].

To address the problem for the nanopore membrane fabrication, Goto et al. have been developing a controlled dielectric breakdown process to enable rapid nanopore fabrication since 2016, and now, this process can offer an upgraded method to fabricate nanopore membranes for DNA-sequencing technology [206, 208]. Recently, Waugh et al. proposed another low-cost and scalable solid-state nanopore fabrication method, termed controlled breakdown (CBD), which is rapidly becoming the method of choice for fabricating solid-state nanopores. Since its initial development, nanopore research groups around the world have applied and adapted the CBD method in a variety of ways, with varying levels of success that present their accumulated knowledge of nanopore fabrication by CBD, including a detailed description of the instrumentation, software, and procedures required to reliably fabricate low-noise and precisely sized solid-state nanopores with a yield of >85% in less than one hour. The general platform for this method is illustrated by Waugh et al. [207]. Unlike traditional beam-based nanopore fabrication technologies, the methods presented here are low-cost and low technical barriers for the fabrication of nanoscale pores in thin solid-state membranes, which is accessible for non-experts. In addition, Song and coworkers, also developed an alternative method to fabricate nanopore membranes of controlled sizes, shapes, and substrates, named as template transfer nanoimprinting [18, 20, 22, 47, 209], which will be summarized in Chapters

1.6 Thin Films and Coatings for Biomedical Engineering and Life Science

10 and 11 of this book. We believed that industrial applications that take advantage of this sensing modality include DNA sequencing DNA barcoding and investigating single-molecule capture and transport shall be realized soon with the progress of the marriage of the nanofabrication technique to the biomedical engineering. The third promising progress of thin films or coatings for biomedical and life science exists in some functional thin films formed by 2D materials, such as graphene. As a new nanomaterial, graphene has shown great promise in drug delivery, cancer therapy, and other nano-biomedical techniques due to its unique microstructures and mechanical and electronic properties. Graphene has been used in the fabrication of other kinds of permeable membranes with enhanced permeability [210], such as reverse osmosis membranes that can preserve great potential in the purification of wastewater and soft water produced by desalination of ocean water for life science, which were summarized in Chapter 14 of this book comprehensively with other kinds of permeable membranes. Recently, one of the promising methods based on the electric properties of GO composites is to introduce into the separation membrane to control the water permeability using an electric field, which is very important in the healthcare technologies related to the controlled water permeability in capillaries or membranes (e.g. artificial skins) [211]. Previous attempts to control water permeation through membranes (mainly polymeric ones) have concentrated on modulating the structure of the membrane and the physicochemical properties of its surface by varying the pH, temperature, or ionic strength [212, 213]. Electrical control over water transport is an attractive alternative. Many micrometer-thick GOMs have been invented for ultrafast permeation of water [178, 214]and molecular sieving [169, 215], with the potential for industrial-scale production. To achieve electrical control over water permeation, Nair and coworkers created conductive filaments in the GOMs via controllable electrical breakdown in 2018 [211]. The scheme of how to fabricate the E-controlled GO composite membranes was given as follows [211]. The GO multilayered membranes were first deposited in a well formed by a polymer molding method supported on a porous silver substrate [211]. The metal/GO/metal sandwich structures were then formed by depositing a thin (≈ 10 nm) gold (Au) film on top of the GOM prepared on the porous silver (Ag) substrate. One of the metal–GO–metal sandwich membranes was attached to the Polyethylene Terephthalate (PET) sheet. Such a thin layer of gold is sufficiently porous and shows the discontinuities and voids in a 10-nm gold thin film on a GOM. Therefore, the coated gold layer does not change the permeation properties of the membranes. However, the water permeability depended on the thickness of the gold layer and the concentration of hydrogen ions and hydroxide ions. Then, the electric field can be imposed between the silver layer and the gold layer to fulfill the E-controlled permeability testing. The electric field that concentrates around these current-carrying filaments ionizes water molecules inside graphene capillaries within the GOMs, which impedes water transport. Thus, the water permeation can be precisely controlled from ultrafast permeation to complete blocking. This work opens up an avenue for developing smart membrane technologies for artificial biological systems, tissue engineering, and filtration.

37

38

1 Introduction: Progress of Thin Films and Coatings

1.7 Thin Films and Coatings for National Defense and Homeland Security There are many sophisticated thin films and coatings used in varieties of instruments and equipment for the national defense and homeland security. In this book, we will emphasize some distinctive key thin films and coatings for aeronautics, information security, and marine voyages. The armor coatings for tanks or armored cars will not be discussed in this book. As for coatings for aeronautics, the most key coatings are the thermal barrier coatings (TBCs) and radar stealth coatings. TBCs are a kind of ceramic layers grown or deposited on the superalloy substrates constructing the turbine blades in the “heart” of aircraft or the jet engine in the propulsion system. They function as the thermal and corrosion protection layers of the superalloy substrates in the combustion chamber of jet engines free from high corrosion and oxidizing gas spraying at high temperatures (∼2000 ∘ C or higher) and high speed (sometimes several times of sound velocity). Readers can refer to Chapter 19 of this book for detailed microstructures and fabrication processes of varieties of TBCs. In order to evade the detection by the hostile radar system, stealth coatings for electromagnetic (EM) waves have to be equipped on the top surface of air vehicles together with the whole structure stealth design, particularly for fighters and strategic/long-range bombers. These coatings are usually called radar-absorbing materials made of magnetic nanomaterials and/or carbon-based porous materials, which are discussed in detail in Chapter 18 of this book. The modern information communication, particularly for long-distance communication, is mainly based on the encoding-emission-transmission-receiving-decoding technologies of EM waves up to now even though the quantum communication that is still a long time for commercialization has emerged. Therefore, electromagnetic interference (EMI) compatibility coatings are still the key technique for the information security related to the national defense and homeland safety. The microstructure and material design of these coatings is also based on the EM-wave–matter interaction, which will be summarized in Chapter 18 together with the radar stealth coatings. In the marine voyages, the most key coatings are possibly those coatings related to the anti-marine corrosion coatings and the anechoic coating systems in the sonar (sound navigation and ranging) systems for the underwater vehicles and surface ships, which are intimately related to the national coastal defense safety and the maritime trade profit security. The detailed analyses on these coatings are summarized in Chapters 16 and 17, respectively. Here, we just want to mention one interesting progress in the hydrogel microphones for stealthy underwater listening related to the future marine source exploration [216]. Vehicles traveling in oceans, particularly for those underwater vehicles (e.g. submarines, unmanned undersea vehicles), usually navigate based on their sonar systems to monitor flow velocities and sound waves to navigate, to identify hostile objects, to track ocean currents and surface waves, and to communicate with each other [217, 218]. However, the conventional ceramic-piezoelectric (PZT)-based

1.7 Thin Films and Coatings for National Defense and Homeland Security

sonar systems suffer from a large acoustic impedance mismatch with water, causing them to be easily detected by hostile vehicles during the current era of stealthy navigation because they efficiently reflect incoming acoustic signals. [216, 219, 220]. In addition, the detection efficiency of PZT-based acoustic sensors is relatively poor at low frequencies [216]. Alternatively, suspended thin membranes of poly(vinylidene fluoride) [221] or graphene stretched over air cavities [222] have been proposed as microphones to afford a higher sensitivity than PZT [223], but these configurations introduce even larger mismatch in acoustic impedance between the device (air) and water [216]. Considering the advances in acoustic metamaterial cloaking, which greatly attenuates incoming acoustic signals, thereby concealing submarine bodies from sonar detection, a ceramic PZT detector or cavity-based microphone, which is necessarily kept outside of this “invisibility cloak,” remains a strong acoustic reflector [224–226]. In contrast to a rigid solid such as ceramics or a low-density compliant medium such as air, hydrogels have almost perfect acoustic impedance matching with water [216]. Polar functional groups from the backbone or side chains allow hydrogels to absorb a large amount of liquid into three-dimensional polymer networks without leaking. Different from dielectric capacitors, where their capacitance is governed by the distance between two parallel electrodes [227], hydrogel capacitors derive their capacitance from electrical double layers (EDLs) [216, 228]. With the excellent acoustic impedance match to water, a hydrogel capacitor seems to be a promising acoustic transducer. The problem, however, is that the low compressibility of water means that an EDL capacitor would have low sensitivity to pressures. In addressing this limitation, a suitable sensor can be made by incorporating a deformable network of metal nanoparticles (MNPs) into the hydrogel. The MNP network makes the capacitor highly sensitive to mechanical stimuli through a coupling between the deformation of the MNP network and the ion modulation. As a result, this MNP–hydrogel capacitor is able to detect deformation, pressure, and acoustic waves. The key to this hydrophone is the fabrication of the MNP–hydrogel network. Figure 1.11a schemes the synthesis process of the MNP–hydrogel network. In step 1, hydrogel is presoaked in an aqueous bath of AgNO3 (e.g. 10 mM) [216]. Then, hydrogel is sandwiched and biased between an amorphous silicon (a-Si) and an ITO plate (step 2). After that, photoactivated a-Si reduces Ag+ into Ag0 nanoparticles at specific locations (step 3). Finally, MNP–hydrogel is soaked in a copper sulfate bath to prepare a smooth and robust layer of surface electrodes (step 4). Figure 1.11b,c gives the photograph of the patterned Ag nanoparticles in the skin depth of the hydrogel, the high-resolution SEM image, and the schematic of the dendritic MNP network inside the hydrogel. Furthermore, hydrogel microphones can be fabricated by the cavity-free coatings via integrating the easily deformable MNP–hydrogel network in the hydrogel matrix, as highlighted in Figure 1.12a–g [216]. Figure 1.12a shows an example of a 9 mm2 hydrogel microphone fabricated by forming an MNP network consisting of dendritic structures 2–3 mm in size and being buried inside the soft and translucent hydrogel matrix (Figure 1.12b). This MNP–hydrogel microphone was electrically biased at 1 V and submerged in water (Figure 1.12c), where it picked

39

40

1 Introduction: Progress of Thin Films and Coatings

(a) –

a-Si

hv

– +

+ Ag+ ITO Step 1

Hydrogel Step 2

Step 3

Cu2+ Step 4

(b)

(c)

Ag + 500 nm

Figure 1.11 Fabrication of hydrogel microphones. (a) Steps to fabricate a deformable network of metal nanoparticles (MNPs) and surface electrode. (b) Photos of patterned Ag nanoparticles in the skin depth of the hydrogel. Scale bar, 3.0 mm. (c) High-resolution SEM image and schematic of a dendritic MNP network inside the hydrogel. Source: Gao et al. [216].

up acoustic waves and produced a signal 30 dB stronger at low frequencies than a commercial hydrophone (Figure 1.12d). Moreover, the hydrogel microphone has a wide frequency response, up to 2 kHz (Figure 1.12e), and has a pronounced directional sensitivity perpendicular to the sensor surface (Figure 1.12f). MNP–hydrogel (Figure 1.12g: solid lines) responds to a static pressure of 5.4 kPa with more than four times in relative capacitance change or seven to eight times in capacitance change than MNP-free device (Figure 1.12g: dashed lines). Since MNPs can be densely implanted as inclusions and can even be arranged in coherent arrays, the general performance testing results suggest that this microphone can detect static loads and air breezes from different angles, as well as underwater acoustic signals from 20 Hz to 3 kHz at amplitudes as low as 4 Pa [216]. Unlike dielectric capacitors or cavity-based microphones that respond to stimuli by deforming the device in thickness directions, this hydrogel device responds with a transient modulation of electric double layers, resulting in an extraordinary sensitivity (217 nF/kPa or 24 μC/N at a bias of 1.0 V) without using any signal amplification tools [216]. Due to their perfect acoustic impedance matching with water, ultrasensitive for low-frequency acoustic waves, the nanomaterial–hydrogel-based hydrophones have currently become an

Acknowledgments

(a)

3m

m θ = 0°

Signal (mV)

15

θ 30°

10 60° 5

(b)

90° 0

0

2

1

10 20 30 40 50

0

Signal (mV)

Time (min)

(e)

(f)

1 μm

Water ITO

Hydrogel Ag NPs

(c)

Hydrogel microphone –150 –160 ˜30 dB –170

20

(d)

40 Frequency (Hz)

16.0 14.0 12.0

20.0

10.0

15.0

8.0 6.0

10.0

4.0

5.0

2.0 0.0

0.0

Hydrophone –180

18.0 Static pressure of 5.4 kPa

25.0

1.0 60

Capacitance change (ΔC, nF)

R

–140 Signal (dB re 1 V/μPa

1V

Relative capacitance change (ΔC/C0, %)

30.0 Signal

–2

1.0 10.0 100.0 Salt concentration (mM)

1000.0

(g)

Figure 1.12 Highlight of hydrogel microphone. (a) Photos (full view and sliced) and (b) scanning electron microscopy (SEM) image of the hydrogel membrane implanted with a patch (3 × 3 mm2 ) of silver dendrites (highlighted yellow). (c) Setup and circuit using the membrane as a microphone. (d) Better performance of the hydrogel microphone at low frequencies than a commercial device (hydrophone). (e) Hydrogel microphone is capable of detecting underwater sound at 2 kHz and (f) at all angles. Note: the 0_ orientation is for the top surface of the microphone facing toward the loudspeaker. (g) MNP–hydrogel (solid lines) responds to a static pressure of 5.4 kPa with more than four times in relative capacitance change or seven to eight times in capacitance change than MNP-free device (dashed lines). The relative errors of direct current (DC) and DC/C 0 for MNP–hydrogel are, respectively, 21–25% and 3.5–4.5%. For MNP-free hydrogel, the relative errors of DC and DC/C 0 are, respectively, 13.0–13.5% and 3.0–4.0%. Source: Gao et al. [216].

exciting area for developing hydrophones for potential antiscouting sonar [229, 230] or ultrasensitive stretchable pressure sensors [231, 232].

Acknowledgments This chapter was supported by the NSFC-BRICS STI Framework Program (No. 51861145309), the National Natural Science Foundation of China (No. 51971029), the National S&T Major Project (No. 2018ZX10301201), the “All English teaching demonstration course construction project of University of Science and Technology Beijing” (No. KC2015QYW06, 2016), the “1125” Zhihui Zhengzhou Talent project of Henan province (Fund No. in USTB: 39080070), the “100 talent plan” fund of Fujian province (Fund No. in USTB: 39080067), and the development of a highly sensitive magneto-optical biomolecular sensor experimental prototype (Fund No. in USTB: 2019-0649) by Hangzhou Ruidi Biotechnology Co. Ltd.

41

42

1 Introduction: Progress of Thin Films and Coatings

List of Abbreviations 2D 3D AEMs AIBs AND BDA BIF CBD CCD C-dots CIGS CMOS CMP CMR CO2 RR CVD DJ EDLs EG EM EMI ESSs ETL-free PSCs F4TCNQ FCs FEA FET FOLEDs GDEs GMR GNRs GO GON Gp GPU h-BN (hBN) HBPs HER HR-TEM HTM ITO LDH LED

two-dimensional three-dimensional atom electronic mechanics aluminum-ion batteries agree nor disagree butanediamine build-in field controlled breakdown charge-coupled device carbon quantum dots copper indium gallium selenide complementary metal-oxide semiconductor conjugated microporous polymer colossal magnetoresistance reduction reaction of CO2 chemical vapor deposition Dion Jacobson electrical double layers expansion reduced graphene electromagnetic electromagnetic interference energy storage systems electron transport layer-free perovskite solar cells 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquino dimethane fuel cells pentafluorophenylethyl ammonium field emission transistor foldable OLEDs gas diffusion electrodes giant magnetoresistance graphene nanoribbons graphene oxide germanium-on-nothing graphene graphics processing unit hexagonal boron nitride hyperbranched polymers hydrogen evolution reaction high-resolution transmission electron microscopy hole-transporting material indium tin oxide layered bimetallic hydroxide laser emission diodes

List of Abbreviations

LSC MAPbI3 MBE MEMs MNC MNPs MO MOSFET MoTe2 NAND NCPV NEMs NFs NPN NPs NREL OER OLED ORR PBES PCE PDA PEM PFE PG P-Hg PHJ PLQY PN PNPL PSCs poly-Si PV PZT QAHE QW rGO RTG SACs SEIs SEM SFCMs SiNx SMEMs SSCs

luminous solar concentrator methylammonium lead iodide molecular beam epitaxy microelectronic mechanics metal–nitrogen–carbon metal nanoparticles magneto-optical metal-oxide-semiconductor field-effect transistor molybdenum ditelluride neither agree nor disagree National Center for Photovoltaics nanoelectronic mechanics nanofibers nanoporous silicon nitride nanoparticles National Renewable Energy Laboratory oxygen evolution reaction organic light emitting diode oxygen reduction reaction protein-based electronic skin power conversion efficiency propanediamine proton-exchange membrane polar ferroelectric PNIPAm modified graphene black phosphorus plane heterojunctions photoluminescence quantum yield poly(N-isopropylacrylamide) (PNIPAm) halide perovskite nanoflake perovskite solar cells polysilicon photovoltaic piezoelectric quantum anomalous hall effect quantum well reduced graphene oxide thermoelectric power generator single-atom catalysts solid electrolyte interfaces scanning electronic microscope silk fibroin composite membranes silicon nitride single-molecule electronic mechanics single-site catalysts

43

44

1 Introduction: Progress of Thin Films and Coatings

TBCs TE TM TMDs TMR TSCs UCD XeF2 XOR

thermal barrier coatings thermoelectric transition metal transition metal dichalcogenides tunneling magnetoresistance tandem solar cells Universal Communication DeviceTM xenon difluoride exclusive-OR

References 1 Kroemer, H. (2001). Nobel Lecture: quasielectric fields and band offsets: teaching electrons new tricks. Reviews of Modern Physics 73: 783. 2 Alferov, Z.I. (2001). Nobel lecture: the double heterostructure concept and its applications in physics, electronics, and technology. Reviews of Modern Physics 73: 767. 3 Gao, W., Zhu, X., Zheng, F., et al. (2018). A possible candidate for triply degenerate point fermions in trigonal layered PtBi2 . Nature Communications 9: 3249. https://doi.org/ 10.1038/s41467-018-05730-3. 4 Xiang, R., Inoue1, T., Zheng, Y., et al. (2020). One-dimensional van der Waals heterostructures. Science 367: 537–542. 5 Schwierz, F. (2010). Graphene transistors. Nature Nanotechnology 5, 487–496. 6 Ge, S., Zhao, Z., Pai, J.-H. et al. (2016). Highly sensitive, wearable, durable strain sensors, and stretchable conductors using graphene/silicon rubber composites. Advanced Functional Materials 26: 7614–7625. 7 Yang, H., Yao, X., Yuan, L. et al. (2019). Strain-sensitive electrical conductivity of carbon nanotube-graphene-filled rubber composites under cyclic loading. Nanoscale Research Letters 11: 578–586. 8 Yin, F., Lu, H., Pan, H., et al. (2019). Highly sensitive and transparent strain sensors with an ordered array structure of AgNWs for wearable motion and health monitoring. Scientific Reports 9: 1–10. 9 Han, S., Chand, A., Araby, S., et al. (2019). Thermally and electrically conductive multifunctional sensor based on epoxy/graphene composite. Nanotechnology 31: 075702. 10 Pan, C., Fu, Y., Wang, J., et al. (2018). Vertical transistors: analog circuit applications based on ambipolar graphene/MoTe2 vertical transistors (Adv. Electron. Mater. 3/2018). Advanced Electronic Materials 4: 1870015. 11 Qi, X.-L. and Zhang, S.-C. (2010). The quantum spin Hall effect and topological insulators. Physics Today 63: 33–38. 12 Kong, D. and Cui, Y. (2011). Opportunities in chemistry and materials science for topological insulators and their nanostructures. Nature Chemistry 3: 845–849.

References

13 Rienks, E.D.L., Wimmer, S., Sánchez-Barriga, J., et al. (2019). Large magnetic gap at the Dirac point in Bi2 Te3 /MnBi2 Te4 heterostructures. Nature 576: 423–428. https://doi.org/10.1038/s41586-019-1826-7. 14 Cao, Y., Fatemi, V., Demir, A., et al. (2018). Correlated insulator behaviour at half-filling in magic-angle graphene superlattices. Nature 556: 80–84. https://doi .org/10.1038/nature26154. 15 Rhodes, D., Chae, S.H., Ribeiro-Palau, R., and Hone, J. (2019). Disorder in van der Waals heterostructures of 2D materials. Nature Materials 18: 541–549. https://doi.org/10.1038/s41563-019-0366-8. 16 Wang, E., Lu, X., Ding, S., et al. (2016). Gaps induced by inversion symmetry breaking and second-generation Dirac cones in graphene/hexagonal boron nitride. Nature Physics 12: 1111–1115. https://doi.org/10.1038/nphys3856. 17 Liu, Y., Weiss, N.O., Duan, X., et al. (2016). Van der Waals heterostructures and devices. Nature Reviews Materials 1: 16042. https://doi.org/10.1038/natrevmats .2016.42. 18 Zhang, W., Wang, Y., Wang, Q. et al. (2018). The surface and interface engineering multi-layered nanopore films for enhanced Fabry–Pérot interferences. The Journal of Physical Chemistry C 122: 29457–29463. 19 Knyazev, G.A., Kapralov, P.O., Gusev, N.A. et al. (2018). Magnetoplasmonic crystals for highly sensitive magnetometry. ACS Photonics 5: 4951–4959. https://doi.org/10.1021/acsphotonics.8b01135. 20 Zhang, W., Li, J., Ding, X., et al. (2017). Tunable magneto-optical Kerr effects of nanoporous thin films. Scientific Reports 7: 2888. 21 Diaz-Valencia, B.F., Mejía-Salazar, J.R., Oliveira, O.N. et al. (2017). Enhanced transverse magneto-optical Kerr effect in magnetoplasmonic crystals for the design of highly sensitive plasmonic (bio)sensing platforms. ACS Omega 2: 7682–7685. 22 Zhang, W., Wang, Q., Zhao, C., and Song, Y. (2019). The optical cavity enhanced magneto-optical Kerr effect signals of AAO/Al-based CoFeB nanostructure arrays. Optics Communications 437: 44–49. 23 Assouar, B., Liang, B., Wu, Y., et al. (2018). Acoustic metasurfaces. Nature Reviews Materials 3: 460–472. https://doi.org/10.1038/s41578-018-0061-4. 24 Fan, Y., Shen, N.-H., Zhang, F., et al. Photoexcited graphene metasurfaces: significantly enhanced and tunable magnetic resonances. ACS Photonics https:// doi.org/10.1021/acsphotonics.8b00057. 25 Fasold, S., Linß, S., Kawde, T. et al. (2018). Disorder-enabled pure chirality in bilayer plasmonic metasurfaces. ACS Photonics 5: 1773–1778. 26 Collins, J.T., Hooper, D.C., Mark, A.G. et al. (2018). Second-harmonic generation optical rotation solely attributable to chirality in plasmonic metasurfaces. ACS Nano: 10.1021/acsnano.8b00601. 27 Lee, H., Jung, M., Kim, M., et al. (2018). Acoustically sticky topographic metasurfaces for underwater sound absorption. The Journal of the Acoustical Society of America 143: 1534–1547. 28 Langfeldt, F. and Gleine, W. (2019). Improved sound transmission loss of glass wool with acoustic metamaterials. Proceeding of the 26th ICSV, pp. 7–11.

45

46

1 Introduction: Progress of Thin Films and Coatings

29 Talirz, L., Ruffieux, P., and Fasel, R. (2016). On-surface synthesis of atomically precise graphene nanoribbons. Advanced Materials 28: 6222–6231. 30 Verlhac, B., Bachellier, N., Garnier, L., et al. (2019). Atomic-scale spin sensing with a single molecule at the apex of a scanning tunneling microscope. Science 366: 623–627. 31 Jiang, Y., Chen, Z., Han, Y., et al. (2018). Electron ptychography of 2D materials to deep sub-ångström resolution. Nature 559: 343–349. 32 Xiong, X., Huang, M., Hu, B., et al. (2020). A transverse tunnelling field-effect transistor made from a van der Waals heterostructure. Nature Electronics 3: 106–112. https://doi.org/10.1038/s41928-019-0364-5. 33 Liu, Y., Huang, Y., and Duan, X. (2019). Van der Waals integration before and beyond two-dimensional materials. Nature 567: 323–333. 34 Yan, R., Khalsa, G., Vishwanath, S., et al. (2018). GaN/NbN epitaxial semiconductor/superconductor heterostructures. Nature 555: 183–189. 35 Cao, Y., Fatemi, V., Fang, S., et al. (2018). Unconventional superconductivity in magic-angle graphene superlattices. Nature 556: 43–50. 36 Lin, Z., Liu, Y., Halim, U., et al. (2018). Solution-processable 2D semiconductors for high-performance large-area electronics. Nature 562: 254–258. 37 Gibaja, C., Rodriguez-San-Miguel, D., Ares, P. et al. (2016). Back cover: few-layer antimonene by liquid-phase exfoliation (Angew. Chem. Int. Ed. 46/2016). Angewandte Chemie 55: 14470–14470. 38 Dong, A., Chen, J., Vora, P.M. et al. (2010). Binary nanocrystal superlattice membranes self-assembled at the liquid–air interface. Nature 466: 474–477. 39 Goerbig, M., Montambaux, G., and Piéchon, F. (2014). Measure of Diracness in two-dimensional semiconductors. EPL (Europhysics Letters) 105: 57005. 40 Cai, J., Ruffieux, P., Jaafar, R., et al. (2010). Atomically precise bottom-up fabrication of graphene nanoribbons. Nature 466: 470–473. 41 Gröning, O., Wang, S., Yao, X., et al. (2018). Engineering of robust topological quantum phases in graphene nanoribbons. Nature 560: 209–213. https://doi.org/ 10.1038/s41586-018-0375-9. 42 Son, Y.W., Cohen, M.L., and Louie, S.G. (2006). Energy gaps in graphene nanoribbons. Physical Review Letters 97: 216803. 43 Cao, T., Zhao, F., and Louie, S.G. (2017). Topological phases in graphene nanoribbons: junction states, spin centers, and quantum spin chains. Physical Review Letters 119: 076401. 44 Song, Y., Nallathamby, P.D., Huang, T. et al. (2010). Correlation and characterization of 3D morphological dependent localized surface plasmon resonance spectra of single silver nanoparticles using dark-field optical microscopy and AFM. The Journal of Physical Chemistry C 114: 74–81. 45 Song, Y. (2011). Controlled Fabrication of Noble Metal Nanomaterials via Nanosphere Lithography and their Optical Properties. INTECH Open Access Publisher. 46 Zhang, W., Liu, Z., Belotelov, V.I. et al. (2018). The magnetic properties of CoFeB and CoFeB/Ag nanodot arrays fabricated by a template transfer imprinting method. Thin Solid Films 660: 301–305.

References

47 Song, Y., Yin, W., Wang, Y.H., et al. (2014). Magneto-plasmons in periodic nanoporous structures. Scientific Reports 4: 4991. https://doi.org/10.1038/ srep04991. 48 Song, Y., Zhang, Z., Elsayled-Ali, H.E., et al. (2011). Identification of single nanoparticles. Nanoscale 3: 31–44. 49 Kumar, G., Tang, H.X., and Schroers, J. (2009). Nanomoulding with amorphous metals. Nature 457: 868–872. 50 Jeong, E.G., Kwon, J.H., Kang, K.S. et al. (2019). A review of highly reliable flexible encapsulation technologies towards rollable and foldable OLEDs. Journal of Information Display: 1–14. 51 Rogers, J., Bao, Z., and Lee, T.-W. (2019). Wearable bioelectronics: opportunities for chemistry. Accounts of Chemical Research 52: 521–522. 52 Sugimoto, A., Ochi, H., Fujimura, S., et al. (2004). Flexible OLED displays using plastic substrates. IEEE Journal of Selected Topics in Quantum Electronics 10: 107–114. 53 Park, J.-S., Chae, H., Chung, H.K., and Lee, S.I. (2011). Thin film encapsulation for flexible AM-OLED: a review. Semiconductor Science and Technology 26: 034001. 54 Zhang, W., Zhao, C., Chen, Z., et al. (2020). Ultrasensitive surface-enhanced raman scattering (SERS) nanopore arrayed silver thin film substrates for R6G detection. RSC Advances 10: 23908–23915. 55 Ye, C., Chao, D., Shan, J., et al. (2020). Unveiling the advances of 2D materials for Li/Na-S batteries experimentally and theoretically. Matter 2: 323–344. 56 Seebeck, T. and Magnetische, J. (1826). Polarisation der Metalle und Erze durch Temperatur-Differenz. Annalen der Physik 82: 133–160. 57 Peltier, J.C.A. (1834). New experiments on the heat effects of electric currents. Annales de Chimie Physique 56: 371–386. 58 Satterthwaite, C. and Ure, R. Jr., (1957). Electrical and thermal properties of Bi2 Te3 . Physical Review A 108: 1164–1170. 59 Goldsmid, H.J. (1955). XXVII. Thermoelectric applications of semiconductors. International Journal of Electronics 1: 218–222. 60 Dismukes, J., Ekstrom, L., Steigmeier, E. et al. (1964). Thermal and electrical properties of heavily doped Ge–Si alloys up to 1300 K. Journal of Applied Physics 35: 2899–2907. 61 Allen, A.W. (1997). Detector handbook. Laser Focus World 33: S15. 62 He, J. and Tritt, T.M. (2010). Thermal-to-Electrical Energy Conversion from the Nanotechnology Perspective (ed. J. Garcia-Martinez), 47–77. John Wiley and Sons, Ltd.. https://doi.org/10.1002/9783527629299.ch3. 63 Bansal, P., Vineyard, E., and Abdelaziz, O. (2011). Advances in household appliances - a review. Applied Thermal Engineering 31: 3748–3760. 64 Zhao, L.-D., Lo, S.-H., Zhang, Y., et al. (2014). Ultralow thermal conductivity and high thermoelectric figure of merit in SnSe crystals. Nature Catalysis 508: 373–377. 65 Yang, J., Xi, L., Qiu, W., et al. (2016). On the tuning of electrical and thermal transport in thermoelectrics: an integrated theory–experiment perspective. npj Computational Materials 2: 1–17.

47

48

1 Introduction: Progress of Thin Films and Coatings

66 Tan, G., Hao, S., Cai, S., et al. (2019). All-scale hierarchically structured p-type PbSe alloys with high thermoelectric performance enabled by improved band degeneracy. Journal of the American Chemical Society 141: 4480–4486. 67 Zhang, J., Song, L., and Iversen, B.B. (2019). Insights into the design of thermoelectric Mg3 Sb2 and its analogs by combining theory and experiment. npj Computational Materials 5: 76. 68 Sales, B.C., Mandrus, D., and Williams, R.K. (1996). ChemInform abstract: filled skutterudite antimonides: a new class of thermoelectric materials. Science 272: 1325–1328. 69 Butt, S., Liu, Y.-C., Lan, J.-L., et al. (2014). High-temperature thermoelectric properties of La and Fe co-doped Ca–Co–O misfit-layered cobaltites consolidated by spark plasma sintering. Journal of Alloys and Compounds 588: 277–283. 70 Hao, S., Dravid, V.P., Kanatzidis, M.G., and Wolverton, C.J. (2019). Computational strategies for design and discovery of nanostructured thermoelectrics. npj Computational Materials 5: 58. 71 Morelli, D.T. (2017). Chapter 57. Thermoelectric materials. In: Handbook of Electronic and Photonic Materials, Chapter 57, Part E (eds. S. Kasap and P. Capper), 1379–1395. Springer. 72 Tritt, T.M., Böttner, H., and Chen, L. (2008). Thermoelectrics: direct solar thermal energy conversion. MRS Bulletin 33: 366–368. 73 Hu, Y., Song, L., Chen, Y., and Huang, W. (2019). Two-terminal perovskites tandem solar cells: recent advances and perspectives. Solar RRL 3, https://doi .org/10.1002/solr.201900080. 74 Grinberg, I., West, D.V., Torres, M., et al. (2013). Perovskite oxides for visible-light-absorbing ferroelectric and photovoltaic materials. Nature Catalysis 503: 509–512. 75 Sum, T.C. and Mathews, N. (2014). Advancements in perovskite solar cells: photophysics behind the photovoltaics. Energy & Environmental Science 7: 2518–2534. 76 Roldán-Carmona, C., Malinkiewicz, O., Soriano, A., et al. (2014). Flexible high efficiency perovskite solar cells. Energy & Environmental Science 7: 994–997. 77 Green, M.A., Ho-Baillie, A., and Snaith, H.J. (2014). The emergence of perovskite solar cells. Nature Photonics 8: 506–514. 78 Dualeh, A., Moehl, T., Tétreault, N., et al. (2014). Impedance spectroscopic analysis of lead iodide perovskite-sensitized solid-state solar cells. ACS Nano 8: 362–373. 79 Elumalai, N.K., Mahmud, M.A., Wang, D., and Uddin, A. (2016). Perovskite solar cells: progress and advancements. Energies 9: 861. 80 Rose, G. (1839). De Novis Quibusdam Fossilibus Quae in Montibus Uraliis Inveniuntur. AG Schade. 81 Zhu, X. (2016). The perovskite fever and beyond. Accounts of Chemical Research 49: 355–356. 82 Wu, W.-Q., Wang, Q., Fang, Y., et al. (2018). Molecular doping enabled scalable blading of efficient hole-transport-layer-free perovskite solar cells. Nature Communications 9: 1–8.

References

83 Rong, Y., Hu, Y., Mei, A., et al. (2018). Challenges for commercializing perovskite solar cells. Science 361: eaat8235. 84 Zheng, Y., Niu, T., Qiu, J., et al. (2019). Oriented and uniform distribution of Dion–Jacobson phase perovskites controlled by quantum well barrier thickness. Solar RRL 3: 1900090. 85 Liu, Y., Akin, S., Pan, L., et al. (2019). Ultrahydrophobic 3D/2D fluoroarene bilayer-based water-resistant perovskite solar cells with efficiencies exceeding 22%. Science Advances 5: eaaw2543. 86 Benetti, D., Jokar, E., Yu, C.-H., et al. (2019). Hole-extraction and photostability enhancement in highly efficient inverted perovskite solar cells through carbon dot-based hybrid material. Nano Energy 62: 781–790. 87 Zhang, C.C., Wang, Z.K., Yuan, S., et al. (2019). Polarized ferroelectric polymers for high-performance perovskite solar cells. Advanced Materials 31: 1902222. 88 Braukyla, T., Xia, R., Daskeviciene, M., et al. (2019). Inexpensive hole-transporting materials derived from Tröger’s base afford efficient and stable perovskite solar cells. Angewandte Chemie 131: 11388–11394. 89 Liu, T., Luo, Z., Chen, Y. et al. (2019). A nonfullerene acceptor with a 1000 nm absorption edge enables ternary organic solar cells with improved optical and morphological properties and efficiencies over 15%. Energy & Environmental Science 12: 2529–2536. 90 Chen, Y., Liu, T., Hu, H. et al. (2018). Modulation of end groups for low-bandgap nonfullerene acceptors enabling high-performance organic solar cells. Advanced Energy Materials 8: 1801203. https://doi.org/10.1002/aenm .201801203. 91 Fan, Q., Zhu, Q., Xu, Z. et al. (2018). Chlorine substituted 2D-conjugated polymer for high-performance polymer solar cells with 13.1% efficiency via toluene processing. Nano Energy 48: 413–420. 92 Cui, Y., Yao, H., Zhang, J., et al. (2019). Over 16% efficiency organic photovoltaic cells enabled by a chlorinated acceptor with increased open-circuit voltages. Nature Communications 10: 1–8. 93 Park, S., Simon, J., Schulte, K. L., et al. (2019). Germanium-on-nothing for epitaxial liftoff of GaAs solar cells. Joule 3: 1782–1793. 94 Cai, L. and Yu, G. (2019). Recent advances in growth and modification of graphene-based energy materials: from chemical vapor deposition to reduction of graphene oxide. Small Methods 3: 1900071. 95 Liu, Z., You, P., Xie, C. et al. (2016). Ultrathin and flexible perovskite solar cells with graphene transparent electrodes. Nano Energy 28: 151–157. 96 Li, Z., Johnston, A., Wei, M., et al. (2020). Solvent-solute coordination engineering for efficient perovskite luminescent solar concentrators. Joule 4: 631–643. 97 Huang, L., Zhang, D., Bu, S., et al. (2020). Synergistic interface energy band alignment optimization and defect passivation toward efficient and simple-structured perovskite solar cell. Advanced Science 7: 1902656. https:// doi.org/10.1002/advs.201902656. 98 Polman, A., Knight, M., Garnett, E.C. et al. (2016). Photovoltaic materials: present efficiencies and future challenges. Science 352: aad4424.

49

50

1 Introduction: Progress of Thin Films and Coatings

99 Tiedje, T., Yablonovitch, E., Cody, G.D., and Brooks, B.G. (1984). Limiting efficiency of silicon solar cells. IEEE Transactions on Electron Devices 31: 711–716. 100 De Vos, A. (1980). Detailed balance limit of the efficiency of tandem solar cells. Journal of Physics D: Applied Physics 13: 839. 101 Strandberg, R. (2015). Detailed balance analysis of area de-coupled double tandem photovoltaic modules. Applied Physics Letters 106: 033902. 102 Futscher, M.H. and Ehrler, B. (2016). Efficiency limit of perovskite/Si tandem solar cells. ACS Energy Letters 1: 863–868. 103 Lamanna, E., Matteocci, F., Calabrò, E., et al. (2020). Mechanically stacked, two-terminal graphene-based perovskite/silicon tandem solar cell with efficiency over 26%. Joule 4: 865–881. https://doi.org/10.1016/j.joule.2020.01.015. 104 Shi, L. and Zhao, T. (2017). Recent advances in inorganic 2D materials and their applications in lithium and sodium batteries. Journal of Materials Chemistry A 5: 3735–3758. 105 Kühne, M., Boerrnert, F., Fecher, S., et al. (2018). Reversible superdense ordering of lithium between two graphene sheets. Nature 564: 234–239. https://doi .org/10.1038/s41586-018-0754-2. 106 Nguyen, D.D., Hsiao, C.H., Su, T.Y., et al. (2019). Bioinspired networks consisting of spongy carbon wrapped by graphene sheath for flexible transparent supercapacitors. Communications Chemistry 2: 137. https://doi.org/10.1038/ s42004-019-0238-9. 107 Yu, Z., Jiao, S., Li, S. et al. (2019). Flexible stable solid-state Al–ion batteries. Advanced Functional Materials 29, https://doi.org/10.1002/adfm.201806799. 108 Winter, M. and Besenhard, J.O. (2011). III. Materials for alkali metal batteries. In: Handbook of Battery Materials (eds. C. Daniel and J.O. Besenhard), 433–478. Wiley-VCH. 109 Kaskhedikar, N.A. and Maier, J. (2009). Lithium storage in carbon nanostructures. Advanced Materials 21: 2664–2680. 110 Zheng, H., Meng, Y.S., and Zhu, Y. (2015). Frontiers of in situ electron microscopy. MRS Bulletin 40: 12–18. 111 Qian, D., Ma, C., More, K.L. et al. (2015). Advanced analytical electron microscopy for lithium–ion batteries. NPG Asia Materials 7: e193. 112 Liu, X.H., Liu, Y., Kushima, A., et al. (2012). In situ TEM experiments of electrochemical lithiation and delithiation of individual nanostructures. Advanced Energy Materials 2: 722–741. 113 Shao-Horn, Y., Croguennec, L., Delmas, C. et al. (2003). Atomic resolution of lithium ions in LiCoO2 . Nature Materials 2: 464–467. 114 Oshima, Y. and Murakami, Y. (2017). Challenges for lithium detection (special issue). Microscopy 66: 1–61. 115 Reimer, L. and Kohl, H. (2008), Springer Series in Optical Sciences (eds. W.T. Rhodes, A. Adibi, T. Asakura, et al.). New York: Springer, ISSN: 0342-4111. 116 Linck, M., Hartel, P., Uhlemann, S. et al. (2016). Chromatic aberration correction for atomic resolution TEM imaging from 20 to 80 kV. Physical Review Letters 117: 076101.

References

117 Enoki, T., Suzuki, M., and Endo, M. (2003). Graphite Intercalation Compounds and Applications. Oxford University Press. 118 Li, Z., Gadipelli, S., Li, H., et al. (2020). Tuning the interlayer spacing of graphene laminate films for efficient pore utilization towards compact capacitive energy storage. Nature Energy 5: 160–168. https://doi.org/10.1038/s41560020-0560-6. 119 Stamenkovic, V.R., Strmcnik, D., Lopes, P.P., and Markovic, N.M. (2016). Energy and fuels from electrochemical interfaces. Nature Materials 16: 57–69. https:// doi.org/10.1038/nmat4738. 120 Chu, S., Cui, Y., and Liu, N. (2016). The path towards sustainable energy. Nature Materials 16: 16–22. https://doi.org/10.1038/nmat4834. 121 Ma, J., Wang, L., Deng, Y. et al. (2020). Mass production of high-performance single atomic FeNC electrocatalysts via sequenced ultrasonic atomization and pyrolysis process. Science China Materials, https://doi.org/10.1007/s40843-020-1464-6. 122 Chen, Y., Ji, S., Zhao, S., et al. (2018). Enhanced oxygen reduction with single-atomic-site iron catalysts for a zinc-air battery and hydrogen-air fuel cell. Nature Communications 9: 5422. https://doi.org/10.1038/s41467-018-078502. 123 Tiwari, J.N., Tiwari, R.N., Singh, G., and Kim, K.S. (2013). Recent progress in the development of anode and cathode catalysts for direct methanol fuel cells. Nano Energy 2: 553–578. https://doi.org/10.1016/j.nanoen.2013.06.009. 124 Shao, M., Chang, Q., Dodelet, J.-P., and Chenitz, R. (2016). Recent advances in electrocatalysts for oxygen reduction reaction. Chemical Reviews 116: 3594–3657. https://doi.org/10.1021/acs.chemrev.5b00462. 125 Mori, K., Sano, T., Kobayashi, H., and Yamashita, H. (2018). Surface engineering of a supported PdAg catalyst for hydrogenation of CO2 to formic acid: elucidating the active Pd atoms in alloy nanoparticles. Journal of the American Chemical Society 140: 8902–8909. https://doi.org/10.1021/jacs.8b04852. 126 Yan, Z., He, G., Shen, P. K. et al. (2014). MoC–graphite composite as a Pt electrocatalyst support for highly active methanol oxidation and oxygen reduction reaction. Journal of Materials Chemistry A 2: 4014. https://doi.org/10.1039/ c3ta14251e. 127 Yao, Y., Hu, S., Chen, W. et al. (2019). Engineering the electronic structure of single atom Ru sites via compressive strain boosts acidic water oxidation electrocatalysis. Nature Catalysis 2: 304–313. https://doi.org/10.1038/s41929-0190246-2. 128 Ma, J., Tong, X., Wang, J. et al. (2019). Rare-earth metal oxide hybridized PtFe nanocrystals synthesized via microfluidic process for enhanced electrochemical catalytic performance. Electrochimica Acta 299: 80–88. https://doi.org/10.1016/j .electacta.2018.12.132. 129 Wang, Z., Fan, H., Liang, H., et al. (2017). Microfluidic synthesis and characterization of FePtSn/C catalysts with enhanced electro-catalytic performance for direct methanol fuel cells. Electrochimica Acta 230: 245–254. https://doi.org/10 .1016/j.electacta.2017.01.1590013-4686.

51

52

1 Introduction: Progress of Thin Films and Coatings

130 Ott, S., Orfanidi, A., Schmies, H., et al. (2020). Ionomer distribution control in porous carbon-supported catalyst layers for high-power and low Pt-loaded proton exchange membrane fuel cells. Nature Materials 19: 77–85. https://doi .org/10.1038/s41563-019-0487-0. 131 Wang, J., Ye, H., and Song, Y. (2018). In-situ reaction-growth of PtNiX nanocrystals on supports forenhanced electrochemical catalytic oxidation of ethanol via continuous flow microfluidic process. Electrochimica Acta 278: 149–155. 132 Glüsen, A., Dionigi, F., Paciok, P., et al. (2019). Dealloyed PtNi-core–shell nanocatalysts enable significant lowering of Pt electrode content in direct methanol fuel cells. ACS Catalysis 9: 3764–3772. https://doi.org/10.1021/acscatal .8b04883. 133 Dionigi, F., Weber, C.C., Primbs, M., et al. (2019). Controlling near-surface Ni composition in octahedral PtNi(Mo) nanoparticles by Mo doping for a highly active oxygen reduction reaction catalyst. Nano Letters 19: 6876–6885. https:// doi.org/10.1021/acs.nanolett.9b02116. 134 Sun, Y., Silvioli, L., Sahraie, N.R., et al. (2019). Activity–selectivity trends in the electrochemical production of hydrogen peroxide over single-site metal–nitrogen–carbon catalysts. Journal of the American Chemical Society 141: 12372–12381. https://doi.org/10.1021/jacs.9b05576. 135 Garcon, M., Bakewell, C., Sackman, G.A., et al. (2019). A hexagonal planar transition-metal complex. Nature 574: 390–393. https://doi.org/10.1038/s41586019-1616-2. 136 Song, Y., Hart, K.T., and Dooley, K.M. (2004). Waste-reducing catalysis for acylation of a secondary amine: synthesis of DEET. Catalysis Letters 98: 69–75. https://doi.org/10.1007/s10562-004-7933-5. 137 Song, Y., Hart, K.T., and Dooley, K.M. (2016). Waste-reducing catalytic oxidation of m-xylene to m-toluic acid. Catalysis Letters 146: 1213–1220. 138 Jasinski, R. (1964). A new fuel cell cathode catalyst. Nature 201: 1212–1213. https://doi.org/10.1038/2011212a0. 139 Huang, K., Zhang, L., Xu, T., et al. (2019). 60∘ C solution synthesis of atomically dispersed cobalt electrocatalyst with superior performance. Nature Communications 10: 606. https://doi.org/10.1038/s41467-019-08484-8. 140 Liu, X., Wang, L., Yu, P. et al. (2018). A stable bifunctional catalyst for rechargeable zinc-air batteries: iron-cobalt nanoparticles embedded in a nitrogen-doped 3D carbon matrix. Angewandte Chemie 57: 16166–16170. https://doi.org/10.1002/anie.201809009. 141 Huang, X., Zhang, Y., Shen, H., et al. (2018). N-doped carbon nanosheet networks with favorable active sites triggered by metal nanoparticles as bifunctional oxygen electrocatalysts. ACS Energy Letters 3: 2914–2920. https://doi.org/ 10.1021/acsenergylett.8b01717. 142 Sun, T., Tian, B., Lu, J., and Su, C. (2017). Recent advances in Fe (or Co)/N/C electrocatalysts for the oxygen reduction reaction in polymer electrolyte membrane fuel cells. Journal of Materials Chemistry A 5: 18933–18950. https://doi .org/10.1039/c7ta04915c.

References

143 Nie, Y., Li, L., and Wei, Z. (2015). Recent advancements in Pt and Pt-free catalysts for oxygen reduction reaction. Chemical Society Reviews 44: 2168–2201. https://doi.org/10.1039/c4cs00484a. 144 Inamdar, A.I., Chavan, H.S., Hou, B., et al. (2020). Electrochemical water splitting: a robust nonprecious CuFe composite as a highly efficient bifunctional catalyst for overall electrochemical water splitting. Small 16: 1905884. 145 Chen, Z., Song, Y., Cai, J., et al. (2018). Tailoring the d band centers enables Co4 N nanosheets to be highly active for hydrogen evolution catalysis. Angewandte Chemie International Edition 130: 5170–5174. https://doi.org/10.1002/ anie.201801834. 146 Li, G., Chen, Z., Li, Y., et al. (2020). Engineering substrate interaction to improve hydrogen evolution catalysis of monolayer MoS2 films beyond Pt. ACS Nano 14: 1707–1714. https://doi.org/10.1021/acsnano.9b07324. 147 Ye, G., Gong, Y., Lin, J. et al. (2016). Defects engineered monolayer MoS2 for improved hydrogen evolution reaction. Nano Letters 16: 1097–1103. https://doi .org/10.1021/acs.nanolett.5b04331. 148 Cho, K.H., Seo, H., Park, S., et al. (2020). Uniform, assembled 4 nm Mn3 O4 nanoparticles as efficient water oxidation electrocatalysts at neutral pH. Advanced Functional Materials 30: 1910424. https://doi.org/10.1002/adfm .201910424. 149 Alvarez, P.J.J., Chan, C.K., Elimelech, M. et al. (2018). Emerging opportunities for nanotechnology to enhance water security. Nature Nanotechnology 13: 634–641. https://doi.org/10.1038/s41565-018-0203-2. 150 Schiermeier, Q. (2014). Water on tap. Nature 510: 327–328. 151 Song, Y., Sun, P., Henry, L.L., and Sun, B. (2005). Mechanism for structure and performance controlled preparation of thin film composite membrane via interfacial polymerization. Journal of Membrane Science 251: 67–72. 152 Song, Y., Liu, F., and Sun, B. (2005). Preparation, characterization, and application of thin film composite nanofiltration membranes. Journal of Applied Polymer Science 95: 1251–1256. 153 Song, Y. (2000). Nanofiltration thin film composite membranes. PhD thesis. Beijing University of Chemical Technology. 154 Inukai, S., Cruz-Silva, R., Ortiz-Medina, J. et al. (2015). High-performance multi-functional reverse osmosis membranes obtained by carbon nanotube⋅polyamide nanocomposite. Scientific Reports 5: 13562. https://doi .org/10.1038/srep13562. 155 Geng, H., Xu, Q., Wu, M., et al. (2019). Plant leaves inspired sunlight-driven purifier for high-efficiency clean water production. Nature Communications 10: 1512. https://doi.org/10.1038/s41467-019-09535-w. 156 Chen, W., Chen, S., Liang, T., et al. (2018). High-flux water desalination with interfacial salt sieving effect in nanoporous carbon composite membranes. Nature Nanotechnology 13: 345–350. https://doi.org/10.1038/s41565-018-0067-5. 157 Song, Y., Wang, D.K., Birkett, G., et al. (2016). Mixed matrix carbon molecular sieve and alumina (CMS-Al2 O3 ) membranes. Scientific Reports 6: 30703. https:// doi.org/10.1038/srep30703.

53

54

1 Introduction: Progress of Thin Films and Coatings

158 Khorshidi, B., Thundat, T., Fleck, B.A., and Sadrzadeh, M. (2016). A novel approach toward fabrication of high performance thin film composite polyamide membranes. Scientific Reports 6: 22069. https://doi.org/10.1038/ srep22069. 159 Vlassiouk, I.V. (2017). A scalable graphene-based membrane. Nature Nanotechnology 12: 1022–1023. https://doi.org/10.1038/nnano.2017.184. 160 Chen, L., Shi, G., Shen, J., et al. (2017). Ion sieving in graphene oxide membranes via cationic control of interlayer spacing. Nature 550: 380–383. https:// doi.org/10.1038/nature24044. 161 Liang, B., Wang, H., Shi, X., et al. (2018). Microporous membranes comprising conjugated polymers with rigid backbones enable ultrafast organic-solvent nanofiltration. Nature Chemistry 10: 961–967. https://doi.org/10.1038/s41557018-0093-9. 162 Enrico, D., Lidietta, G., and Enrica, F. (2017). Comprehensive Membrane Science and Engineering, 2e. Elsevier Science. 163 Al-Abri, M., Al-Ghafri, B., Bora, T., et al. (2019). Chlorination disadvantages and alternative routes for biofouling control in reverse osmosis desalination. NPJ Clean Water 2: 2. https://doi.org/10.1038/s41545-018-0024-8. 164 Byun, Y. and Coskun, A. (2015). Bottom-up approach for the synthesis of a three-dimensional nanoporous graphene nanoribbon framework and its gas sorption properties. Chemistry of Materials 27: 2576–2583. https://doi.org/10 .1021/acs.chemmater.5b00246. 165 Ye, H., Li, D., Ye, X., et al. (2019). An adjustable permeation membrane up to the separation for multicomponent gas mixture. Scientific Reports 9: 7380. https://doi.org/10.1038/s41598-019-43751-0. 166 Han, Y., Xu, Z., and Gao, C. (2013). Ultrathin graphene nanofiltration membrane for water purification. Advanced Functional Materials 23: 3693–3700. 167 Keerthi, A., Geim, A.K., Janardanan, A., et al. (2018). Ballistic molecular transport through two-dimensional channels. Nature 558: 420–424. https://doi.org/10 .1038/s41586-018-0203-2. 168 Geim, A.K. and Grigorieva, I.V. (2013). Van der Waals heterostructures. Nature 499: 419–425. https://doi.org/10.1038/nature12385. 169 Joshi, R.K., Carbone, P., Wang, C., et al. (2014). Precise and ultrafast molecular sieving through graphene oxide membranes. Science 343: 752–754. https://doi .org/10.1126/science.1245711. 170 Liu, G.P., Jin, W.Q., and Xu, N.P. (2015). Graphene-based membranes. Chemical Society Reviews 44: 5016–5030. 171 Yang, Q., Su, Y., Chi, C., et al. (2017). Ultrathin graphene-based membrane with precise molecular sieving and ultrafast solvent permeation. Nature Materials 16: 1198–1202. https://doi.org/10.1038/nmat5025. 172 Li, B., Japip, S., and Chung, T.-S. (2020). Molecularly tunable thin-film nanocomposite membranes with enhanced molecular sieving for organic solvent forward osmosis. Nature Communications 11: 1198. https://doi.org/10.1038/ s41467-020-15070-w.

References

173 Mahmoudi, E., Ng, L.Y., Ang, W.L., et al. (2019). Enhancing morphology and separation performance of polyamide 6,6 membranes by minimal incorporation of silver decorated graphene oxide nanoparticles. Scientific Reports 9: 1216. https://doi.org/10.1038/s41598-018-38060-x. 174 Celebi, K., Buchheim, J., Wyss, R.M., et al. (2014). Ultimate permeation across atomically thin porous graphene. Science 344: 289–292. 175 Dikin, D.A., Stankovich, S., Zimney, E.J., et al. (2007). Preparation and characterization of graphene oxide paper. Nature 448: 457–460. https://doi.org/10 .1038/nature06016. 176 Elimelech, M. and Phillip, W.A. (2011). The future of seawater desalination: energy, technology, and the environment. Science 333: 712–717. 177 Gin, D.L. and Noble, R.D. (2011). Designing the next generation of chemical separation membranes. Science 332: 674–676. 178 Radha, B., Esfandiar, A., Wang, F.C., et al. (2016). Molecular transport through capillaries made with atomic-scale precision. Nature 538: 222–225. https://doi .org/10.1038/nature19363. 179 Sun, P., Zheng, F., Zhu, M., et al. (2014). Selective trans-membrane transport of alkali and alkaline earth cations through graphene oxide membranes based on cation−π interactions. ACS Nano 8: 850–859. https://doi.org/10.1021/nn4055682. 180 Surwade, S.P., Smirnov, S.N., Vlassiouk, I.V., et al. (2015). Water desalination using nanoporous single-layer graphene. Nature Nanotechnology 10: 459–464. https://doi.org/10.1038/nnano.2015.37. 181 Lin, L.-C. and Grossman, J.C. (2015). Atomistic understandings of reduced graphene oxide as an ultrathin-film nanoporous membrane for separations. Nature Communications 6: 8335. https://doi.org/10.1038/ncomms9335. 182 Kim, H.W., Yoon, H.W., Yoon, S.-M., et al. (2013). Selective gas transport through few-layered graphene and graphene oxide membranes. Science 342: 91–95. https://doi.org/10.1126/science.1236098. 183 Li, H., Song, Z., Zhang, X., et al. (2013). Ultrathin, molecular-sieving graphene oxide membranes for selective hydrogen separation. Science 342: 95–98. https:// doi.org/10.1126/science.1236686. 184 Koenig, S.P., Wang, L., Pellegrino, J., and Bunch, J.S. (2012). Selective molecular sieving through porous graphene. Nature Nanotechnology 7: 728–732. https://doi .org/10.1038/nnano.2012.162. 185 De Volder, M.F.L., Tawfick, S.H., Baughman, R.H., and Hart, A.J. (2013). Carbon nanotubes: present and future commercial applications. Science 339: 535–539. https://doi.org/10.1126/science.1222453. 186 Sun, P., Wang, K., and Zhu, H. (2016). Reent developments in graphene-based membranes: structure, mass-transport mechanism and potential applications. Advanced Materials 28: 2287–2310. 187 Seo, D.H., Pineda, S., Woo, Y.C., et al. (2018). Anti-fouling graphene-based membranes for effective water desalination. Nature Communications 9: 683. https://doi.org/10.1038/s41467-018-02871-3. 188 Lee, S.Y., Jung, H., Kim, N.-K., et al. (2018). Mixed copper states in anodized Cu electrocatalyst for stable and selective ethylene production from CO2

55

56

1 Introduction: Progress of Thin Films and Coatings

189

190

191

192

193

194

195 196

197

198

199

200

201 202

reduction. Journal of the American Chemical Society 140: 8681–8689. https:// doi.org/10.1021/jacs.8b02173. Yi, Z., Xin-feng, Q., and Zi-li, Z. (2014). Technical analysis and application prospect on resource utilization of carbon dioxide resource technology. Science Technology & Engineering 16: 175–183. Varela, A.S., Ju, W., Bagger, A., et al. (2019). Electrochemical reduction of CO2 on metal–nitrogen-doped carbon catalysts. ACS Catalysis 9: 7270–7284. https:// doi.org/10.1021/acscatal.9b01405. Hussain, J., Jónsson, H., and Skúlason, E. (2018). Calculations of product selectivity in electrochemical CO2 reduction. ACS Catalysis 8: 5240–5249. https://doi .org/10.1021/acscatal.7b03308. Ju, W., Bagger, A., Hao, G.-P., et al. (2017). Understanding activity and selectivity of metal–nitrogen-doped carbon catalysts for electrochemical reduction of CO2 . Nature Communications 8: 944. Nudelman, R., Alhmoud, H., Delalat, B., et al. (2019). Jellyfish-based smart wound dressing devices containing in situ synthesized antibacterial nanoparticles. Advanced Functional Materials 29: 1902783. https://doi.org/10.1002/adfm .201902783. Madejski, G.R., Briggs, K., DesOrmeaux, J.-P., et al. (2019). Monolithic fabrication of NPN/SiNx dual membrane cavity for nanopore-based DNA sensing. Advanced Materials Interfaces 6: 1900684. https://doi.org/10.1002/admi .201900684. Ulubayram, K., Cakar, A.N., Korkusuz, P. et al. (2001). EGF containing gelatin-based wound dressings. Biomaterials 22: 1345–1356. Madhumathi, K., Sudheesh Kumar, P.T., Abhilash, S., et al. (2010). Development of novel chitin/nanosilver composite scaffolds for wound dressing applications. Journal of Materials Science: Materials in Medicine 21: 807–813. https://doi.org/10.1007/s10856-009-3877-z. Jin, H., Huang, W., Zhu, X. et al. (2012). Biocompatible or biodegradable hyperbranched polymers: from self-assembly to cytomimetic applications. Chemical Society Reviews 41: 5986–5997. Choueiri, R.M., Galati, E., Thérien-Aubin, H., et al. (2016). Surface patterning of nanoparticles with polymer patches. Nature 538: 79–83. https://doi.org/10 .1038/nature19089. Huang, J., Xu, Z., Qiu, W., et al. (2020). Stretchable and heat-resistant protein-based electronic skin for human thermoregulation. Advanced Functional Materials 30: 1910547. https://doi.org/10.1002/adfm.201910547. Nie, Z.H., Petukhova, A., and Kumacheva, E. (2010). Properties and emerging applications of self-assembled structures of inorganic nanoparticles. Nature Nanotechnology 5: 15–25. Lattuada, M. and Hatton, T.A. (2011). Synthesis, properties and applications of Janus nanoparticles. Nano Today 6: 286–308. Andala, D.M., Shin, S.H.R., Lee, H.Y., and Bishop, K.J.M. (2012). Templated synthesis of amphiphilic nanoparticles at the liquid–liquid interface. ACS Nano 6: 1044–1050.

References

203 Vilain, C., Goettmann, F., Moores, A. et al. (2007). Study of metal nanoparticles stabilised by mixed ligand shell: a striking blue shift of the surface-plasmon band evidencing the formation of Janus nanoparticles. Journal of Materials Chemistry 17: 3509–3514. 204 Jackson, A.M., Myerson, J.W., and Stellacci, F. (2004). Spontaneous assembly of subnanometre ordered domains in the ligand shell of monolayer-protected nanoparticles. Nature Materials 3: 330–336. 205 Bao, C., Tang, S., Wright, R.A.E., et al. (2014). Effect of molecular weight on lateral microphase separation of mixed homopolymer brushes grafted on silica particles. Macromolecules 47: 6824–6835. https://doi.org/10.1021/ma501474m. 206 Goto, Y., Yanagi, I., Matsui, K. et al. (2016). Integrated solid-state nanopore platform for nanopore fabrication via dielectric breakdown, DNA-speed deceleration and noise reduction. Scientific Reports 6: 31324. https://doi.org/10.1038/ srep31324. 207 Waugh, M., Briggs, K., Gunn, D., et al. (2020). Solid-state nanopore fabrication by automated controlled breakdown. Nature Protocols 15: 122–143. https://doi .org/10.1038/s41596-019-0255-2. 208 Goto, Y., Akahori, R., Yanagi, I., and Takeda, K.-i. (2020). Solid-state nanopores towards single-molecule DNA sequencing. Journal of Human Genetics 65: 69–77. https://doi.org/10.1038/s10038-019-0655-8. 209 Li, J., Zhang, W., Song, Y. et al. (2016). Template transfer nanoimprint for uniform nanopores and nanopoles. Journal of Nanomaterials 2016: 9354364. 210 Cao, K., Jiang, Z., Zhao, J., et al. (2014). Enhanced water permeation through sodium alginate membranes by incorporating graphene oxides. Journal of Membrane Science 469: 272–283. https://doi.org/10.1016/j.memsci.2014.06.053. 211 Zhou, K.G., Vasu, K.S., Cherian, C.T., et al. (2018). Electrically controlled water permeation through graphene oxide membranes. Nature 559: 236–240. https:// doi.org/10.1038/s41586-018-0292-y. 212 Zhao, C., Nie, S., Tang, M., and Sun, S. (2011). Polymeric pH-sensitive membranes-a review. Progress in Polymer Science 36: 1499–1520. 213 Liu, Z., Wang, W., Xie, R. et al. (2016). Stimuli-responsive smart gating membranes. Chemical Society Reviews 45: 460–475. 214 Nair, R.R., Wu, H.A., Jayaram, P.N. et al. (2012). Unimpeded permeation of water through helium-leak-tight graphene-based membranes. Science 335: 442–444. 215 Abraham, J., Vasu, K.S., Williams, C.D., et al. (2017). Tunable sieving of ions using graphene oxide membranes. Nature Nanotechnology 12: 546–550. https:// doi.org/10.1038/nnano.2017.21. 216 Gao, Y., Song, J., Li, S., et al. (2016). Hydrogel microphones for stealthy underwater listening. Nature Communications 7: 12316. https://doi.org/10.1038/ ncomms12316. 217 Fish, F.E. and Kocak, D.M. (2011). Biomimetics and marine technology: an introduction. Marine Technology Society Journal 45: 8–13. 218 Becker, G. and Gudesen, A. (2000). Passive sensing with acoustics on the battlefield. Applied Accoustics 59: 149–178.

57

58

1 Introduction: Progress of Thin Films and Coatings

219 Asadnia, M., Kottapalli, A.G.P., Shen, Z.Y. et al. (2013). Flexible and surface-mountable piezoelectric sensor arrays for underwater sensing in marine vehicles. IEEE Sensors Journal 13: 3918–3925. 220 Sanchez, A., Blanc, S., Yuste, P. et al. (2012). An ultra-low power and flexible acoustic modem design to develop energy-efficient underwater sensor networks. Sensors 12: 6837–6856. 221 Xu, J., Headings, L.M., and Dapino, M.J. (2015). High sensitivity polyvinylidene fluoride microphone based on area ratio amplification and minimal capacitance. IEEE Sensors Journal 15: 2839–2847. 222 Zhou, Q., Zheng, J.L., Onishi, S. et al. (2015). Graphene electrostatic microphone and ultrasonic radio. Proceedings of the National Academy of Sciences of the United States of America 112: 8942–8946. 223 Polcawich, R.G., Scanlon, M., Pulskamp, J., et al. (2003). Design and fabrication of a lead zirconate titanate (PZT) thin film acoustic sensor. Integrated Ferroelectrics 54: 595–606. https://doi.org/10.1080/10584580390259010. 224 Zigoneanu, L., Popa, B.I., and Cummer, S.A. (2014). Three-dimensional broadband omnidirectional acoustic ground cloak. Nature Materials 13: 352–355. 225 Zhu, J., Chen, T., Liang, Q., et al. (2015). A unidirectional acoustic cloak for multilayered background media with homogeneous metamaterials. Journal of Physics D: Applied Physics 48: 305502. https://doi.org/10.1088/0022-3727/48/30/ 305502. 226 Song, X., Chen, T., Zhu, J. et al. (2019). Broadband acoustic cloaking and disguising with full-rangle incident angles based on reconfigurable metasurface. International Journal of Modern Physics B 33: 1950273. https://doi.org/10.1142/ S0217979219502734. 227 Nohara, S., Wada, H., Furukawa, N., et al. (2003). Electrochemical characterization of new electric double layer capacitor with polymer hydrogel electrolyte. Electrochimica Acta 48: 749–753. 228 Keplinger, C., Sun, J.-Y., Foo, C.C., et al. (2013). Stretchable, transparent, ionic conductors. Science 341: 984–987. https://doi.org/10.1126/science.1240228. 229 Wen, X.-Y., Liu, Z., Wang, J., et al. (2019). Nanocomposite hydrogels with optic–sonic transparency and hydroacoustic-sensitive conductivity for potential antiscouting sonar. ACS Applied Materials & Interfaces 11: 20386. https://doi .org/10.1021/acsami.9b04463. 230 Li, S., Zheng, J., Yan, J., et al. (2018). Gate-free hydrogel–graphene transistors as underwater microphones. ACS Applied Materials & Interfaces 10: 42573–42582. https://doi.org/10.1021/acsami.8b14034. 231 Zhou, Y., He, B., Yan, Z., et al. (2018). Touch locating and stretch sensing studies of conductive hydrogels with applications to soft robots. Sensors 18: 569. https://doi.org/10.3390/s18020569. 232 Zhang, X., Sheng, N., Wang, L., et al. (2019). Supramolecular nanofibrillar hydrogels as highly stretchable, elastic and sensitive ionic sensors. Materials Horizons 6: 326. https://doi.org/10.1039/C8MH01188E.

59

2 Fundamental in Functional Thin Films and Coatings Weiwei Zhang 1 and Yujun Song 1,2 1 University of Science and Technology Beijing, Center for Modern Physics Technology, Applied Physics Department, School of Mathematics and Physics, 30 Xueyuan Road, Beijing 100083, China 2 Zhejiang Key Laboratory for Pulsed Power Technology Translational Medicine, Hangzhou Ruidi Biotechnology Company, Hangzhou 310000, China

2.1 Introduction Films and coatings with multi-functions have novel physical and chemical properties resulting from the combination of their interfacial multi-physical field couplings (i.e. magneto-electric coupling [1–3], magneto-optical (MO) coupling [4, 5], magnetothermal coupling [6–9], and photoacoustic coupling [10, 11]). This field has been developed with new materials, processing, and applications being envisaged. In this chapter, we summarize the underlying mechanisms associated with thin films and coatings.

2.2 Theory of Magneto-electric Coupling in Magnetic Thin Films To date, various magnetoresistances (i.e. giant magnetoresistance (GMR), colossal magnetoresistance (CMR), tunneling magnetoresistance (TMR)) and quantum anomalous Holzer effect (QAHE) combining the electron spin and charge properties with magneto-electric coupling at mesoscopic sales are capable of modulating electron transports. GMR based on the dependence of electron scattering on the spin orientation is a quantum mechanical magnetoresistance effect, which significantly change electrical resistance depending on the magnetization of adjacent ferromagnetic layers is in a parallel or an antiparallel alignment in the presence of an external magnetic field. It is found that the overall resistance is relatively low for parallel alignment and relatively high for antiparallel alignment. Usually, the value of the GMR can be defined as MR =

𝜌(H) − 𝜌(0) Δ𝜌 = 𝜌(0) 𝜌(0)

Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications, First Edition. Edited by Yujun Song. © 2021 WILEY-VCH GmbH. Published 2021 by WILEY-VCH GmbH.

(2.1)

60

2 Fundamental in Functional Thin Films and Coatings

Here, 𝜌(H) represents the electrical resistivity in the presence of the applied magnetic field H, and 𝜌(0) is the electrical resistivity without the applied magnetic field. CMR effect associated with a ferromagnetic to paramagnetic phase transition under an external magnetic field is usually based on manganese-based perovskite oxides, and CMR can reach many orders of magnitude exceeding GMR of multilayered and granulated systems [12, 13]. TMR that depends dramatically on the alignment of the ferromagnets, in parallel or antiparallel, in magnetic tunnel junctions made up of ferromagnets/insulator/ferromagnets system. As shown in Eq. (2.2), the TMR effect is a typical example of the spin-dependent electron transport and can be evaluated by the ratio of conductance: TMR =

Gp − GAp Gp

(2.2)

where Gp and GAp are the conductances for the parallel and antiparallel alignments of the magnetic tunnel junctions, respectively. The value of TMR is mainly dependent on the intensity of two ferromagnets’ spin polarization at the Fermi energy. It is concluded that the conductance will be larger if they have the same sign of the spin polarization and the magnetic layers are aligned parallel, which is considered as the positive sign of the TMR. However, it is experimentally demonstrated that the conductance of the TMR can be widely decided by a broad range of atomic and electronic factors such as an insulator and the ferromagnet/insulator interfaces, materials of electrodes, the height, the shape and even the disorder of the barriers, and the impurities in the barrier. QAHE is a quantized version of the Holzer effect observed in two-dimensional (2D) electron systems under approximately 10 T magnetic field strength that makes the experimental realization challenging and greatly hinders real-world applications. However, QAHE, which may be a consequence of the combined spin–orbit coupling and reduction of the time-reversal symmetry (TRS) due to intrinsic magnetization, can be realized without an external magnetic field that may lead to the development of low-power-consumption electronics [14]. Nowadays, broad investigations have been carried out to seek new platforms for the realization of the QAHE [15–17]. Among the investigations, graphene-like honeycomb materials [18, 19] and magnetically doped topological insulators [14, 20] have mainly attracted increasing attention.

2.3 Theory of Electronic Thin Films: Electronic Percolation and Spintronic Theory on the Semiconductor Thin Film Spintronics refers commonly to phenomena, where the spin of electrons in a solid-state environment plays the determining role. Particularly, semiconductor spintronics as an emerging research discipline and an important advanced field in

2.3 Electronic Percolation and Spintronic Theory on the Semiconductor Thin Film

physics has developed quickly and obtained fruitful results and remarkable success in recent decades [21–24]. An important research topic in semiconductor spintronics is to use semiconductor devices that are based on the long spin coherence time of electron spin and nuclear spin to complete quantum information processing [25, 26]. However, a crucial problem in making spintronic devices is how to inject spinpolarized electrons from magnetic semiconductors into nonmagnetic ones without a strong magnetic field or at room temperature. Up to now, there are five main injection methods: ohmic injection, tunnel junction injection, ballistic electron spin injection, hot electron injection, and dilute magnetic semiconductor injection. Ohmic injection. The most direct spin injection structure for injecting spin-polarized current into a semiconductor is the ohmic contact formed by the ferromagnetic material/semiconductor. Since electrons are spin-polarized in the ferromagnetic material, it is desirable to inject spin-polarized electrons into the semiconductor. However, the typical metal–semiconductor ohmic contact requires heavy doping on the semiconductor surface, which will cause the spin-flip scattering of carriers and result in the loss of spin polarization. Because the semiconductor surface is heavily doped, spin inversion scattering and spin polarizability decrease. Therefore, the spin injection rate of this method is very low. Tunnel junction injection. The tunnel junction in the vacuum can effectively inject the spintronics into the semiconductor. High impedance ferromagnetic/ insulating layer/ferromagnetic structures have also demonstrated that spin polarization can be maintained during tunneling, suggesting that tunneling may be a more efficient spin injection method than diffusion transport. Hot electron injection. This method uses spin-polarized hot electrons to be injected into the ferromagnetic film through a tunnel junction, and the energy injected into the hot electrons is regulated by adjusting the bias of the tunnel junction. When the inelastic mean free path is different, the hot electrons passing through the ferromagnetic metal layer can produce an electron current with polarizability of over 90%. Depending on the transmission probability determined by the band structure of the semiconductor and metal at the interface, a highly polarized current can retain a considerable portion. If there is less spin-flip scattering at the interface, the ballistic current entering the semiconductor is still highly polarized. Spin detection technology. This method is also a crucial concept for spintronic devices. There are optical and electrical methods for spin detection. The optical detection method is relatively mature and has made great progress. The greatest advantage of the optical method is that it avoids other electrical effects. The most direct electrical method for detecting spin polarization is to utilize the spin-dependent transport properties of the semiconductor/ferromagnetic interface. However, the electrode in this method uses ohmic contact, and the conductivity mismatch still exists.

61

62

2 Fundamental in Functional Thin Films and Coatings

2.4 Theory of Metal Structural Thin Films: Metamaterials and the Negative Permeability Theory and Maxwell Theory A metamaterial is an artificially structured material that has extraordinary electromagnetic properties involving physics, electrical engineering, materials science, optics, and nanoscience. The properties of the metamaterial can be widely tailored by their internal physical structure design. An attractive interest in metamaterial is its effects on light propagating. Here we introduce some models to deal with the interaction between an electromagnetic wave and metamaterial. Jones vector and Jones matrix [27] compose a simple approach to model scatterers on metasurfaces. We assume a light wave propagating along the z-direction. As shown in Eq. (2.3), the incident electric field is composed of two components, x and y, and can be described by the Jones vector without time-harmonic factor: ( ) i (2.3) Ei (r, t) = x e−ikz iy Here, k is the wavenumber and the complex amplitudes ix and iy are the polarization states of the incident waves. After the incident wave impinging on the metasurface, the transmitted wave is described in the same manner: ( ) t Et (r, t) = x e−ikz (2.4) ty where tx and ty represent the polarization states along the x and y directions of the transmitted wave. Then, the localized scatterer on the metasurface can be expressed by Jones matrix, J. It connects the transmitted field components, tx, and ty to the incident ones: )( ) ( ) ( ) ( J J ix i tx = xx xy =J x (2.5) ty iy iy Jyx Jyy where the first and second subscripts of J denote the polarization states of the transmitted wave and incident wave, respectively. The reflected field can be modeled in the same way. Additionally, the Jones matrix, J can be transformed into a circular basis by coordinate transformation so that circularly polarized fields can be directly manipulated: ) ( J J Jc = ++ +− J−+ J−− ( ) 1 (Jxx + Jyy ) + i(Jxy − Jyx ) (Jxx − Jyy ) − i(Jxy + Jyx ) . (2.6) = 2 (Jxx − Jyy ) + i(Jxy + Jyx ) (Jxx + Jyy ) − i(Jxy − Jyx ) where Jc connects the incident circularly polarized Jones vectors to the transmitted or reflected circularly polarized ones. + and − represent the left circularly polarized and right polarized components, respectively. We should see that only the normal incidence and reflection/transmission are considered by using the Jones matrix. However, the coupling between scatterers is not taken into consideration rigorously. Polarizability model. Incident waves can bring polarization electric current, leading to the discontinuities of field components when crossing the metasurface

2.4 Metamaterials and the Negative Permeability Theory and Maxwell Theory

plane, which results in the conventional boundary conditions unable to describe the system that should be replaced by the generalized sheet transition conditions (GSTCs) [28]. n × (H1∕∕ − H2∕∕ ) = Jtoe = i𝜔P∕∕ − n × ∇∕∕ Mn P n × (E1∕∕ − E2∕∕ ) = −Jtom = −i𝜔𝜇0 M∕∕ − n × ∇∕∕ n 𝜀0 P ∕∕ n ⋅ (E1∕∕ − E2∕∕ ) = −∇ ⋅ 𝜀0 n ⋅ (H1∕∕ − H2∕∕ ) = −∇ ⋅ M∕∕

(2.7) (2.8) (2.9) (2.10)

where P and M are the surface electric and magnetic polarization densities, correspondingly. Subscripts “//” and “n” denote the tangential and normal components, respectively, while superscripts “1” and “2” refer to the fields at the two sides of the metasurface. Jtoe and Jtom denote the effective total electric and magnetic currents. Equations (2.6)–(2.9) provide us the information on the induced polarization density that transforms the impinging field in the desired manner, but still, it cannot give an intuitive insight of the metasurface design. To investigate this problem, models based on different homogenized parameters such as the polarizability [29], susceptibility [30], and equivalent impedance [31] have been proposed and demonstrated. Polarization density can be expressed as a form of the polarizability and the incident field [29]: P=

𝛼⌢ee 𝛼⌢ Ei + em Hi S S

(2.11)

𝛼⌢me 𝛼⌢ Ei + mm Hi (2.12) S S where Ei and Hi are the known incident fields, S is the area of the unit cell of the 𝜇M =

̂ is the effective polarizability dyadic, which represents the collective metasurface, 𝛼 effect of a single scatterer itself, together with the coupling and interaction from the whole metasurface array [31]. To synthesize a metasurface, one should first determine the polarization by substituting the known incident fields (Ei ,Hi ), desired ̂ components transmitted fields (Et ,Ht ), and reflected fields (Er ,Hr ), then find the 𝛼 through the obtained (P, M). For normal incident plane wave and the metasurface with uniaxial symmetry, a simple form of the polarizability and fields can be derived [29]. ) [( i𝜔 1 co co cr cr 𝜂0 𝛼 Er = − ̂ee +𝛼 ̂em +𝛼 ̂me − 𝛼 ̂mm I t 2S 𝜂0 ) ] ( 1 cr cr co co + 𝜂0 𝛼 ̂ee −𝛼 ̂em −𝛼 ̂me − 𝛼 ̂mm J t ⋅ Ei (2.13) 𝜂0 ( )) (( i𝜔 1 co co cr cr 𝜂0 𝛼 Et = 1− ̂ee + 𝛼 ̂em − 𝛼 ̂me + 𝛼 ̂ It 2S 𝜂0 mm ( ) ) i𝜔 1 cr cr co co 𝜂0 𝛼 − ̂ee −𝛼 ̂em +𝛼 ̂me + 𝛼 ̂ J Ei (2.14) 2S 𝜂0 mm t

63

64

2 Fundamental in Functional Thin Films and Coatings

where I t = I − z ⋅ z is the tangential unit dyadic, while J t = z × I t is the transverse rotation dyadic. Superscripts “co” and “cr” mean the symmetric and anti-symmetric components of 𝛼 ̂, respectively ⎧ ⌢cr ̂ee = 𝛼⌢co ee I t + 𝛼 ee J t ⎪𝛼 ⎨ ⌢co ⌢cr ⎪𝛼 ⎩̂em = 𝛼 em I t + 𝛼 em J t

(2.15)

⎧ ⌢cr ̂me = 𝛼⌢co me I t + 𝛼 me J t ⎪ 𝛼 (2.16) ⎨ ⌢co ⌢cr ⎪𝛼 ̂ = 𝛼 I + 𝛼 J mm t mm t ⎩ mm Susceptibility model. Alternatively, the polarizability of the metasurface can be homogenized by surface susceptibility. Polarization density is described by [30]. √ (2.17) P = 𝜀𝜒 ee Eav + 𝜇𝜀 𝜒 em Hav √ 𝜀 𝜒 E (2.18) M = 𝜒 mm Hav + 𝜇 me av u = [Eu + (Eu + Eu )]∕2 and H u = [H u + (H u + H u )]∕2 (u = x, y, z) where Eav r av r t t i i Closed-form relation of the fields and the susceptibility tensor can be obtained for the simplified case, assuming only tangential components of the polarizations are induced, so that Pz = Mz = 0. Substitute Eqs. (2.17) and (2.18) into Eqs. (2.7)–(2.10), it leads to ( xx ( ( xx y) xy ) ( x ) xy ) ( x ) √ 𝜒ee 𝜒ee Eav Hav −HΔ 𝜒em 𝜒em = i𝜔𝜀 + i𝜔 𝜀𝜇 (2.19) yx yy yx yy HxΔ 𝜒ee 𝜒ee Eyav 𝜒em 𝜒em Hyav ( xx ( y ) ( xx xy ) ( x ) xy ) ( x ) √ 𝜒mm 𝜒mm Hav Eav EΔ 𝜒me 𝜒me = i𝜔𝜇 + i𝜔 (2.20) 𝜀𝜇 yx yy y yx yy y x −EΔ 𝜒mm 𝜒mm Hav 𝜒me 𝜒me Eav

with EΔu = Etu − (Eiu + Eru ) and HΔu = Htu − (Hiu + Hru ) denoting the differences of the fields at the two sides of the metasurface, correspondingly. To this point, the susceptibility tensor matrix still cannot be completely determined for a specified field transformation. The number of unknown matrix components should be reduced by enforcing some extra conditions. For example, to monoanisotropic and uniaxial xy yx xy yx medium, so that 𝜒 em = 𝜒 me = 0 and 𝜒ee = 𝜒ee = 𝜒mm = 𝜒mm = 0, Eqs. (2.19) and (2.20) degrades to a simple relation. y

xx = 𝜒ee

yy = 𝜒ee

−HΔ x i𝜔𝜀Eav

HΔx y

i𝜔𝜀Eav

(2.21)

(2.22)

y

xx 𝜒mm =

yy = 𝜒mm

EΔ x i𝜔𝜇Hav

−EΔx y i𝜔𝜇Hav

(2.23)

(2.24)

2.4 Metamaterials and the Negative Permeability Theory and Maxwell Theory

Figure 2.1 Equivalent transmission line model (T-circuit) of a metasurface. Source: Asadchy et al. [32].

i2

i1 Z1 V1

Z2 Z3

V2

Therefore, the metasurface can be synthesized according to the desired fields on the two sides. Equivalent Impedance Model. The equivalent impedance model based on the transmission line theory is also a powerful method for the metasurface design [31, 32]. Impinging plane waves to the metasurface is an analogy to a propagating signal in a transmission line with proper equivalent parameters. Metasurface described by Eqs. (2.7)–(2.10) that can be modeled by a T-circuit as shown in Figure 2.1, with the equivalent impedance matrix that connects the voltages and currents by ( 1) ( ) ( 1) v Z11 Z12 i = (2.25) v2 Z21 Z22 i2 where Z 11 = Z 1 + Z 3 , Z 22 = Z 2 + Z 3 , Z 12 = Z 21 = Z 3 , linking the tangential fields at the two sides of the metasurface to the voltages and currents of the transmission line: ( ) ( ) )( E1∕∕ n × H1∕∕ Z11 Z12 = (2.26) E2∕∕ Z21 Z22 −n × H2∕∕ It has been reported that metamaterial with negative-permeability has been widely investigated [33, 34], which means that the solutions of Maxwell magnetostatic equations to include negative permeability values are extended. The understanding of these new solutions allows us to devise a negative-permeability material as a suitably tailored set of currents arranged in space, overcoming the fact that passive materials with negative permeability do not exist in magnetostatics, making the metamaterials particularly attractive. It is known that in the static case, electric and magnetic fields decouple, which means controlling magnetic fields requires only dealing with permeabilities. However, natural materials exist with extreme permeability values, such as 𝜇 → 0 and 𝜇 → ∞. In contrast, it is very difficult to fabricate materials with zero permittivity and only approximate results can be achieved based on resonances. However, some advantages of the full electromagnetic case have not yet a counterpart in magnetostatics. One of these is the possibility of having negative –𝜇 materials, whereas resonances in different kinds of natural and artificial substances can yield negative values of 𝜇 and 𝜀 at nonzero frequencies [35], no such negative –𝜇 materials exist in magneto-statics. Negative values of 𝜇 and 𝜀 have enabled very interesting novel phenomena for electromagnetic waves [36]. Devising ways to create the effective response of negative −𝜇 materials would pave the way toward the realization of these properties and also for static magnetic fields.

65

66

2 Fundamental in Functional Thin Films and Coatings

2.5 Theory of Surface Plasmon Resonance and Magnetoplasmonic Thin Films Surface plasmon resonance (SPR) is the resonant oscillation of conduction electrons at the interface between a negative and positive permittivity material stimulated by incident light. The resonance condition is established when the frequency of incident photons matches the natural frequency of surface electrons oscillating against the restoring force of positive nuclei. SPR in subwavelength scale nanostructures can be polaritonic or plasmonic in nature. SPR is the basis of many standard tools for measuring adsorption of material onto planar metal (typically gold or silver) surfaces or onto the surface of metal nanoparticles. It is the fundamental principle behind many color-based biosensor applications and different lab-on-a-chip sensors. Since the wave is on the boundary of the conductor and the external medium (for example, air, water, or vacuum), these oscillations are very sensitive to any change of this boundary, such as the adsorption of molecules to the conducting surface. Localized surface plasmon resonances (LSPRs) are collective electron charge oscillations in metallic nanoparticles that are excited by light. They exhibit enhanced near-field amplitude at the resonance wavelength. This field is highly localized at the nanoparticle and decays rapidly away from the nanoparticle/dielectric interface into the dielectric background, though far-field scattering by the particle is also enhanced by the resonance. Light intensity enhancement is a very important aspect of LSPRs and localization means the LSPR has a very high spatial resolution (subwavelength), limited only by the size of nanoparticles. Because of the enhanced field amplitude, the MO effect that depends on the amplitude is also enhanced by LSPRs. In order to excite surface plasmons in a resonant manner, one can use an electron bombardment or incident light beam (visible and infrared ranges are typical). The incoming beam has to match its momentum to that of the plasmon. [4] In the case of p polarized light (polarization occurs parallel to the plane of incidence), this is possible by passing the light through a block of glass to increase the wavenumber (and the momentum) and achieve the resonance at a given wavelength and angle. An s polarized light (polarization occurs perpendicular to the plane of incidence) cannot excite SPR. In order to excite SPR waves, Otto and Kretschmann setups are two main configurations. As shown in Figure 2.2, in the Otto setup, the light illuminates the wall of a glass block, typically a prism, and is totally internally reflected. A thin metal film is positioned close enough to the prism wall so that an evanescent wave can interact with the plasma waves on the surface to excite the SPR.

Prism

Prism

Kretschmann

Otto (a)

(b)

Figure 2.2 Schematics of (a) Otto configuration and (b) Kretschmann configuration.

2.5 Theory of Surface Plasmon Resonance and Magnetoplasmonic Thin Films

(a)

(b)

(c)

Figure 2.3 (a) One-dimensional strip grating; (b) two-dimensional nanorod grating; and (c) two-dimensional hexagonal nanoarray.

In the Kretschmann configuration, the metal film is evaporated onto the glass block. The light again illuminates the glass block, and an evanescent wave penetrates through the metal film. The SPR is excited at the outer side of the film. This configuration is used in most practical applications. Another method is to use the grating to realize the excitation of SPR. The basic mechanism of grating coupling is to use the increment of wave vector generated by grating diffraction. The wave vector can be given as [37] −−−−−⇀ − ⇀ − ⇀ (2.27) Rel(Kspp ) = |k=incident + mGx + nGy | −−−−−⇀ where k=incident is the horizontal component of the incident wave, m and n are the − ⇀ − ⇀ diffraction series, and Gx and Gy are the base vectors. Here we introduce three main grating nanostructures whose schematic are shown in Figure 2.3. The light is incident to the surface of the grating from the medium with a refractive index n, incident angle 𝜃, and azimuth angle 𝜑. The relative position of the incident wave vector and the grating is shown in Figure 2.4a (𝜃 ≠ 0, 𝜑 = 0). Here the wave vector matching condition is given as | 2π | (2.28) Rel(Kspp ) = |k0 n sin 𝜃 + mG | = ||k0 n sin 𝜃 + m || p| | As shown in Figure 2.4b, when 𝜃 ≠ 0, 𝜑 ≠ 0, the wave vector matching condition can be shown as | 2𝜋 | Rel(Kspp ) = ||k0 n sin 𝜃 cos φ⃑i + k0 n sin 𝜃 sin φ⃑j + m ⃑i|| (2.29) p | | where ⃑i and ⃑j are the unit vectors in the x and y directions respectively, and a and b are the periods in the unit vector directions.

67

68

2 Fundamental in Functional Thin Films and Coatings

z

z

y

o

y

o x

x

θ

θ φ

(a)

(b)

Figure 2.4 The relative position of the incident wave vector and the grating. (a) azimuth angle is 0 and (b) azimuth angle is not 0.

Up to now, SPR has been used to enhance the surface sensitivity of several spectroscopic measurements including fluorescence, Raman scattering, second harmonic generation, and MO effects. MO effects can be described by the interaction of incident photons with the electron spins through spin–orbit coupling that has the general form. Hso =

h P ⋅ (𝝈 × ∇V) 8𝜋c2 m20

(2.30)

where m0 is the free electron mass, c is the velocity of light, 𝝈 = 𝝈(x, y, z) is a vector, which components are the Pauli matrices, V is the electric potential, and P represents the canonical momentum. In the presence of an external magnetic field, B = r × A, P should be replaced by the kinetic momentum P = p + eA. But MO effects can be completely explained by the classical magnetic properties of the material. For instance, MO effects in an anisotropic material that can be conveniently expressed in terms of its relative dielectric permittivity, where 𝜀 is a 3 × 3 tensor as follows: ⎡𝜀xx 𝜀xy 𝜀xz ⎤ 𝜀 = ⎢𝜀yx 𝜀yy 𝜀yz ⎥ ⎥ ⎢ ⎣𝜀zx 𝜀zy 𝜀zz ⎦

(2.31)

Here, the subscripts x, y, z represent the three Cartesian coordinate axes. For a linear anisotropic material, it is further reduced to just a diagonal tensor as follows: ⎡𝜀xx 0 0 ⎤ 𝜀 = ⎢ 0 𝜀yy 0 ⎥ ⎥ ⎢ ⎣ 0 0 𝜀zz ⎦

(2.32)

However, in the presence of an internal magnetization M pointing, for instance, in the y-direction, the off-diagonal elements of the permittivity tensor in the x- and z-direction become coupled. Microscopically, this can be explained using classical electron theory. When an incident electromagnetic wave (EM) field illuminates

2.5 Theory of Surface Plasmon Resonance and Magnetoplasmonic Thin Films

a material, a Lorentz force acts on the material’s electrons and consequently, an oscillation in one direction tends to create another oscillation in its transverse direction manifesting itself as nonzero off-diagonal elements. The material’s permittivity gets modified as: 0 ⎤ ⎡𝜀xx 0 𝜀 = ⎢ 0 𝜀yy −ig⎥ ⎥ ⎢ ⎣ 0 ig 𝜀zz ⎦

(2.33)

where the off-diagonal elements represent the magnetically-induced part and are directly proportional to M. g is referred to as the MO constant of a material. Its value varies from material-to-material and is a function of the incident wavelength. g = g(𝜆, M)

(2.34)

The tensor elements of the permittivity tensor are complex in general. For an isotropic material at room temperature, the diagonal elements can be expressed as a square of the complex refractive index of the material. 𝜀xx = 𝜀yy = 𝜀zz = N 2

(2.35)

N = n + ki

(2.36)

Here, n represents the dispersion and k represents the extinction coefficient of the material. Here, we will mainly summarize the mechanism of Kerr and Faraday effects. Magneto-optical Kerr effect (MOKE): MO effects in reflection mode are called MOKE and have the same physical origin as the Faraday effect. They lead to a change of the polarization properties and/or of the light intensity when a linearly polarized light gets reflected from a magnetized medium. When a linearly polarized optical wave travels through or is reflected by an isotropic, non-magnetic medium; it could be in either of its two eigenmodes (waves that preserve their polarization while propagating through a medium). However, when it is incident on a magnetized medium, the off-diagonal elements of the permittivity tensor of this medium induce a coupling between the two polarizations or modify the reflectivity depending on the magnetization orientation. There are three configurations of the MOKE based on the direction of magnetization. We describe the different forms of dielectric tensors of three different MOKE. It is assumed that the sample plane is parallel to the xoy plane, and the incident plane of the light is parallel to the xoz plane. Table 2.1 summarizes the MO Kerr effect configurations with a schematic depiction of each effect. Polar magneto-optical Kerr effect (P-MOKE): The magnetization direction of the material caused by the external magnetic field is perpendicular to the sample surface and parallel to the incident surface. In this case, the dielectric tensor can be expressed as: ⎡𝜀xx −gi 0 ⎤ 𝜀 = ⎢ gi 𝜀yy 0 ⎥ ⎥ ⎢ 0 𝜀zz ⎦ ⎣0

(2.37)

69

70

2 Fundamental in Functional Thin Films and Coatings

Table 2.1

The different configurations of MOKE.

MOKE and its diagram z

P-MOKE

M

M

M y o

z

z x

L-MOKE

y o

x

T-MOKE

y o

Tensor

⎡𝜀xx −gi 0 ⎤ ⎢ ⎥ 𝜀 = ⎢ gi 𝜀yy 0 ⎥ ⎢ ⎥ 0 𝜀zz ⎦ ⎣0

⎡𝜀xx 0 0 ⎤ ⎢ ⎥ 𝜀 = ⎢ 0 𝜀yy −gi⎥ ⎢ ⎥ ⎣ 0 gi 𝜀zz ⎦

⎡𝜀xx 0 −gi⎤ ⎢ ⎥ 𝜀 = ⎢ 0 𝜀yy 0 ⎥ ⎢ ⎥ ⎣ gi 0 𝜀zz ⎦

Effects

Rotation and ellipticity

Rotation and ellipticity

Change of intensity

x

For the P-MOKE, both s polarized light and p polarized light have an electric field component in the vertical direction of the magnetic field, so both s polarized light and p polarized light can produce Kerr angle and ellipticity, and their magnitude depends on the angle of incident light and the magnitude of the magnetic field that determines the magnitude of g. The P-MOKE can be expressed as: ] [√ (Nr 2 − tan2 𝜃in ) ∓ sin 𝜃in tan 𝜃in Nr 2 gi 𝛹= (2.38) 𝜀xx (Nr 2 − 1)(Nr 2 − tan2 𝜃in ) where, N r is the ratio of the refractive index of the magnetic medium to the refractive index of the medium. In this article, we assume that in the ∓ and ± symbols, the upper sign represents s polarized light and the lower indicates the p polarized light. The polar Kerr deflection angle and ellipticity can be expressed as: { 𝜃 = Rel(𝛹 ) (2.39) 𝜑 = Im(𝛹 ) In general, when studying the P-MOKE, the incident ray from the air (refractive index 1), the incidence angle is usually close to 0∘ , and for isotropic substances (namely 𝜀xx = 𝜀yy = 𝜀zz = N 2 ), Eq. (2.38) can be simplified to 𝛹= √

gi 𝜀xx (𝜀xx − 1)

(2.40)

Longitudinal magneto-optical Kerr effect (L-MOKE): the direction of magnetization caused by the external magnetic field is in the incident plane and parallel to the surface of the sample. The dielectric tensor of the L-MOKE can be given as: 0 ⎤ ⎡𝜀xx 0 𝜀 = ⎢ 0 𝜀yy −gi⎥ ⎥ ⎢ ⎣ 0 gi 𝜀zz ⎦

(2.41)

Different from the P-MOKE, the incident angle of the L-MOKE is generally not 0, and the change of the Kerr angle is also greatly related to the incident angle. The

2.5 Theory of Surface Plasmon Resonance and Magnetoplasmonic Thin Films

L-MOKE angle and ellipticity can be expressed as: [ ] √ ⎧ ⎛ sin 𝜃in Nr 2 gi sin 𝜃in tan 𝜃in ± (Nr 2 −tan2 𝜃in ) ⎞ ⎪𝜃 = Rel ⎜ ⎟ √ ⎪ ⎜ 𝜀xx (Nr 2 −1)(Nr 2 −tan2 𝜃in ) (Nr 2 −sin2 𝜃in ) ⎟ ⎝ ⎠ ⎪ [ ] ⎨ √ ⎛ sin 𝜃in Nr 2 gi sin 𝜃in tan 𝜃in ± (Nr 2 −tan2 𝜃in ) ⎞ ⎪ ⎟ ⎪ 𝜑 = Im ⎜ √ ⎜ 𝜀xx (Nr 2 −1)(Nr 2 −tan2 𝜃in ) (Nr 2 −sin2 𝜃in ) ⎟ ⎪ ⎝ ⎠ ⎩

(2.42)

Transverse magneto-optical Kerr effect (T-MOKE): the direction of magnetization caused by an external magnetic field is parallel to the surface of the sample but perpendicular to the light incident surface. The primitive tensor can be expressed as: ⎡𝜀xx 0 −gi⎤ 𝜀 = ⎢ 0 𝜀yy 0 ⎥ ⎢ ⎥ ⎣ gi 0 𝜀zz ⎦

(2.43)

From the relative position relationship between the incident plane, the sample surface and the plane where the magnetization direction is located in the T-MOKE system, we can see that when the incident light is an s polarized light, the magnetization direction is parallel to the polarization direction, so the T-MOKE device will not affect the s polarized light wave. The p polarized light doesn’t produce s polarization component that is parallel to the magnetic field, and it just changes the reflectivity of p-polarized light. The following equation is usually used to value the intensity of the T-MOKE. I+ − I− R+ − R− = (2.44) 𝛿= R0 I0 R+ and I + represent the reflectivity and intensity of reflected light when the magnetic field direction is positive. R− and I − represent the reflectivity and reflected light intensity of samples with the magnetic field direction being the opposite. R0 and I 0 represent the reflectivity and reflected light intensity of samples with demagnetization. To reflect the relationship between the T-MOKE and the incident angle of the light wave, the intensity of the T-MOKE can be a function of the incident angle. ) ( 4(tan 𝜃in )Nr 2 gi (2.45) 𝛿 = Rel 𝜀xx (Nr 2 − 1)(Nr 2 − tan2 𝜃in ) Faraday effect: MO effects in transmission mode. Another important MO phenomenon is the Faraday effect or Faraday rotation. The Faraday effect causes a rotation of the plane of polarization that is linearly proportional to the component of the magnetic field in the direction of propagation. An application of an external magnetic field to the sample (Figure 2.5) can lead to magneto-induced anisotropy and to the appearance of additional off-diagonal terms in the electric permittivity. The static magnetic field is perpendicular to the surface and along the z-axis. The direction of the wavevector is along the −z-direction and the polarization direction of the electric field is parallel to the x-axis In this configuration, the external magnetic field is perpendicular to the surface and along the z-axis. The direction of the wavevector is along the −z-direction and the polarization direction of the electric field is parallel to the x-axis. Here,

71

72

2 Fundamental in Functional Thin Films and Coatings

θ M

Figure 2.5 Schematic drawing of Faraday configuration.

z X

we assume that the MO medium is isotropic and 𝜀xx = 𝜀yy = 𝜀zz = 𝜀f . Then, the permittivity tensor can be expressed as follows. ⎡𝜀f −gi 0 ⎤ 𝜀 = ⎢ gi 𝜀f 0 ⎥ ⎥ ⎢ ⎣ 0 0 𝜀f ⎦

(2.46)

We are looking for a solution to Maxwell’s equations in the form of a monochromatic plane wave with the frequency 𝜔 propagating along the −z axis, with electric field: ⃑ =E ⃑ 0 e−i(kz z−𝜔t) E Substituting this into Maxwell’s equations we obtain 1 ⃑ ) = i𝜔𝜀 E ⃑ − ∇ × (∇ × E i𝜔𝜇

(2.47)

(2.48)

This gives the following values for the wave vectors kzL and kzR in the transmission. √ √ 2𝜋 𝜀f −g ⎧ L ⎪kz = 𝜔 𝜇(𝜀f − g) = √𝜆 √ (2.49) ⎨ R 2𝜋 𝜀f +g k = 𝜔 𝜇(𝜀 + g) = ⎪ z f 𝜆 ⎩ The phase difference through thickness L can be expressed as √ √ 2𝜋L( 𝜀f + g − 𝜀f − g) R L (2.50) 𝛿 = (kz − kz )L = 𝜆 This complex phase-difference makes the polarization plane rotated by a certain angle. This is known as the Faraday effect. The Faraday angle is expressed as: 𝜃F = Rel(𝛿∕2)

(2.51)

The expression for the ellipticity (the ratio of minor to major axis) 𝜑F can be given as: ( ( )) 𝛿 𝜑F = − tanh Im (2.52) 2 Replacing 𝜀f and g by the refractive elements using equation N± = 𝜀f ± g we get the following expressions for Faraday effect: ) ( 𝜋L(N+ − N− ) 𝜃F = Rel 𝜆 ) ( 𝜋L(N+ − N− ) 𝜑F = − tanh 𝜆

(2.53)

(2.54) (2.55)

2.6 Heterojunction Theory

M

X

θ

z θ 2θ

Figure 2.6

Mechanism of the magneto-optical isolator based on the Faraday effect.

A very useful property of the Faraday effect is the irreversibility of its optical path. Its mechanism schematic is shown in Figure 2.6. When the incident light passes through the MO medium, the polarization direction of the transmitted light deflects by an angle of theta from the original direction. When the transmitted light is reflected back to the original medium, the angle of deflection will not change to 0 but to 2 thetas with respect to the original polarization direction. This special property can be used in MO isolators to prevent reflected light from entering the laser and in one-way communication devices. To quantify the intrinsic Faraday effect activity of a material, it is convenient to define a new parameter called specific Faraday rotation Q ( ∘∕cm) that can be expressed as: Q=

𝜃F L

(2.56)

2.6 Heterojunction Theory Heterojunction with an interface between two different materials consisting of unequal bandgaps is composed of two or more semiconductor materials [38, 39]. Nowadays, heterojunction is a distinctly important component in electronics, transistors, photovoltaic cells, diodes, and sensors, memory devices, photodetectors, and optoelectronic devices that have made remarkable changes in our daily lives. A well-defined heterojunction can be formed using different materials with a tight interface and its physicochemical properties can be tailored to obtain desirable functionalities. Based on the various materials, heterostructures can be classified into five categories [40]. They are metal or semimetal/semiconductor, semimetal/insulator-based, semiconductor/semiconductor, semiconductor/insulator, and metal/semimetal-based heterojunctions. However, according to the arrangement and interface of the different materials, heterostructures can be classified into four categories: (1) spherical zero-dimensional (prominent heterostructures include nanoclusters, nanodispersions, quantum dots, quantum wells, and core–shell structures [41]), (2) cylindrical one dimensional, (3) planar two dimensional, and (4) cubic three dimensional.

73

74

2 Fundamental in Functional Thin Films and Coatings

Various heterostructures show different synergistic relations between two or more building blocks that improve functional characteristics. To enhance the physical properties of heterostructures, crucial issues are the design and synthesis of complex heterostructures with the controlled assembly of each section of materials, including the size, shape, and uniformity of the building blocks. Fundamental understanding of the systematic mechanisms is needed to understand the morphological evolution of heterostructures, which could be highly desirable for developing effective heterostructured devices. Among the various heterojunctions, the abrupt and isotype heterojunction is found to be a good approximation for many heterojunctions. The mechanisms for charge carrier transport in abrupt an isotype heterojunctions have been explored by various models including diffusion model, emission model, emission-recombination model, tunneling model, tunneling-recombination model [40]. The schematic of diffusion model with a typical energy band profile of two isolated pieces of p- and n-type semiconductors and an equilibrium energy band profile of an abrupt p–n heterojunction formed by bringing them into intimate contact are shown in Figure 2.7a,b. In this model, the two semiconductors have different energy gaps (Eg ), dielectric constants (𝜀), work functions (𝜑) and electron affinities (𝜒). Neglecting the generation-recombination current, the predicted current–voltage relation is given by )[ ( ) ( )] ( qV2 qV1 qVD2 exp − exp − (2.57) I = A exp − KT KT KT where V D2 and V 2 are the portions of the applied voltage appearing in p- and n-type semiconductors, K is the Boltzmann constant, T is the absolute temperature. Although this model is invariably used to predict the energy band diagrams of heterojunctions, neither the voltage nor the temperature dependences observed by various workers are adequately described. The emission model is the combination of a model for the evaluation of emission currents and a diffusion model. The emission-recombination model is according to two considerations: there is a thin layer at the interface with a strongly disturbed lattice and fast recombination and the electrons and holes reach the interface via thermal emission over their respective barriers; tuning model is according to the consideration that the electrons can go through the potential barrier in the n-type wide-bandgap material; in tunneling-recombination model, either the electrons tunnel from the conduction band of the wide-bandgap material into empty interbond states located in the narrow-band gap material, then recombine with holes, or holes tunnel from p-type material into occupied states in the n-type material, then recombine with electrons.

2.7 Topological Insulator, Topological Semi-metal, and Perovskite Topological insulators (TIs), a class of materials that present unique quantummechanical properties and a new quantum state of matter, which is characterized

2.7 Topological Insulator, Topological Semi-metal, and Perovskite

75

Electron energy

Vacuum level

χ2

ϕ2 χ1

ϕ1

EC2 ΔEC

EC1

Fermi level 2

Eg1

Eg2 Fermi level 1

EV1

ΔEV

EV2

(a)

Electron energy

Vacuum level VD = ϕ1 – ϕ2

χ1 EC1

ϕ1

“NOTCH”

“SPIKE” VD1

ΔEC

VD2

ϕ2

χ2 EC2

EV1

Fermi level ΔEV EV2 X1 X0

X2

(b)

Figure 2.7 Equilibrium energy band diagrams (a) before and (b) after the formation of an abrupt heterojunction. Source: Sharma and Purohit [40].

by peculiar edge or surface states that show up due to a topological character of the bulk wave functions, are particularly important. TIs concern a qualitatively new aspect of quantum mechanics, i.e. the topology of the Hilbert space, they opened a new window for understanding the elaborate workings of nature and wave functions describing their electronic states span a Hilbert space that has a nontrivial topology. A large part of the unique quantum-mechanical properties of TIs come from the peculiar characteristics of the edge/surface states. Currently, the TI research is focused mostly on time-reversal (TR) invariant systems, where the nontrivial topology is protected by TRS. In those systems, the edge/surface states present Dirac dispersions (Figure 2.8), hence the physics of relativistic Dirac fermions becomes relevant. Furthermore, spin degeneracy is lifted in the Dirac fermions residing in the edge/surface states of TR-invariant TIs and their spin is locked to the momentum. Such a spin state is said to have “helical spin polarization”

76

2 Fundamental in Functional Thin Films and Coatings

E

E k

(a)

Figure 2.8 (a) Surface state ordinary insulator and (b) surface state topological insulator. k

(b)

and it provides an opportunity to realize Majorana fermions in the presence of proximity-induced superconductivity, not to mention its obvious implications for spintronics applications. As shown in Figure 2.8, the electron motion of the topological insulator does not conform to the usual metallic electron dispersion relation E = k2 /2m, but E = v * k. Here, v is the velocity of the electron motion when k0 = 0. Notice that, for the light, E is equal to light velocity c times k. In this regard, the electron does not like non-relativistic particles, but like a light. This property makes the electrons respond sensitively to external electric fields and they can be used as the basis for semiconductor devices such as field-effect tubes. Topological insulators are a very special type of insulators. Due to the spin-orbit interaction (relativistic effect), there is always a massless Dirac-type electronic state on the surface of such insulators, so the surface is always metallic and conductive. Electrons are carriers of electricity, and in addition to having an electric charge, they have a property of spin, just as the earth goes around the sun, and the earth itself spins. In today’s widely used computer chips, the spin state of the electron is uncertain. As the electrons travel from one end of the transistor to the other, it is as if they were going through a disarranged disco. Inevitably, they collide, which heats up and dissipates heat. In topological insulators, there is a definite relation between the direction of electron rotation and the direction of the current. Electrons moving in different directions behave like cars on a highway, each electron does not interfere with the other, resulting in low energy dissipation. The unique properties of topological insulators make them potentially important in the fields of low-energy electronic devices and fault-tolerant quantum computing. Therefore, the discovery of topological insulators quickly aroused the great interest of scientists in physics, material science, and other fields. Topological semi-metal is a new class of topological electronic states, which are different from topological insulators. The band crossing degenerate point, the Weyl node, which happens to sit on the Fermi surface, gives a very special class of topological semi-metals. Topological metals have special energy band structures that contain some singularities of energy band structures. In short, the intersection point with two energy bands can be described by the relativistic Weyl equation with chirality. Completely different from two-dimensional space (e.g. graphene), in three-dimensional momentum space, such band intersections are a very stable topological structure and cannot introduce mass terms, that is, they cannot open the energy gap through perturbation, so they are very stable. Such band crossing degeneracy points are called Weyl nodes. If we examine the Weyl node in detail,

2.8 Acoustic Theory

we will find that there are two completely different types of Weyl nodes, which can be described as ± symbols in Hamiltonian, corresponding to left-handed and right-handed Weyl nodes, so they are topologically different. When a left-handed and a right-handed Weyl node coincide in the momentum space, it needs to be described by the 4 × 4 Dirac equation. Such a 4∘ degeneracy point is called a three-dimensional Dirac node, and its existence requires the protection of crystal symmetry (since mass terms can be introduced in the 4 × 4 equation). In most metal materials, such a Weyl/Dirac node, would be far from the Fermi surface, but if such a Weyl/Dirac node happened to be located on the Fermi surface, it would give rise to a very special class of electronic structures: “topological semi-metals” whose Fermi surface is reduced to Fermi points, with zero energy gap, and linear dispersion. Such topological semi-metallic states will exhibit wonderful physical properties, for example, their surface states have Fermi arcs, their bodies have magnetic monopoles in momentum space, unique transport properties, magnetism, and so on. Perovskite is a calcium titanium oxide mineral composed of calcium titanate (CaTiO3 ). Its name is also applied to the class of compounds, which have the same type of crystal structure as CaTiO3 , known as the perovskite structure. Many different cations can be embedded in this structure, allowing the development of diverse engineered materials. Perovskites have a cubic structure with the general formula of ABO3 . In this structure, an A-site ion, on the corners of the lattice, is usually alkaline earth or rare-earth element. B-site ions, on the center of the lattice, could be 3d, 4d, and 5d transition metal elements. A large number of metallic elements are stable in the perovskite structure if the Goldschmidt’s tolerance factor t is in the range of 0.75–1.0. R + RO (2.58) t= √ A 2(RA + RO ) where RA , RB , and RO are the ionic radii of A and B site elements and oxygen, respectively [42]. Perovskites have sub-metallic to metallic luster, colorless streak, cube-like structure along with imperfect cleavage and brittle tenacity. Colors include black, brown, gray, orange to yellow. Crystals of perovskite appear as cubes, but are pseudocubic and crystallize in the orthorhombic system. Perovskite crystals have been mistaken for galena; however, galena has a better metallic luster, greater density, perfect cleavage, and true cubic symmetry [43]. Perovskites are a class of materials that share a similar structure, which displays a myriad of exciting properties like superconductivity, magnetoresistance, and more. These easily synthesized materials are considered the future of solar cells, as their distinctive structure makes them perfect for enabling low-cost, efficient photovoltaics. They are also predicted to play a role in next-gen electric vehicle batteries, sensors, lasers and much more.

2.8 Acoustic Theory It is assumed that the medium is non-viscous and the propagation of sound waves in the medium is lossless. In the case of silent disturbance, the medium is

77

78

2 Fundamental in Functional Thin Films and Coatings

macroscopically stationary and its initial velocity is zero. The static pressure P0 and the static density 𝜌0 are constants and sound waves travel adiabatically. For the multidimensional, the wave equations can be given as: ⎧ ∇2 p = ⎪ ⎪ ⎪ 2 ′ ⎨∇ 𝜌 = ⎪ ⎪ ⎪ ∇2 v = ⎩

1 𝜕2 p c0 𝜕t2 1 𝜕 2 𝜌′ c0 𝜕t2

(2.59)

1 𝜕2 v c0 𝜕t2

Here we focus on the propagation of sound waves in one dimension and the wave equations can be given as: 2 2 ⎧𝜕 p = 1 𝜕 p ⎪ 𝜕x2 c0 2 𝜕t2 ⎪ ⎪ 𝜕 2 𝜌′ 1 𝜕 2 𝜌′ ⎨ 2 = 2 2 c0 𝜕t ⎪ 𝜕x ⎪ 2 2 ⎪ 𝜕 v = 1 𝜕 v 2 ⎩ 𝜕x2 c0 𝜕t2

(2.60)

Here, p is the acoustic pressure, 𝜌′ is the density increment, and 𝜈 is the velocity at the x position. The solution to the wave equation can be given as: p(x, t) = Aej(𝜔t−kx) + Bej(𝜔t+kx)

(2.61)

If there is no barrier, the acoustic pressure p(x, t), velocity v(x, t) and displacement 𝜉 can be given as: p(x, t) = pa ej(𝜔t−kx)

(2.62)

j(𝜔t−kx)

v(x, t) = va e (2.63) ) ( 𝜋 v −j kx0 + j𝜔t 2 e 𝜉 = ae = 𝜉a ej(𝜔t−𝛼) (2.64) 𝜔 The particle at position x0 just moves back and forth near the equilibrium position but does not move a distance. In fact, it is this kind of back and forth the vibration of the media particle near the equilibrium position that affects the surrounding and even further media particles to vibrate back and forth near the equilibrium position, thus spreading out the vibration energy of the sound source. For an ideal gas: ( ) 𝛾P dP 2 c0 = = 0 (2.65) 𝜌0 d𝜌 s,0 The velocity of acoustic in an ideal gas can be shown as follows under the temperature t (∘ C). √ c (0∘ C) 𝛾P0 t (2.66) c0 (t∘ C) = (273 + t) ≈ C0 (0∘ C) + 0 𝜌0 273 × 2 c0 (t∘ C) ≈ 331.6 + 0.6t(m∕s)

(2.67)

2.9 Theory of Magnetoacoustic and Photoacoustic Coupling

The kinetic energy: 1 (𝜌 V )v2 2 0 0 The potential energy:

(2.68)

ΔEk =

p

ΔEp = −

∫0

p dV =

V0 2 p 2𝜌0 c0 2

(2.69)

The total energy of sound: ΔE = ΔEk + ΔEp = = V0

( ) V0 1 𝜌0 v2 + 2 2 p2 2 𝜌0 c0

pa 2 cos2 (𝜔t − kx) 𝜌0 c0 2

(2.70)

The energy density (the acoustic energy contained in a unit volume medium in a sound field): ( ) 1 1 ΔE (2.71) = 𝜌0 v2 + 2 2 p2 𝜀= V0 2 𝜌0 c0 Mean energy density: 𝜀=

pa 2 pe 2 ΔE = = V0 2𝜌0 c0 2 𝜌0 c0 2

(2.72)

where Pe is the effective pressure. Mean acoustic power W = 𝜀c0 S

(2.73)

Acoustic intensity (the mean acoustic power per unit area perpendicular to the direction of acoustic propagation): T

I=

W Re(p)Re(v) dt = pe ve = 𝜀c0 = ∫0 S

(2.74)

2.9 Theory of Magnetoacoustic and Photoacoustic Coupling Magnetoacoustic coupling. The phenomenon of energy exchange or mutual excitation between spin waves (magnetons) and sound waves (phonons) in magnetic materials is called the magnetoacoustic effect. It results from magnetoelastic coupling, the coupling between magnetization and elastic strain, which also results in magnetostriction. Strong magnetic objects subjected to an alternating magnetic field will produce corresponding mechanical vibration. This principle has been used in ultrasonic transducers and the acoustic vibration on magnetization. In a static magnetic field, the magnetization of a strong magnet varies with acoustic vibration. In remanence state, acoustic vibration will cause remanence reduction. The magnetoelastic effect coupled the spin-wave (magneton) to the lattice vibration (phonon). When the frequency and wavelength of the spin-wave and the acoustic wave in the

79

80

2 Fundamental in Functional Thin Films and Coatings

strong magnet are equal, it is called crossover. They are strongly coupled and can be converted to each other. The result is a magnetoelastic wave of two coupled waves. The acoustic waves that cause this effect are strongly attenuated. When the stable magnetic field and high-frequency sound field act together and meet certain conditions, some weakly magnetic (diamagnetic or paramagnetic) materials will conduct the resonance transition of electrons between the space moving energy level (landau energy level) or the resonance transition of electron spin energy level, which is called magnetoacoustic resonance.

2.9.1 The Mechanism of the Sound Pressure in the Presence of the Pulse Magnetic Field [44] The sample with the electrical conductivity 𝜎(r ′ ) is put in the pulse magnetic field B(r ′ , t). Here there will be two effects in the sample. One is the magnetoacoustic effect with the induced eddy. The sample in an alternating magnetic field and static magnetic field that can generate eddy current J(r ′ , t) due to the alternating magnetic field. Then the eddy current result in Lorentz force in the presence of the alternating magnetic field and static magnetic field. This causes the organization’s acoustic vibration of the ultrasonic signals, which can be collected by the ultrasonic transducer to reconstruct the distribution of the conductivity of biological tissue. The Lorentz force and can be given as: F𝟏 (r′ , t) = J(r′ , t) × B(r′ , t)

(2.75)

Anther effect is a thermoacoustic effect. Its basic principle is that the alternating magnetic field produces eddy current in the sample and this current introduces the heat generation. The sample absorbs joule heat, which generates transient thermal expansion and emits thermoacoustic signals. The acoustic source is a function of heat, defined as the energy absorbed per unit volume per unit time. And this can be given as: Q(r′ , t) = 𝜎E2 (r′ , t)

(2.76)

E2 (r ′ ,

Here, t) is the model of the electric field. Therefore, the pressure that comes from the heat function and Lorentz force follows the wave equation: 𝛁𝟐 P(r, t) −

𝛽 𝜕Q(r ′ , t) 1 𝜕2 P(r, t) = 𝛁 ⋅ F1 (r ′ , t) − 2 𝜕t2 Cp 𝜕t cs

(2.77)

where P(r, t) is the pressure of the location r and the time t. cs is the speed of sound in a medium, Cp is specific heat capacity, and 𝛽 is the coefficient of cubical expansion. It is can be concluded that the intensity of the ultrasonic signal depends on the value of the pulsed magnetic field, the value of the induced electric field strength, and the distribution of the conductivity.

2.9.2 The Mechanism of the Sound Pressure in the Presence of the Pulsed Magnetic Field and Static Magnetic Field [44] The sample with the electrical conductivity 𝜎(r ′ ) is put in the pulse magnetic field and static magnetic field B0 . The expression of the heat function is not affected, and

2.9 Theory of Magnetoacoustic and Photoacoustic Coupling

the expression of Lorentz is modified: F𝟐 (r′ , t) = J(r′ , t) × (B(r′ , t) + B𝟎 )

(2.78)

The pressure is given as: 𝛁𝟐 P(r, t) −

𝛽 𝜕Q(r′ , t) 1 𝜕2 P(r, t) = 𝛁 ⋅ F𝟐 (r′ , t) − 2 𝜕t2 Cp 𝜕t cs

(2.79)

It can be concluded that the intensity of the ultrasonic signal depends on the value of the pulsed magnetic field and the static magnetic field, the value of the induced electric field strength, and the distribution of the conductivity. Photoacoustic coupling. The interaction of acoustic and optical signals are under intensive studies these years. One of the typical examples is the coupling of optical wave and acoustic wave in the Bragg gratings. Bragg gratings have a periodic variation of the refractive index, a very large group velocity dispersion, and nonlinear effects. Optical waves propagating in the Bragg gratings may form optical solitons. Moreover, the light may drive sound through compressing the medium due to variations of the intensity of light; conversely, acoustic waves react into the optical wave through the dependence of the refractive index on the material density. Considering the electrostriction, optical gap solitons may couple to acoustic waves and generate the optoacoustic solitons. Photoacoustic coupling equations in Bragg gratings can be given as [45]: ik0 ′ 𝜇t + i𝜇z + 𝜅𝜈 +

2𝜋( 𝜔0∕c)2 (𝜒s |𝜇|2 + 𝜒𝜒 |𝜈|2 )𝜇 + 𝜒es 𝜔𝜇 = 0 k0 A

(2.80)

ik0 ′ 𝜈t − i𝜈z + 𝜅 ∗ 𝜇 +

2𝜋( 𝜔0∕c)2 (𝜒s |𝜇|2 + 𝜒𝜒 |𝜈|2 )𝜈 + 𝜒es 𝜔𝜇 = 0 k0 A

(2.81)

𝜔𝜇 − 𝛤 𝜔tzz − 𝛽s2 𝜔zz + 𝜆(|𝜇|2 + |𝜈|2 )zz = 0

(2.82)

Here, 𝜇 = 𝜇(z, t) and 𝜈 = 𝜈(z, t) are the envelope functions of a light wave in opposite directions (i.e. the amplitude envelope function of the light wave propagating forward and backward in the grating). 𝜔 = 𝜔(z, t) is the density function of the medium in the acoustic region, 𝜇 t , 𝜈 t , and 𝜔zz are the derivatives with respect to the time and dk (𝜔 = 𝜔0 ) is the group velocity of a light wave, 𝜅 is the Bragg space variables, k0 ′ = d𝜔 reflectance, and 𝜒 𝜒 and 𝜒 s are cross-phase modulation coefficient and self-phase modulation coefficient, respectively. A is the effective cross-section of a waveguide, ( ) 𝜔 dn 𝜆 is the electrostrictive coefficient associated with energy density, 𝜒es = c0 d𝜔 is the electrostrictive coefficient associated with the variation of the wave vector, 𝛽 s is the speed of sound, and Γ is the acoustic viscosity coefficient that can be neglected in the grating nanostructure. It has been reported that the photoacoustic coupling equations in Bragg gratings can be reduced to the standard nonlinear Schrodinger equation by the multi-scale method, and the approximate photoacoustic soliton analytical solutions in Bragg gratings, such as single-photoacoustic soliton solutions and double-photoacoustic soliton solutions [45].

81

82

2 Fundamental in Functional Thin Films and Coatings

2.10 Theory of Acoustooptic Effect When the ultrasonic wave passes through the medium, it will cause local compression and elongation strains in the medium. The strain changes periodically with time and space, so that the medium appears the phenomenon of dense phase, just like a phase grating. Diffraction occurs when light passes through this medium disturbed by ultrasonic waves. This phenomenon is called the acoustooptic effect. Acoustooptic effect is the light scattering or diffraction when it passes through a medium disturbed by sound waves. Due to the elastic-optical effect, when the ultrasonic longitudinal wave propagates in the medium in the form of the traveling wave, the refractive index of the medium will change in sine or cosine law and will propagate with the ultrasonic wave. When the laser passes through the medium, light diffraction, namely acoustooptic diffraction, will occur. The change of refractive index caused by elastic deformation can be written as: 1 𝛥n = − n3 ps (2.83) 2 where n is the refractive index of the medium, s is the degree of deformation of the medium, and p is the acoustooptic (or elastomeric) coefficient. The characteristics of acoustooptic diffraction are related to the length L of acoustooptic interaction. The characteristic length of acoustooptic diffraction is defined as: L0 =

nv2 𝜆0 f 2

(2.84)

where 𝜆0 is the light wavelength in vacuum, v is the speed of sound, and f is the frequency. It can be seen from the above equation that L0 is very small at high frequency. When a beam of monochromatic collimation light is incident vertically on the ultrasonic grating (the direction of the light propagation is within the gate plane of the grating), the outgoing light is diffraction light, as shown in Figure 2.9. In the figure, m is the number of diffraction order, and 𝜃 is the diffraction angle of the diffraction light of the mth order. It can be proved that, like the optical grating, the condition for the formation of diffraction at all levels is: m𝜆 sin 𝜃m = ± (m = 0, ±1, ±2, …) (2.85) 𝜆s where 𝜆 is the wavelength of incident light and 𝜆s is the wavelength of ultrasonic wave. According to the frequency of the ultrasonic wave and the length of the acoustooptic interaction in the medium, there are two common extreme diffractions, Raman–Nath diffraction, and Bragg diffraction. The parameter for evaluating these two types of diffraction can be given as: 𝜆 Q = 2𝜋L 2 (2.86) 𝜆s where L is the length of the acoustooptic interaction, 𝜆s is the ultrasonic wavelength, and 𝜆 is the wavelength of light passing through the acoustooptic medium. Raman–Nath diffraction occurs when Q is less than 1 (accurately, Q ≤ 0.3). Bragg diffraction occurs when the ultrasonic frequency is high, the acoustooptic region

2.11 Magnetothermal Thin Films: Phonon Thermal Theory

Figure 2.9 Diffraction effect of ultrasonic grating on the light beam.

Direction of ultrasonic propagation Acoustic wall +m +2

θ1

Laser Acostooptic medium

+1 0 –1 –2

Electroacous converter

–m d

is long, and the ray and ultrasonic wave surface have a certain angle of oblique incidence. Bragg diffraction occurs when Q is more than 1 (accurately, Q ≥ 4). In the region of 0.3 < Q < 4, the diffraction phenomenon is relatively complex, and the conventional acoustooptic devices do not work in this range.

2.11 Magnetothermal Thin Films: Phonon Thermal Theory The magnetothermal effect refers to the phenomenon that paramagnetic or soft ferromagnetic materials will release heat under the action of the external magnetic field, and absorb heat during demagnetization. In some paramagnets, magnetothermal effects are used to produce extremely low temperatures. The magnetothermal effect is the inherent nature of all magnetic materials, which is determined by the microstructure of such materials [46]. According to Boltzmann statistics, the entropy of a system can be expressed as a function of Boltzmann’s constant k and the number of states of the system W: S = k ln W

(2.87)

If the number of particles in the system is N and the total quantum number is j, when the system is not affected by the external magnetic field, the energy of particles is randomly distributed in 2j + 1 degenerate state, and its angular momentum and magnetic moment orientation are 2j + 1, so the number of states of the system is W = (2j + 1)N . When the external magnetic field is introduced, the probability of the distribution of the particle’s angular momentum ( ) and magnetic moment orien−g𝜇B mi B tation in each energy level state is P ∝ exp , where g is the Landau factor, KT 𝜇 B is the Bohr magnetic moment, mi is the magnetic quantum number written as 2j + 1, B is the magnetic induction intensity of the external magnetic field, and T is

83

2 Fundamental in Functional Thin Films and Coatings

the temperature of the system. If the external magnetic field intensity is large and the system temperature is relatively low, the magnetic moment orientation will tend to be the same, and the number of states will be nearly 1, leading to the approximately zero entropy. The magnetic entropy change of the system in this process is ΔS ≈ − Nk ln(2j + 1). The system release or absorb heat energy ΔQ ∝ ΔS. So the larger the magnetic field with the lower the temperature, the greater the entropy of the system that leads to the greater magnetic heat effect.

2.12 Theory of Thermoelectric Effect The thermoelectric effect is due to the coupling of heat and electricity flows, which is composed of the Seebeck effect, the Peltier effect, and the Thomson effect. The Seebeck effect was discovered in 1827 by the German physicist Thomas Seebeck. As shown in Figure 2.10, metals A and B were connected into a ring, and the temperature of one joint was at T. When another joint was heated at T + ΔT, it is distinctly found that the compass inside the ring moved due to the generated magnetic field. In this process, the carriers on the hot joint moved fast, causing negative charges to accumulate on the cold side. This creates a potential difference in the metal, creating an electric field. This electric field impedes the carriers’ movement and when the carrier distribution reached an equilibrium state, there is a stable voltage at both ends of the metal, and a stable electric field is formed in the metal. The voltage is proportional to the temperature difference and the proportional coefficient is called the Seebeck coefficient S (T), which is a physical quantity of the material itself and is related to the temperature. The voltage can be given as: T+ΔT

V=

∫T

(2.88)

(SB (T) − SA (T))dT

where SA (T) and SB (T) represent the Seebeck coefficients of metal A and metal B. If SA (T) SB (T) and do not change with the change of temperature, the above equation can be expressed as follows: (2.89)

V = (SB (T) − SA (T))ΔT Figure 2.10 effect.

Cold T

E

N

Metal A W

84

S

Hot T+ΔT

Metal B

e‒

e‒ e‒ e‒

The diagram of the Seebeck

2.12 Theory of Thermoelectric Effect

Figure 2.11 The diagram of the Peltier effect. Metal A

Jq

Jq

Metal B Je Power supply

Another thermoelectric effect is the Peltier effect discovered in 1834, which can be considered as the inverse effect of the Seebeck effect. As shown in Figure 2.11, when an electric current passes through a circuit composed of two different conductors, in addition to producing joule heat, the Peltier heat is absorbed or given off at the two joints. This is due to the fact that the applied voltage on the two joints can drive the carriers that can make the energy transferred from one joint to the other with the heat absorbed and released. It is found that the process is reversible. When the current direction is changed, the endothermic and exothermic joints exchange. Peltier heat Q can be expressed as a function of the Peltier coefficient P. (2.90)

Jq = P × Je

The third one is the Thomson effect. As shown in Figure 2.12, when an electric current passes through a conductor with a certain temperature gradient, in addition to joule heat, another transverse heat flow flows called the Thomson effect into or out of the conductor (i.e. endothermic or exothermic). The temperature gradient (pointing in the direction of temperature rise) is opposite to the direction of the current, that is, the current flows from the hot end to the cold end-exothermic (flowing out laterally).The average velocity of free electrons at the high-temperature end is greater than that at the low-temperature end, so more electrons are diffused from the high-temperature end to the low-temperature end than from the low to the high end, forming a potential difference of temperature difference V(T1, T2), and the direction from the high-temperature end to the low-temperature end. The external current is in the same direction as V(T1, T2), and the electrons are accelerated to obtain energy. Apart from the increase of kinetic energy needed for the electrons to reach the high-temperature end, the remaining energy is transferred to the lattice through the collision between the electrons and the lattice, so that the temperature of the whole metal rises and heat is released. The direction of the temperature gradient is the same as the direction of the current – endothermic (flowing in laterally). The

Je

T QT

Δ

Thomson effect

Δ

Figure 2.12 The diagram of the Thomson effect.

QT = –K × Je × T

K is Thomson coefficient

85

86

2 Fundamental in Functional Thin Films and Coatings

electrons are slowed down by a temperature difference electric field, and when they collide with the lattice, they obtain energy from the metal atoms, reducing lattice energy and the whole metal temperature and absorbing heat from the environment. The amount of heat absorbed or released per unit time dQ/dt is proportional to the current I and temperature gradient dT/dx. dQ dT = SI dt dx

(2.91)

where S represents the Seebeck coefficient. When the circuit composed of two conductors is analyzed, the three thermal effects will be generated at the same time when the temperature of the two contact ends is different.

2.13 Thermal Barrier Insulation Theory for TBC Coating Thermal barrier coatings (TBCs) enable the engines to operate at higher temperatures without raising the base metal temperatures using cooling systems inside the hot section components and thus, enhance the operating efficiency of the engines [47]. TBCs with low thermal conductivity, phase stability, and high resistance to sintering have ever-increasing demands [48]. TBCs are primarily a two-layer system, which is consisting of a porous ceramic topcoat layer and an alumina forming bond coat layer. Topcoat layer (thermal insulation), thermally grown oxide (TGO) layer (bonding of TBC to bond coat and slows subsequent oxidation), topcoat layer (providing thermal insulation), TGO layer (bonding of TBC to bond coat and slows subsequent oxidation), bond coat layer (containing the source of elements to create TGO in an oxidizing environment and provides oxidation protection), and superalloy substrate (mechanical load) are four main components with unique functions that influence TBC life [49]. Each of these components has markedly different physical, thermal and mechanical properties that are strongly affected by processing conditions. During fabrication and most notably during use, these components interact chemically and mechanically. Dynamic relationships between these layers control the durability of TBC. Impedance spectroscopy (IS) is used for structure detection and life prediction of the thermal barrier layers. The impedance characteristics of a material or device depend on dry temperature, thickness, microstructure, or damage, and the damage is detected by measuring the impedance changes caused by such microstructure changes. Applying a sinusoidal amplitude voltage V(t) = V m sin(𝜔t) to the thermal barrier coating will measure the steady-state current i(t) = I m sin(𝜔t + 𝜃). Here, 𝜔 is the angular velocity and 𝜃 is the phase difference between voltage and current. The impedance of the coating is the ratio of the applied voltage to current, which can be given as: Z(𝜔) =

V(t) i(t)

(2.92)

2.14 Permeability Theory: Fick First Diffusion Theory and Fick Second Diffusion Theory

2.14 Permeability Theory: Fick First Diffusion Theory and Fick Second Diffusion Theory [50] Fick’s law proposed by Adolf Fick in 1855 describes the relationship between the mass transfer flux and the concentration gradient in the process of molecular diffusion when the mass transfer occurs without macroscopic mixing. Diffusion occurs in response to a concentration gradient expressed as the change in concentration due to a change in position 𝜕C . The local rule for movement or flux J is given by Fick’s 𝜕x first law of diffusion: J = −D

𝜕C 𝜕x

(2.93)

Here, J is the diffusion flux: the mass of a diffuser per unit area in the direction of diffusion per unit time, D is diffusion coefficient, C is the diffusion concentration, and x is the distance in the diffusion direction. This law reflects a steady-state diffusion, which concentrations do not vary with time. The flux is proportional to the diffusivity and the negative gradient of concentration. The negative sign indicates that J is positive when movement is down the gradient. As shown in Figure 2.13 the negative sign cancels the negative gradient along the direction of positive flux. Consider diffusion at the front and rear surfaces of an incremental planar volume. Fick’s second law of diffusion describes the rate of accumulation (or depletion) of concentration within the volume as proportional to the local curvature of the concentration gradient. The local rule for accumulation is given by Fick’s second law of diffusion: ( ) 𝜕C 𝜕C 𝜕C = D (2.94) 𝜕t 𝜕x 𝜕x Figure 2.13 The flux is driven by the negative gradient in the direction of increasing.

C J

J=

X0

–D д C дX

X

87

88

2 Fundamental in Functional Thin Films and Coatings

Figure 2.14 Incremental planar volume accumulates concentration because the front gradient at x1 drives more flux J1 into the volume than the flux J2 driven out of the volume by the rear gradient at x2 . Source: [51].

Volume

C J1 X1

J2 X2

X

If the diffusion coefficient is not a function of concentration, the Fick’s second law of diffusion can be given as: 𝜕2 C 𝜕C =D 2 𝜕t 𝜕x

(2.95)

is proportional to the diffusivity and the second derivawhich the accumulation 𝜕C 𝜕t tive (or curvature) of the concentration. The accumulation is positive when the curvature is positive, i.e. as shown in Figure 2.14, when the concentration gradient is more negative on the front end of the planar volume and less negative on the rear end that more flux is driven into the volume at the front end than is driven out of the volume at the rear end. For the three-dimensional diffusion problem, the Fick’s second law of diffusion is expressed as: 𝜕C = D ⋅ ∇2 C 𝜕t

(2.96)

At steady state, we have 𝜕C = 0, which means no concentration change, where we 𝜕t can get the Fick’s first law that is considered as a specific (simplified) format of the second law.

2.15 Multi-physical Field Coupling Theory and Simulation Software Introduction In engineering, there are various physical fields including the temperature field, the stress field, the humidity field, and so on. It is found that many of the problems we have to solve are the superimposing of these physical fields that affect each other. Multiple physical fields superimposing on each other is called the multi-field coupling. With the rapid development of computer technology, finite element analysis (FEA) is increasingly used in the field of engineering simulation to solve real engineering problems. Over the years, a growing number of engineers applied mathematicians, and physicists have shown that this method of solving partial differential equations, which describe flows, electromagnetic fields, and structural mechanics, can solve many physical phenomena. The finite element method is used to convert these well-known mathematical equations into approximate digital images. Here, we mainly introduce two typical multi-physical field coupling software, i.e. COMSOL Multiphysics and ANSYS Multiphysics. COMSOL Multiphysics is a multiphysical field analysis software based on FEA, which can be used for mechanical, optical, chemical, and other aspects of simulation, has been widely used in scientific research simulation and calculation. This

2.15 Multi-physical Field Coupling Theory and Simulation Software Introduction

software strives to meet all requirements of user simulation and become the preferred simulation tool for users. It is versatile, flexible, easy-to-use, and more powerful than other FEA software due to the fact that it can be easily extended with additional functional modules. Its distinctive features are as follows: (1) Solving multi-field problems. Users only need to choose different professional differential equations for arbitrary combinations. This can easily realize the direct coupling analysis of multi-physical fields. (2) A fully open architecture. Users can easily define the required professional partial differential equations in a graphical interface. (3) Controlled solution parameters, material properties, and boundary conditions. (4) Professional computing model library. (5) Rich embedded CAD modeling tools. (6) Comprehensive CAD import function. It supports the current mainstream CAD software format file import. (7) Powerful grid generation ability. It supports a variety of grid generation and mobile grid functions. (8) Large scale computing power. This software has the 64-bit processing power and parallel computing on Linux, Unix, and Windows systems. (9) Rich post-processing functions. It can output and analyze a variety of data, curves, and pictures. (10) Professional online help documents. Users can easily master the operation and application of the software through the operation manual. (11) Multi-language interface. As shown in Figure 2.15 this software has a convenient and quick interface for load conditions, boundary conditions, and parameter settings. COMSOL Multiphysics is a great innovation in multi-field coupling simulation. It integrates rich algorithms that are functional, flexible, and practical. Moreover, it can be applied and expanded conveniently through additional professional solution modules including AC/DC module, heat transfer module, CFD module, chemical reaction engineering module, RF module, structural mechanics module, microfluidics module, batteries and fuel cells module, MEMS module, geomechanics module, subsurface flow module, electrodeposition module, plasma module, acoustics module, pipe flow module, corrosion module, nonlinear structural materials module, and an optimization module. ANSYS Multiphysics software is a large general FEA software and integrates structure, fluid, electric field, magnetic field, and sound field analysis. It is developed by ANSYS company that is one of the largest FEA software companies in the world. It can be compatible with most CAD software for data sharing and exchange, such as Pro/Engineer, NASTRAN, Alogor, i-deas, AutoCAD, etc. It is one of the most advanced CAE tools in modern product design. This software mainly consists of three parts: pre-processing module, analysis, and calculation module and post-processing module. The preprocessing module provides a powerful entity modeling and mesh generation tool that allows users to easily construct finite element models; the analysis and calculation module includes structural

89

90

2 Fundamental in Functional Thin Films and Coatings Untitled.mph - COMSOL Multiphysics File

Home

Definitions

Geometry

Materials

Physics

Mesh

Update COMSOL Application Libraries

Results

Developer

Normal Modes of a Biased Resonator–3D

Application Libraries Refresh

Study

Application Gallery Search

Composite Materials Module Corrosion Module Design Module ECAD Import Module Electrochemistry Module Electrodeposition Module Fatigue Module Geomechanics Module Heat Transfer Module MEMS Module Actuators biased_resonator_2d_basic

An electrostatically actuated MEMS resonator is simulated. The device is driven by an AC + DC bias voltage applied across a parallel plate capacitors. In this example, the normal mode shapes and frequencies are computed, as a function of applied bias.

biased_resonator_2d_freq biased_resonator_2d_modes biased_resonator_2d_pull_in_pull_out biased_resonator_2d_pull_in biased_resonator_3d_basic biased_resonator_3d_freq

Run Application

Open

Open PDF Document

Name Used products Physics interfaces

Help

Close

Figure 2.15

biased_resonator_3d_modes COMSOL Multiphysics and MEMS Module Electrostatics Solid Mechanics

Modules window of the COMSOL Multiphysics. Source: COMSOL Multiphysics.

analysis (linear analysis, nonlinear analysis, and highly nonlinear analysis), fluid dynamics analysis, electromagnetic field analysis, sound field analysis, piezoelectric analysis, and coupling analysis of multiple physical fields, which can simulate the interaction of multiple physical media with sensitivity analysis and optimization analysis capabilities; and the post-processing module can display the calculation results by color-contour display, gradient display, vector display, particle flow trace display, three-dimensional section display, transparent and translucent display and other graphics, and can also display or output the calculation results in the form of charts and curves. The software provides more than 100 unit types to simulate various structures and materials in engineering. The software comes in various versions and can run on a variety of computer devices from personal computers to mainframes, such as PCS, SGI, HP, SUN, DEC, IBM, CRAY, etc.

Acknowledgments This chapter is supported by the NSFC–BRICS STI Framework Program (No. 51861145309), the National Natural Science Foundation of China (No. 51971029), the National S&T Major Project (No. 2018ZX10301201), the Postdoctor Research Foundation of Shunde Graduate School of University of Science and Technology Beijing (No. 2020BH005), the Project funded by China Postdoctoral Science Foundation (No. 2020M680336), the “All English teaching demonstration course construction project of University of Science and Technology Beijing” (No. KC2015QYW06, 2016), the “1125” Zhihui Zhengzhou Talent project of Henan province (Fund No. in USTB: 39080070), the “100 talent plan” fund of Fujian province (Fund No. in USTB: 39080067), and the development of a highly sensitive

References

MO biomelecular sensor experimental prototype (Fund No. in USTB: 2019-0649) by Hangzhou Ruidi Biotechnology Co. Ltd.

List of Abbreviation CMR TMR QAHE 2D GSTCs SPR MO MOKE P-MOKE L-MOKE T-MOKE TIs TR TRS IS

colossal magnetoresistance tunneling magnetoresistance quantum anomalous Holzer effect two-dimensional generalized sheet transition conditions surface plasmon resonance magneto-optical magneto-optical Kerr effect polar magneto-optical Kerr effect longitudinal magneto-optical Kerr effect transverse magneto-optical Kerr effect Topological insulators time reversal time-reversal symmetry impedance spectroscopy

References 1 Acharjee, S. and Goswami, U.D. (2019). Spin transport and spin tunnelling magneto-resistance (STMR) of F|NCSC|F spin valve. Journal of Magnetism and Magnetic Materials 495: 165844. 2 Li, Y., Huang, W., Wang, B., and Weng, L. (2019). High-frequency output characteristics of giant magnetostrictive transducer. IEEE Transactions on Magnetics 55 (6): 1–5. 3 Rinkevich, A.B., Pakhomov, Y.A., Kuznetsov, E.A. et al. (2019). Microwave giant magnetoresistance in [CoFe/Cu]n superlattices with record-high magnetoresistance. Technical Physics Letters 45 (3): 225–227. 4 Zhang, W., Li, J., Ding, X. et al. (2017). Tunable magneto-optical Kerr effects of nanoporous thin films. Scientific Reports 7 (1): 2888. 5 Zhang, W., Wang, Q., Zhao, C., and Song, Y. (2019). The optical cavity enhanced magneto-optical Kerr effect signals of AAO/Al-based CoFeB nanostructure arrays. Optics Communications 437: 44–49. 6 Chen, B., Rojo, A.G., Uher, C. et al. (1997). Magnetothermal conductivity of La0.8 Ca0.2 MnO3 . Physical Review B 55 (23): 15471–15474. 7 Evangelisti, M., Luis, F., de Jongh, L.J., and Affronte, M. (2006). Magnetothermal properties of molecule-based materials. Journal of Materials Chemistry 16 (26): 2534–2549.

91

92

2 Fundamental in Functional Thin Films and Coatings

8 Kurita, N., Ronning, F., Tokiwa, Y. et al. (2009). Low-temperature magnetothermal transport investigation of a Ni-based superconductor BaNi2 As2 : evidence for fully gapped superconductivity. Physical Review Letters 102 (14): 147004. 9 Pons, J.A., Miralles, J.A., and Geppert, U. (2009). Magneto-thermal evolution of neutron stars. Astronomy and Astrophysics 496 (1): 207–216. 10 Huang, P., Lin, J., Li, W. et al. (2013). Biodegradable gold nanovesicles with an ultrastrong plasmonic coupling effect for photoacoustic imaging and photothermal therapy. Angewandte Chemie International Edition 52 (52): 13958–13964. 11 Mallidi, S., Larson, T., Tam, J. et al. (2009). Multiwavelength photoacoustic imaging and plasmon resonance coupling of gold nanoparticles for selective detection of cancer. Nano Letters 9 (8): –2825, 2831. 12 Ramirez, A.P. (1997). Colossal magnetoresistance. Journal of Physics: Condensed Matter 9 (39): 8171–8199. 13 Rodriguez-Martinez, L.M. and Attfield, J.P. (1996). Cation disorder and size effects in magnetoresistive manganese oxide perovskites. Physical Review B: Condensed Matter 54 (22): R15622–R15625. 14 Chang, C., Zhang, J., Feng, X. et al. (2013). Experimental observation of the quantum anomalous Hall effect in a magnetic topological insulator. Science 340 (6129): 167–170. 15 Hu, J., Zhu, Z., and Wu, R. (2015). Chern half metals: a new class of topological materials to realize the quantum anomalous Hall effect. Nano Letters 15 (3): 2074–2078. ´ B.K. (2017). Monolayer of the 5d transition 16 Sheng, X.-L. and Nikolic, metal trichloride OsCl3 : a playground for two-dimensional magnetism, room-temperature quantum anomalous Hall effect, and topological phase transitions. Physical Review B 95 (20): 201402.1–201402.5. 17 Qiao, Z., Yang, S.A., Feng, W. et al. (2010). Quantum anomalous Hall effect in graphene from Rashba and exchange effects. Physical Review B 82 (16): 161414. 18 Onoda, M. and Nagaosa, N. (2003). Quantized anomalous Hall effect in two-dimensional ferromagnets: quantum Hall effect in metals. Physical Review Letters 90 (20): 206601. 19 Weng, H., Yu, R., Hu, X. et al. (2015). Quantum anomalous Hall effect and related topological electronic states. Advances in Physics 64 (3): 227–282. 20 Liu, C.-X., Qi, X.-L., Dai, X. et al. (2008). Quantum anomalous Hall effect in Hg1−y Mny Te quantum wells. Physical Review Letters 101 (14): 146802. 21 Rajan, P.I., Mahalakshmi, S., and Chandra, S. (2017). Occurrence of spintronics behaviour (half-metallicity, spin gapless semiconductor and bipolar magnetic semiconductor) depending on the location of oxygen vacancies in BiFe0.83 Ni0.17 O3 . Royal Society Open Science 4 (6): 170273. 22 Schliemann, J. (2017). Colloquium: persistent spin textures in semiconductor nanostructures. Reviews of Modern Physics 89: 011001. 23 Suzuki, R., Wakabayashi, Y.K., Okamoto, K. et al. (2018). Quantum size effect in an Fe quantum well detected by resonant tunneling carriers injected from a p-type Ge semiconductor electrode. Applied Physics Letters 112 (15): 152402.

References

24 Yan, W.-W., Li, X.-F., Zhang, X.-H. et al. (2019). Family-dependent magnetism in atomic boron adsorbed armchair graphene nanoribbons. Journal of Materials Chemistry C 7: 6241–6245. 25 Kawakami, E., Jullien, T., Scarlino, P. et al. (2016). Gate fidelity and coherence of an electron spin in an Si/SiGe quantum dot with micromagnet. Proceedings of the National Academy of Sciences of the United States of America 113 (42): 11738–11743. 26 Pershin, V.Y. (2005). Long-lived spin coherence states in semiconductor heterostructures. Physical Review B 71 (15): 155317. 27 Menzel, C., Rockstuhl, C., and Lederer, F. (2010). Advanced Jones calculus for the classification of periodic metamaterials. Physical Review A 82 (5): 053811. 28 Idemen, M.M. (2011). Discontinuities in the Electromagnetic Field. New Jersey: Wiley-IEEE Press. 29 Niemi, T., Karilainen, A.O., and Tretyakov, S.A. (2013). Synthesis of polarization transformers. IEEE Transactions on Antennas and Propagation 61 (6): 3102–3111. 30 Achouri, K., Salem, M.A., and Caloz, C. (2015). General metasurface synthesis based on susceptibility tensors. IEEE Transactions on Antennas and Propagation 63 (7): 2977–2991. 31 Tretyakov, S. (2003). Analytical Modeling in Applied Electromagnetics. Boston: Artech House. 32 Asadchy, V.S., Albooyeh, M., Tcvetkova, S.N. et al. (2016). Perfect control of reflection and refraction using spatially dispersive metasurfaces. Physical Review B 94 (7): 075142. 33 Sakai, O. (2016). Negative-permittivity plasma generation in negative-permeability metamaterial space. Plasma Sources Science and Technology 25 (5): 055019. 34 Machac, J. (2017). Amorphous metamaterial with negative permeability. IEEE Antennas and Wireless Propagation Letters 16: 2138–2141. 35 Sun, K., Fan, R.H., Zhang, Z.D. et al. The tunable negative permittivity and negative permeability of percolative Fe/Al2 O3 composites in radio frequency range. Applied Physics Letters 106 (17): 172902. 36 Zhang, X. and Liu, Z. (2008). Superlenses to overcome the diffraction limit. Nature Materials 7 (6): 435–441. 37 Xia, W.B., Gao, J.L., Zhang, S.Y. et al. (2014). Optical and magneto-optical properties of periodic Co double layer film. Chinese Physics B 23 (10): –103303. 38 Geim, A.K. and Grigorieva, I.V. (2013). Van der Waals heterostructures. Nature 499 (7459): 419–425. 39 Jariwala, D., Sangwan, V.K., Lauhon, L.J. et al. (2014). Emerging device applications for semiconducting two-dimensional transition metal dichalcogenides. ACS Nano 8 (2): 1102–1120. 40 Sharma, B.L. and Purohit, R.K. (1974). Theory of heterojunctions. In: Semiconductor Heterojunctions, vol. 5 (eds. B.L. Sharma and R.K. Purohit), 1–23. Pergamon.

93

94

2 Fundamental in Functional Thin Films and Coatings

41 Kumari, A. and Singh, R.R. (2017). Encapsulation of highly confined CdSe quantum dots for defect free luminescence and improved stability. Physica E: Low-dimensional Systems and Nanostructures 89: 77–85. 42 Peña, M.A. and Fierro, J.L. (2001). Chemical structures and performance of perovskite oxides. Chemical Reviews 101 (7): 1981. 43 Luxová, J., Ulcová, P., and Trojan, M. (2008). Study of perovskite compounds. Journal of Thermal Analysis and Calorimetry 93 (3): 823–827. 44 Yang, Y., Liu, G., and Xia, Z. (2017). Simulation research on magnetoacoustic effect and thermoacoustic effect of pulsed magnetic excitation. Journal of biomedical engineering 34 (1): 21–26. 45 Hua-Xing, L. and Ji, L. (2011). The perturbed optoacoustic solitons in Bragg grating. Acta Physica Sinica 60 (12): 124201. 46 Wang, G.F., Song, L., Ou, Z.Q. et al. (2007). Calculation of the magnetization and magnetocaloric effect in the MnFeP0.45 As0.55 compound. Acta Metallurgica Sinica 20 (4): 265–269. 47 Tang, F., Ajdelsztajn, L., and Kim, G.E. (2006). Effects of variations in coating materials and process conditions on the thermal cycle properties of NiCrAlY/YSZ thermal barrier coatings. Materials Science and Engineering A 425 (1): 94–106. 48 Matsumoto, M., Takayama, H., Yokoe, D. et al. (2006). Thermal cycle behavior of plasma sprayed La2 O3 , Y2 O3 stabilized ZrO2 coatings. Scripta Materialia 54 (12): 2035–2039. 49 Darolia, R. Thermal barrier coatings technology: critical review, progress update, remaining challenges and prospects. International Materials Reviews 58 (6): 315–348. 50 Fantini, S., Walker, S.A., Franceschini, M.A. et al. (1998). Optical characterization of breast tumors by frequency-domain optical mammography. Advances in Optical Imaging and Photon Migration. Orlando, FL: Optical Society of America. 51 Diffusion theory. https://omlc.org/classroom/ece532/class5/ficks1.html.

95

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing: From GMR, CMR, TMR to Quantum Anomalous Holzer Effect Weiwei Zhang 1 and Yujun Song 1,2 1 University of Science and Technology Beijing, Center for Modern Physics Technology, Applied Physics Department, School of Mathematics and Physics, 30 Xueyuan Road, Beijing 100083, China 2 Zhejiang Key Laboratory for Pulsed Power Technology Translational Medicine, Hangzhou Ruidi Biotechnology Company, Hangzhou 310000, China

3.1 Introduction The electron spin and the motion of the electron charge of solids were investigated independently in the last century [1–4]. However, during the past three decades, both the emergence of various fabrication techniques (e.g. magnetron sputtering [5–8], photo etching, microfluidics [9–12], and template transfer imprinting method [6, 13]) that allow the control of materials at the micro- and nanoscale and the discovery of the anisotropic magnetoresistance (AMR), giant magnetoresistance (GMR) that consists of two ferromagnetic layers separated by a nonmagnetic highly conductive metal, colossal magnetoresistance (CMR) whose materials are mostly manganese-based perovskite oxides, tunneling magnetoresistance (TMR) that consists of ferromagnets/insulator/ferromagnets, and quantum anomalous Holzer effect (QAHE) lead to the development and combination of electron spin and charge properties. Figure 3.1 shows their typical structures. This combination is termed as spintronics (magnetoelectronics or spin electronics). Different from the conventional microelectronics, spintronics regulates the spin state at mesoscopic sales through magnetic fields, electric fields, and even electromagnetic wave to modulate electron transport characteristics. Besides its fundamental scientific importance, regulation of the spintronics at the micro- and nanoscale has been a topic of much interest due to its potential exploitation toward electron transport and signal sensing applications. In this chapter, we aim to focus on the currently used magnetic thin films for electron transport control and signal sensing applications based on GMR, CMR, TMR, and QAHE by covering the fabrication technologies, the structure and property characterization, and the spintronic applications.

Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications, First Edition. Edited by Yujun Song. © 2021 WILEY-VCH GmbH. Published 2021 by WILEY-VCH GmbH.

96

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

(b) FM NM FM

Spin current

(c) (a) M θ I

Magnetoresistance

Perovskites

AMR Upper contact Ferromagnet

(d)

Insulator Ferromagnet

Figure 3.1 Types of magnetoresistance. (a) AMR is a consequence of spin-polarized scattering in a ferromagnetic metal; (b) GMR exists in heterogeneous magnetic material that consists of ferromagnets separated by conducting spacers; (c) CMR based on mostly perovskites results from interactions predominantly between adjacent atoms; and (d) TMR in magnetic tunnel junctions results from spin filtering as spin-polarized electrons tunnel across an insulating barrier from one ferromagnetic to another.

3.2 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing Based on Magnetoresistance (GMR) Effect 3.2.1

Introduction of GMR

GMR was first discovered in 1988 [15] and based on the dependence of electron scattering on the spin orientation, which is a quantum mechanical magnetoresistance effect. This nanotechnology makes it possible to miniaturize hard disks so radically in recent years. For instance, sensitive read-out heads based on GMR can read data from the compact hard disks used in laptops and some music players. The 2007 Nobel Prize in Physics was awarded to Albert Fert and Peter Grünberg for the discovery of GMR [16], which was regarded as a global recognition to the rapid development of GMR, in terms of both physics and engineering [17]. This effect is a significant change in the electrical resistance depending on whether the magnetization of adjacent ferromagnetic layers are in a parallel or an antiparallel alignment in the presence of an external magnetic field. It is found that the overall resistance is

3.2 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing

relatively low for parallel alignment and relatively high for antiparallel alignment. Usually, the value of the GMR can be defined as MR =

𝜌(H) − 𝜌(0) Δ𝜌 = 𝜌(0) 𝜌(0)

(3.1)

Here, 𝜌(H) represents the electrical resistivity in the presence of the applied magnetic field H and 𝜌(0) is the electrical resistivity without the applied magnetic field. The most pioneering work of GMR was the study of the interlayer coupling mechanism of Fe/Gr/Fe multilayered film system [18]. Interestingly, there existed a critical Gr thickness (about 1 nm) for the magnetization of adjacent ferromagnetic layers, which could be changed from ferromagnetic coupling to antiferromagnetic (AF) coupling. This phenomenon was further proved by the spin-polarized low-energy electron diffraction measurements experiment [19]. Subsequently, the GMR effect was first observed in multilayers composed of alternating ferromagnetic and non-magnetic conductive layers system. In this system, huge magnetoresistance of the (Fe(001)/Gr(001))n body-centered cubic (bcc) superlattices on GaAs(001) substrates prepared via a molecular beam epitaxy (MBE) method was achieved in the presence of the applied magnetic field. This was attributed to the spin-dependent transmission of the conduction electrons between Fe layers through Cr layers [15] in AF coupling resulting from the indirect exchange interactions through the Cr layers, which was subsequently demonstrated in the neutron diffraction experiment [20]. Different from the conventional magnetoresistance that results from the increase of the electron-scattering cross section attributed to the functional magnetic field imposing on the electric charge, the magnetoresistance of the GMR is much larger. For instance, in the Fe/Gr superlattice multilayered system, the MR can reach 50% at 4.2 K [15]. Subsequently, it was found that both the transport and magnetic properties of these sputtered polycrystalline Fe/Cr superlattice structures were similar to the reported single-crystal structures [21]. Shortly afterward, Co/Cu multilayers grown on Fe buffer layer displayed the saturation magnetoresistance at room temperature of more than 65% [22]. In a Co90 Fe10 /Cu system, the magnetoresistance reached 63% and the magnetoresistance up to 110% for Fe/Pt/Cu/[Co95 Fe5 /Cu]120/Pt grown by epitaxy on a MgO(110) single crystal substrate was also reported. These related multilayers become the fundamental materials of GMR sensors and storage devices today [14].

3.2.2

Fabrication of GMR Multilayered Thin Films

Generally, various systems such as transition metal magnetic multilayer films (such as spin valve structure), magnetic nanoparticles, oxide films, and magnetic tunnel junction have been fabricated to explore the GMR effect on the electron transport control and signal sensing applications. Among the fabrication methods, the MBE method, magnetron sputtering method, and the electron beam deposition method are three main methods to fabricate the samples. Table 3.1 summarized the typical multilayered magnetic thin film fabricated by these three methods. Besides these three main methods, some combinatorial thin-film magnetic structures can be also obtained by the thermionic vacuum arc method. In this process, the

97

98

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

Table 3.1 Typical multilayered magnetic thin film fabricated by MBE, magnetron sputtering and electron-beam evaporation methods. Fabrication technology

Multilayered films

MBE

(Fe/Cr)n /GaAs [15], Co/Ag [23] Co/Cu [24–40], Mn/Fe/Mn/Ru [41], Mn/Ru [41], Fe/Ag/CoxFe1−x [42]

Magnetron sputtering

Fe/Gr [43], Co/Cu/Fe (buffer) [44], Ni81 Fe19 /Cu [45], Fe16 Ni66 Co18 /Cu [46], Co–Cu/Si [47], Co/Ag [48], FexAg100−x [49], Co–Cu [50], Fe/Cr [51], Fe/Ag and Co/Ag [52], Ni66 Fe16 Co18 /Cu/Fe (buffer) and Ni66 Fe16 Co18 /Cu/ NiFeCo [53], Co/Cu/glass [54], NiFeCo/Cu [55], (CoFe)x Ag1−x [56], CO/CU [57], Ni–Co/Cu and Co–Fe/Cu [58], Ta(tTa)/(Ni80 Fe20 )60 Cr40 (50 Å)/[Co90 Fe10 (15 Å)/ Cu(tCuÅ)]n /Ta(50 Å) [59], NiFeCr [59], Ta/NiFeCr [59], NiFeCr [59]

Electron-beam evaporation

Cu/Co/Cu/glass [60], Co/Cu/Co sandwiches with Si buffer [61], Co2 MnSi (CMS)/n–Si [62], CoFe/p–Si [63], Ta–NiFe–Cu–Co–FeMn–Ta [64]

Quartz microbalance Sample holder

Co

MgO

Cu Fe

FeCo + ‒ H.V.

Figure 3.2 Schematic view of both GMR and combinatorial GMR + TMR structures. Source: Jepu et al. [65]. ©2014, Elsevier.

accelerated electrons focus on the anode start the evaporation process, which leads to the ignition of a bright plasma discharge in pure anode vapors. All processes take place in high vacuum conditions. Take the fabrication of the Fe–Co magnetic structure as an example [65]. As shown in Figure 3.2, due to the substrate position with respect to each element to be deposited, different elemental concentrations were obtained in the same deposition batch. The entire coating process was performed in a high vacuum chamber, with a base pressure in the deposition time of 5 × 10−6 Torr. This high vacuum pressure assures the obtaining of high purity structures without any other unwanted inclusions.

3.2 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing

3.2.2.1 MBE Method for the Fabrication of the GMR Devices

MBE method, developed on the basis of vacuum deposition technology that usually reaches more than 6 × 10−9 Pa [66], was mainly used in the early research of GMR films. Usually, MBE system consisting of ultrahigh vacuum (UHV) chambers, substrate preparation part, thin-film growth and analysis parts, and sample exchange load-lock chamber. It is a method of depositing new monocrystalline films with complete crystal structure one by one along the grain axis of substrate materials under appropriate conditions. This method can be used to grow extremely thin monocrystalline films and has the ability to accurately control the film thickness, composition, and dopants. However, it needs high requirements for the substrate and the epitaxial material whose atomic lattice mismatch is not more than 7%. Since the early work on MBE-grown multilayered Fe/Gr/Fe for the GMR investigation, MBE is now a well-established epitaxial process of major importance in the development of electron transport control and signal sensing applications. 3.2.2.2 Magnetron Sputtering Method for the Fabrication of GMR Devices

The magnetron sputtering system uses magnetrons to realize strong electric and magnetic fields to confine charged plasma particles, which will make the charged plasma particles close to the surface of the sputter target. The plasma can be sustained at lower pressure and the sputtered atoms are neutrally charged and unaffected by the magnetic trap. In the magnetron-sputtering setup (Figure 3.3), specially designed sputter guns with unusually strong permanent magnets are often used in compensation. During the sputtering process, electrons move along the magnetic field lines in the presence of an inert gas (such as argon) and the extra argon ions created as a consequence of these collisions lead to a higher deposition rate. It is found that compared with the evaporation of materials in a resistance evaporator or Knudsen cell, for the magnetron sputtering method, even the materials with very high melting points can be easily sputtered and the sputter deposited films have a better adhesion on the substrate with a composition close to that of the source material due to the different elements spreading differently because of their different mass (light

Water

–DC and RF Shutter

Ar gas

Magnet Target Plasma

Ar Ar+ e– Ar

Ar

M M

Ar

e– M

Substrate Platform

(a)

(b)

Figure 3.3 Schematic of the magnetron sputtering system. Source: Zhang and Song (Image permitted from Tianjin Tianzhao Yuhua Science and Technology Company, Tianjin, China).

99

100

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

elements are deflected more easily by the gas). For the insulating targets, charged process can be avoided due to the introduction of a radiofrequency (RF) sputtering where the sign of the anode–cathode bias is varied at a high rate and RF sputtering works well to produce highly insulating oxide films. Up to date, a broad of multilayered systems based on GMR effect have been fabricated via the magnetron sputtering method (Table 3.1).

3.2.3

GMR Applications for Sensors

Due to the microlevel dimension, excellent sensitivity, and complementary metal– oxide-semiconductor (CMOS) compatibility, GMR sensors have attracted more consideration and interest for conventional applications such as read heads in hard disk drives [14], magnetic memory that brings an improvement to the memory devices, and at the same time, a significant increase in the reading speed of this type of storage device, detection of weak magnetic fields, angular velocity and acceleration, inertia navigation and positioning, and recently emerging applications like eddy current sensing probe [67], ultralow pressure measurement [68], biological magnetic sensors (such as gene expression analysis and influenza virus detection) [69–73], some novel technologies in water pollution detection [74], wireless charging for electric vehicles, and even current monitoring for smart power grids [75]. For the biological sensors, GMR devices have been shown to detect proteins, like cancer biomarkers [69], autoantibodies (Figure 3.4 shows GMR biosensor microarray chips. Each chip has 72 effective sensors as the platform for autoantigen microarrays.) [70], and common allergens (Figure 3.5 shows the process of the multiplexed allergen sensing on GMR sensor array) [71], with high sensitivity and specificity. They have also been utilized for DNA detection [72] and have been well characterized for simultaneous mutation and methylation analysis [73]. (i) Individual GMR sensors were spotted with different capture antibodies, specific to their respective allergens. (ii) Allergens were added and bind to respective capture antibodies. (iii) Biotinylated detection antibodies were added and bind to respective

50μm

500μm

(a)

(b)

Figure 3.4 GMR microarrays for the detection of auto antigen. Optical images of a GMR biosensor chip (a) and a cartridge with a reaction well and their sensors (b). Source: Lee et al. [70].

3.2 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing

(2)

(1)

GMR sensor

GMR sensor

GMR sensor

(4)

GMR sensor

GMR sensor

(3)

GMR sensor

GMR sensor

GMR sensor

GMR sensor

GMR sensor

GMR sensor

GMR sensor

GMR sensor

(6)

(5)

GMR sensor

GMR sensor

GMR sensor

GMR sensor

GMR sensor

Figure 3.5 Illustration of the GMR-based multiplexed allergen detection assay. Source: Ng et al. [71].

allergens. (iv) The sandwich immunoassay structures were formed on individual sensors. (v) Streptavidin microbeads were added to the reaction well. (vi) As the magnetic particles bind to the biotinylated detection antibodies, local magnetic fields were generated, and changes in the resistance of the GMR sensor produce a signal that could be correlated to each allergen concentration [71]. A real-time detection system for water pollution status and pollutant concentration based on GMR sensor considering water pollution detection has been reported [74]. It can realize multipoint automatic sampling, detection, analysis, data uploading analysis, real-time monitoring, alarm, and other functions and overcome the defects of existing technologies. It has been verified that the real-time detection system for water pollution status and pollutant concentration applied with GMR detector exhibit excellent characteristics. A simple, cost-effective, magnetic pressure sensor consisting of a polymer diaphragm, a permanent magnet, and a GMR-based magnetic field gradient sensor is shown in Figure 3.6. The fabricated prototype was calibrated for ultralow differential pressure range, and it showed sensitivity up to 16.67 μV/(V Pa) and nonlinearity of 1.5% full scale (FS) in the range of 0–300 Pa. The concept was not limited to the low pressure but wide ranges of pressure could be realized depending on the type of the diaphragm [68]. GMR could also be as a sensor to make a magnetic nanoparticle detection scheme without the presence of an external magnetic field. In this device, there is a magnetic sensor in a patterned groove structure and external magnetic field is not needed to magnetize the magnetic nanoparticle. An example is given based on a GMR sensing

101

102

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

P

Diaphragm Magnet GMR sensor PCB Movable stage

Diaphragm Magnet dz

z R1,R3 y

dx d

Aluminium housing

Aluminium spacer R2,R4

GMR sensor PCB Base

x

Stage lock

(a)

Electrical connection

(b)

Figure 3.6 (a) Schematic of magnetic pressure sensing mechanism and (b) Schematic of pressure sensor prototype. Source: Borole et al. [68]. Generation

G GMR Current sensor

Transformation

Transmission

G

GMR sensor networks

G

Distrubution

Consumption

G

Data transmission G Dispatch center

Figure 3.7 Schematic diagram of GMR sensor applications in the smart grid. Source: Ouyang et al. [75].

device with a spin valve structure [76]. In this structure, the detection of magnetic nanoparticles located inside the groove and near the free layer is demonstrated without magnetic field, which will be useful for magnetic biosensing system miniaturization and power consumption control. Since the GMR was first commercialized in 1997 [14], GMR structures have been widely used in the conventionally modern hard drives, magnetoresistive random access memory (MRAM) as cells that store one bit of information and sensor applications in the measurements of current [17, 77], angular displacement, accelerated velocity as well as biological signatures of diseases and smart grid as shown in Figure 3.7.

3.3 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing Based on Colossal Magnetoresistance (CMR) Effect 3.3.1

Introduction of CMR

CMR effect belongs to the magnetoresistance effect associated with a ferromagnetic to paramagnetic phase transition under the application of a magnetic field.

3.3 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing

Materials, mostly manganese-based perovskite oxides, featuring CMR may demonstrate many orders of magnitude exceeding GMR of multilayered and granulated systems [78, 79]. CMR phenomena were first discovered in the hole-doped (e.g. Ca or Sr substitution for La) compound LaMnO3 family, in which large magnetoresistance was observed near a high temperature (T) insulating paramagnetic phase to a low-T conducting ferromagnetic phase [80]. Lately, it is found that CMR materials constitute the broad class of doped cubic perovskite manganites termed as R1−x Ax MnO3 . Here R represents a rare earth and A is alkaline earth. They can exhibit huge resistance in the presence of a magnetic field. Recently, it is also reported that CMR can also be realized in a compressed single-valent LaMnO3 manganite compound by generating an inhomogeneous, nonconductive one with a unique structural distortion and a metallic one without distortion using pressure [80]. Besides that, large CMR has also been found in Tl2 Mn2 O7 and some Cr chalcogenide spinels compounds that are different greatly from the manganite perovskites. The phenomenon of CMR is currently of considerable interest because of its value in fundamental physics and potential applications.

3.3.2

Fabrication of Multilayered Thin Films Based on CMR Effect

Much attention has been paid to the fabrication of artificially CMR structures such as p–n junctions, tunnel junctions, and superlattices, using the magnetron sputtering method, pulsed laser deposition method (PLD), MBE, and other methods such as solid-state method. Table 3.2 summarized the typical multilayered magnetic thin film fabricated by these three methods. Magnetron sputtering method is a physical vapor deposition method and can be controlled by many complex parameters including deposition pressure, time, and temperature of the substrates. This method allows for a large degree of control over the growth and microstructure of the film. In the magnetron sputtering method, the off-axis is usually introduced to fabricate thin films for the investigation of the CMR properties. Table 3.2 Typical multilayered magnetic thin film fabricated by magnetron sputtering, PLD, and MBE methods. Fabrication technology

Multilayered films

Magnetron sputtering

La0.7 Ca0.3 MnO3 [81, 82], La0.67 Sr0.33 MnO3 /La0.75 MnO3 / La0.67 Sr0.33 MnO3 [83], Nd0.52 Sr0.48 MnO3 [84], Sm0.53 Sr0.47 MnO3 [85], BiFeO3 [86], La1−x Srx MnO3 [82, 87, 88], La0.67 Ca0.33 MnO3 [89], La0.67 Ca0.33 MnO3 /YBa2 Cu3 O7−δ bilayers [89]

PLD method

YBa2 Cu3 O7−y /LaMnO3 /YBa2 Cu3 O7−y [90], La1−x Pbx MnO3 [91], LaCaMnO/CeO2 /Si(100) [92], La0.9 Sb0.1 MnO3 [93], Pr0.7 Sr0.3 MnO3 /SrTiO3 (001) [94], Pr0.7 Sr0.3 MnO3 /La0.5 Ca0.5 MnO3 [94, 95], Pr0.6 Sr0.4 MnO3 /SrTiO3 , and Pr0.6 Sr0.4 MnO3 /LaAlO3 [96]

MBE method

La1−x Cax MnO3 [97, 98], La1−x Srx MnO3 [99–104], Sr-doped LaMnO3 , and Nb-doped SrTiO3 [105, 106]

103

104

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing Sputterubg gun Target

Figure 3.8 Diagram of magnetron sputtering system geometry. Source: Eom et al. [107].

Substrate Heater

In situ grown YBa2 Cu3 O7−d thin film [107] and La0.67 Sr0.33 MnO3 film [108] were grown by 90∘ off-axis single-target from a stoichiometric target using magnetron sputtering method that exhibits novel physical and material properties. Here, stoichiometric targets were prepared from “freeze-dried” powders. The deposition processes were performed in a high oxygen partial pressure environment. Substrates were placed on the side of a planar magnetron gun to avoid back sputtering damage from negative oxygen ions as shown in Figure 3.8. The investigation showed that films less than 4 nm thick on are purely oriented and epitaxially aligned in the substrate plane [107]. PLD method is a highly flexible film deposition technique. This method uses a high-power pulsed laser beam inside a vacuum chamber to vaporize target to form a thin film on a substrate under an UHV or in the presence of a background gas. In this fabrication system, high particle energies in the deposition plume can promote surface diffusion and crystal growth [109]. Nowadays, the PLD method can be widely used in most materials, such as metals, refractory metals, and rare earth mainly due to its very successful application to high-temperature superconducting materials. For example, in order to investigate the effects of biaxial strains and microstructure on the magnetic anisotropy, the La0.7 Sr0.3 MnO3 layer was deposited on La0.7 Ca0.3 MnO3 buffer layer to form the epitaxial CMR heterostructure using a PLD system with a 10 Hz of energy density [110]. It was found that the magnetic anisotropy was determined by both the magnetoelastic effect and the magnetocrystalline effect. MBE method is another epitaxy method for thin-film deposition of single crystals, especially for the semiconductor. MBE system can also be modified according to need. Oxygen sources, for example, can be incorporated for depositing oxide materials for advanced electronic, magnetic and optical applications, as well as for fundamental research. For instance, in contrast to the negative CMR of the LaMnO3 compound family, as shown in Figure 3.9, a positive CMR has been discovered in an epitaxial multilayer p–n heterostructure fabricated with Sr-doped LaMnO3 and Nb-doped SrTiO3 by MBE method [106]. The fabrication of artificial crystalline materials through layer-by-layer growth with good control over the composition and structure at the atomic level has become possible and even routine for some oxides. To obtain better p–n interfaces, a computer-controlled MBE system was used to deposit the multilayered p–n heterostructure. In this heterostructure, the CMR effect and the magnetic field modulation of the heterostructure current are very evident in low magnetic fields, even at room temperature [106]. Besides the above mentioned fabrication methods, we also show another two methods to fabricate the CMR devices using typical examples. LaMnO3 sample was

3.3 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing

Figure 3.9 Diagram of the sample and measurement scheme. Source: Lu et al. [106].

I

V In

In SrNb0.05Ti0.95O3 La0.9Sr0.1MnO3

SrNb0.05Ti0.95O3 La0.9Sr0.1MnO3

SrNb0.05Ti0.95O3 La0.1Sr0.1MnO3

SrNb0.05Ti0.95O3 La0.1Sr0.1MnO3

SrNb0.01Ti0.99O3

Monolayer G B Graphene Au

Au BP

SiO2 Heavy doped Si

Vg Multilayer BP

Figure 3.10 Schematic drawing and performance of monolayer graphene on multilayer flake graphene black phosphorus device. Source: Liu et al. [111].

synthesized via a solid-state method using La2 O3 and Mn2 O3 [80]. First, pellets were prepared from the thoroughly mixed powders and allowed to react at 1200 ∘ C for a total time of at least 90 hours, during which they were reground and pelletized at least twice. The pellet was then conditioned in argon at 900 ∘ C to assure the oxygen stoichiometry. Here, oxygen stoichiometry was determined by means of thermogravimetric analysis confirming the correct LaMnO3 stoichiometry, and the CMR effect took place at the percolation threshold. Moreover, a large memory effect was observed together with the CMR, suggesting the presence of magnetic clusters. As shown in Figure 3.10, a phonon-mediated CMR device was fabricated to make a highly sensitive sensor. The fabrication process can be described as follows. The black phosphorus was first exfoliated in a glove box with an argon atmosphere onto a silicon substrate. Then, optical microscopy under the argon atmosphere was used to avoid degradation of black phosphorus and identify the desired black phosphorus. Finally, monolayer graphene was aligned onto the black phosphorus flake with the desired rotation angle in the presence of a transfer platform.

3.3.3

CMR Applications

There exists an increasing demand for magnetoresistance sensors with high sensitivity, low energy consumption, spatial resolution at micro-nanoscales and room temperature operation, which results in a broad of investigations of physical phenomena in advanced materials, and fabrication of novel magnetoresistive device [111].

105

106

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

CMR can be used as phonon-mediated CMR. As shown in Figure 3.10, a phonon-mediated CMR device with a GMR of 775% at 9 T magnetic field exceeding by far the magnetoresistance effects from devices was made from either monolayer graphene or few-layer black phosphorus alone. It shows that electron–phonon coupling between 2D material and a suitable substrate can be exploited to create giant MR effects in Dirac semimetals [111]. Dirsyte et al. etc. [112], proposed and fabricated a novel magnetic field sensor based on the combination of a single-layer graphene and thin nanostructured manganite La0.8 Sr0.2 MnO3 film-hybrid graphene–manganite structure. This hybrid graphene–manganite sensor results in the enhanced sensitivity to the magnetic field of the hybrid sensor on the macroscopic level and has a lower sensitivity to temperature variations in comparison to the manganite sensor and can be applied for position sensing. Based on of CMR-B-scalar sensor that was formed by a thin manganite film deposited on a substrate with metal contacts deposited on it, a pulsed magnetic field measurement system for railgun investigation was also studied [113]. This system was made up of four personal computer-controlled B-scalar meters and each of them was connected by a twisted pair cable. Optionally, each B-scalar meter can be connected to the PC directly via USB. The measurement at each channel can be triggered separately by an electrical or an optical signal. It was reported that the system was able to measure the magnitude of pulsed magnetic fields independently on their orientation in the range from 0.3 to 25 T.

3.4 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing Based on Colossal Tunneling Magnetoresistance (TMR) Effect 3.4.1

Introduction of TMR

There exists a quantum mechanical effect called TMR that depends dramatically on whether the ferromagnets are aligned in parallel or antiparallel in magnetic tunnel junctions made up of ferromagnets/insulator/ferromagnets system. The TMR effect that is a typical example of the spin-dependent electron transport can be evaluated by the ratio of conductance. TMR =

Gp − GAp Gp

(3.2)

Here, Gp and GAp represent the conductances for the parallel and antiparallel alignment of the magnetic tunnel junctions respectively. Illustrated in a single model, the value of TMR is mainly dependent on the intensity of two ferromagnets’ spin polarization at the Fermi energy. The conductance will become larger if they have the same sign of the spin polarization and the magnetic layers are aligned parallel, which is considered as the positive sign of the TMR. However, it is experimentally demonstrated that the conductance of the TMR can be widely

3.4 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing

decided by a broad of atomic and electronic factors including an insulator and the ferromagnet/insulator interfaces, materials of electrodes, the height, the shape and even the disorder of the barriers, and the impurities in the barrier. In the early time, for the barriers, the investigations of tunneling junctions focus on amorphous alumina that makes the TMR ratios of the order of 50%. Later, a new fully crystalline junction with the MgO barrier was proposed in Fe/MgO/Fe system and the ratio exceeded 1000% with just a few atomic layers [114]. The organic single barrier was also investigated and it was found that a large TMR appears in the La2/3 Sr1/3 MnO3 (Co)/quaterthiophene/La2/3 Sr1/3 MnO3 organic magnetic tunnel junction with a large spatial spin polarization [115]. Besides the single insulator layer, metallic spacers were inserted between the insulator and the magnetic film to form the Fe/Au/MgO/Au/Fe(001) system [116]. Interestingly, the ratio was found to oscillate as a function of the metallic spacer thickness, and the epitaxial can endow the system with high TMR ratios. Nowadays, this TMR phenomenon assisted with a broad of fabrication methods has attracted more and more interest due to its broad applications in spin-electronic devices including magnetic sensors and magnetic random access memories.

3.4.2

Fabrication of Multilayered Thin Films of the TMR Effect

In this part, we summarized the main methods to fabricate the TMR devices in the past 10 years. Similar to the fabrication methods of the GMR, CMR sensors and devices, magnetron sputtering, MBE and PLD are still the common tools to be used to fabricate the TMR devices. Furthermore, we will introduce the TMR devices prepared by these three methods and other efficient setups with typical examples respectively. Figure 3.11 presents a magnetic tuning junction consisting of a buffer layer/PtMn/ CoFe/Ru/CoFeB/MgO/CoFeB/cap layer deposited on a SiO2 /Si substrate using magnetron sputtering [117]. In this device, the thickness of every layer was set to be nanoscale. After the deposition process, the multilayer stacks were patterned by milling and annealed in 1 T for 120 minutes at 330 ∘ C to make the CoFeB layers crystallize. It was found the TMR ration in the CoFeB/MgO/CoFeB-based magnetic tunneling junctions can be over 200% at 3 K. Hybrid multilayered system Mo/Ru/Mo/Co20 Fe60 B20 /MgO/Co20 Fe60 B20 /Mo/Ru whose morphology and composition distribution were shown in Figure 3.12, was deposited on the silicon substrate at ambient temperature using a magnetron Figure 3.11 Multilayered magnetic tunnel junction with measurement setup for the resistance. Source: Arakawa et al. [117]. ©2011, AIP Publishing LLC.

Lock-in amplifier

Multimeter

100 kΩ or

PtMn (15)

CoFeB (2) MgO (1.05) CoFeB (3) Ru (0.85) CoFe (2.5) H VTI(2-300 K)

+ – Digitizer + –

107

108

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

(a)

EDX composite

HAADF-STEM (b)

10 nm (c)

Mo

Ru

Fe

Co

Mg

Figure 3.12 (a) TEM image of the cross section of a Mo-based perpendicular magnetic tunnel junction annealed at 673 K for two hours. (b) A composite scanning map energy dispersive X-ray (EDX) spectroscopy of the perpendicular magnetic tunnel junction composition and the colors are assigned in (c). Source: Almasi et al. [118].

STEM

sputtering system equipped with a residual gas analyzer to monitor the H2 O partial pressure [118]. During the deposition process, the Mo and Ru targets were sputtered in a 2 mTorr Ar gas atmosphere using a direct current (DC) sputtering apparatus. The MgO and Co20 Fe60 B20 layer was deposited by a RF magnetron sputtering at 1.3 mTorr of argon gas. After the deposition process, the samples were annealed in a rapid thermal annealing setup in an inert atmosphere. The experimental result highlighted the importance of the heavy-metal layers adjacent to CoFeB electrodes for achieving larger TMR and further demonstrated that the TMR can reach as high as 162% injunctions with Mo layers, due to the improved thermal stability. Figure 3.13 shows a schematic of MgO-based spin-valve multilayered structure fabricated using the magnetron sputtering method with a top Co2 Fe6 B2 free layer and incorporated a single SyAF [Co(0.4 nm)/Pt(0.3 nm)]3 layers and a new buffer layer of Co/Pt/Co. After the deposition process, an ex situ annealing of 350 ∘ C for 30 minutes under a vacuum below 10−6 Torr and a perpendicular magnetic field of 3 T was carried out to ensure a memory margin and avoiding read [119]. As shown in Figure 3.13c, it had a TMR ratio of 180% and anisotropy exchange field of 3.44 kOe. This was due to the fact that the decreasing surface roughness in the fcc–MgO tunneling barrier significantly led to a high level of performance. For the PLD setup, the film grown process can occur in an UHV or in the presence of a background gas. This technique can also be used to grow films to form the magnetic tunnel junction. For instance, epitaxial discontinuous Fe/MgO multilayers was grown by PLD on single-crystal substrates. It was observed that the increasing deposition temperature caused an improvement in crystal quality and was accompanied by higher TMR ratios of 9.2% at room temperature and 18 kOe magnetic field trebles that of polycrystalline samples deposited simultaneously on glass substrates. MBE setup is another method to fabricate TMR devices. Multilayered magnetic tunnel junction made up of a fully epitaxial Fe/MgO/Fe/MgO/Fe was fabricated on MgO single-crystal substrate via an MBE equipment. After the measurement of the bias voltage effects on both TMR and conductance. It was found that the TMR ratio

3.4 Multilayered Magnetic Thin Film for the Electron Transport Control and Sensing

trode Ta/Ru elec W cap p MgO ca e upper fre CoFeB er ac sp W e fre er low Fe/CoFeB er nnel barri MgO tu pinned CoFeB ) 48 (0.18~0. W bridge Co t(0.3)] 3 Co/Pt Ru Co

[Co(0.4)/P

t(0.3)] 6

[Co(0.4)/P

Pt seed Ta buffer

ectrode SiO 2 Si sub.

W/TiN el

(b)

(c) Tex = 350 ºC trode Ta/Ru elec W cap p MgO ca per free up eB CoF er W spac e lower fre Fe/CoFeB er nnel barri MgO tu d ne pin CoFeB 48) (0.18~0. W bridge buffer o Co/Pt/C Ru Co t(0.3)] 3 [Co(0.4)/P Pt seed Ta buffer

ectrode SiO 2 Si sub.

W/TiN el

180

TMRMAX: 156% TMRMAX: 180%

MR ratio (%)

(a) Tex = 350 ºC

150

120

30 Double [Co/Pt]n SyAF Single [Co/Pt]n SyAF

0

0.2

0.3 0.4 tw (nm)

0.5

Figure 3.13 Schematics of double MgO based nanostructures with a top Co2 Fe6 B2 free layer using (a) a double SyAF [Co/Pt]n layers, (b) a single SyAF [Co/Pt]n layer, (c) W thickness-dependent TMR ratio. Source: Choi et al. [119]. © Springer Nature Limited. Licensed under creative commons 4.0.

can reach up to 110% with a single MgO barrier at room temperature. The nanostructure also exhibited a large asymmetry in the bias dependence of the TMR ratio. This was ascribed to the asymmetric conductance curves especially for the parallel magnetization configuration. Besides the abovementioned method, the cluster deposition system was also used to fabricate magnetic tunnel junction. For instance, a tunnel junction was deposited at room temperature on thermally SiO2 substrates, with the layer sequence Ta/Ru/Ta/NiFe/IrMn/CoFe/Ru/Co40 Fe40 B20 (CoFeB)3 /MgO/CoFeB/Ta/Ru. After deposition of the bottom electrodes using DC sputtering, both MgO layers were deposited by RF sputtering using two MgO targets in a target-facing-target gun. Subsequently, top electrodes were sputtered. The pressure for the MgO layers’ growth was ranged from 1.3 to 4.0 mTorr. It was demonstrated that frequency 1/f barrier noise as low as 2.5–3.3 E−12 μm2 with a TMR ratio of up to 330% at room temperature was observed [120]. Chemical vapor deposition (CVD) technique is also an efficient method to fabricate the TMR sample. NiFe/Gr–hBN/Co magnetic tunnel junction was fabricated via thermal CVD technique [121]. First, Cu foil was moved into a CVD furnace tube. After the furnace pressure and temperature were set to be 0.1 mTorr and 1010 ∘ C respectively, the CH4 and H2 gases entered into the furnace at a flow rate of 20 and 5 sccm, respectively, to form graphene. Subsequently, the sample was cooled to room temperature. The hBN was grown on the Cu foil in a similar way. Here, the Cu foil temperature was adjusted to 990 ∘ C for 30 minutes with a H2 gas flow rate of 5 sccm. After the cleaning process, borazine and H2 gases entered into the furnace tube and the temperature was kept at 997 ∘ C for 30 minutes to form the hBN layer

109

110

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

and then, the temperature was adjusted to 500 ∘ C. Then, the polymethyl methacrylate (PMMA) was coated on the hBN surface. The new sample was then immersed in an aqueous solution of ammonium persulfate to remove the Cu foil. After that, deionized water was used to wash the sample. The new sample then entered into CVD grown graphene on Cu foil. The samples were then dried on a hot plate at 80 ∘ C. Again, the bottom Cu foil was removed through etching by keeping the sample in an ammonium persulfate solution. After this, Gr/hBN/PMMA was washed with deionized water for cleaning. Next, Gr/hBN/PMMA was transferred upon SiO2 /Si substrate where the NiFe electrode had already been deposited through metal masks. The samples were then placed in acetone for the lift-off PMMA. Consequently, the top Co electrode was then patterned on the top surface of hBN using metal masks. The final device configuration was Co/Gr–hBN/NiFe/SiO2 . It was observed that the ratio of TMR monotonically increases with decreasing temperature. The magnitude of the TMR ratio is found to progress from −0.85% at room temperature to −1.88% at 4.2 K.

3.4.3

TMR Applications

The TMR effect in magnetic tunnel junctions [122] is the key to developing magnetoresistive random-access-memory [123, 124], novel programmable logic devices [125], and magnetic sensor that poses a bright perspective in micro- and nano-scale strain sensing technologies [126, 127]. Recently, tunnel magneto-resistance sensors made up of magnetic tunnel junctions have been used to measure weak bio-magnetic fields [128]. For instance, CoFeB/MgO/CoFeB magnetic tunnel junctions with a perpendicularly magnetized synthetic antiferromagnetic reference layer were fabricated for magnetic sensor applications and this sensor showed dynamic ranges more than 62.5 kOe [129]. Meiling Wang et al. [128], measured the triaxial magnetocardiography signal using a high sensitivity tunnel magnetoresistance sensor. All the biomagnetic field measurements were performed in a magnetically shielded room to counteract the influence of external magnetic fields. The screening factor of the magnetically shielded room was about 60 [email protected] Hz and the residual magnetic field in the magnetically shielded room was about 5 nT. The measurement can be carried out at room temperature and this would make magnetocardiography instrumentation cheaper for extensive clinical applications. In contrast to the behavior of conventional TMR sensors under mechanical stress as well as their sensitivity to the applied stress that depends on the magnetization configuration of magnetic tunnel junctions with respect to the stress axis, as shown in Figure 3.14, CoFeB/MgO/CoFeB-based system using an inverse effect on the tunnel resistance by tensile and compressive stresses was proposed [130]. Field and strain loops on this proposed configuration were calculated. It was demonstrated that the junction exhibited big factors of 2150±30 and −260 for tensile and compressive stresses, respectively, under a −3.2 kA/m bias magnetic field. This configuration result provided a design path for high sensitivity and ability to detect both tensile and compressive stresses by a single TMR sensor.

3.5 The Multilayered Magnetic Thin Film Based on Quantum Anomalous Holzer Effect (QAHE)

Figure 3.14 Schematic of the TMR sensor in which applied press makes the black ceramic pieces move, and then the ceramic transfers the stress signal to the TMR junction. Source: Tavassolizadeh et al. [130].

Fixed Al body 18 mm TMR junction

6 mm ΔZ

Pusher block H

3.5 The Multilayered Magnetic Thin Film Based on Quantum Anomalous Holzer Effect (QAHE) The quantum Holzer effect is a quantized version of the Holzer effect observed in two-dimensional (2D) electron systems under approximately 10 T magnetic field strength that makes the experimental realization challenging and greatly hinders real-world applications. However, QAHE that may be a consequence of the combined spin–orbit coupling and reduction of the time-reversal symmetry due to intrinsic magnetization, can be realized without an external magnetic field that may lead to the development of low-power-consumption electronics [131]. Nowadays, broad investigations have been carried out to seek new platforms for the realization of the QAHE [132–134]. Among the investigations, graphene-like honeycomb materials [135, 136] and magnetically doped topological insulators [131, 137] have mainly attracted increasing attention. Efforts on the investigations of the magnetically doped topological insulators system lead to the first experimental realization of the QAHE. In 2013, QAHE was experimentally observed in a magnetic topological insulator, namely, thin films of chromium-doped (Bi,Sb)2 Te3 [131], in which it was found that at zero magnetic fields and under low temperature, the gate-tuned anomalous Holzer resistance reached the quantized value of h/e2 (h is Planck’s constant and e is the elementary charge), accompanied by a considerable drop in the longitudinal resistance. Although the temperature of the observation of the QAHE of the intrinsic magnetic topological insulators has been increased much, QAHE is still a challenge for further electron transport control and signal sensing applications. In this chapter, we will mainly introduce the recent efforts on the experimental investigation of the QAHE of intrinsic magnetic topological insulators, including the fabrication method and its application for the electron transport control and signal sensing applications, although the investigations of the QAHE mainly focus on the simulation [132, 138–140]. Intrinsic magnetic topological insulator Cr0.15 (Bi0.1 Sb0.9 )1.85 Te3 with a thickness of five periodic layers grown on dielectric SrTiO3 (111) substrates by MBE setup was investigated [131]. In this structure, the chemical potential can be adjusted to the electron- or hole-conductive regime by a positive or negative gate voltage

111

112

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

respectively, due to the fact that the film thickness in this system was nearly charged neutral. In order to carry out the transport measurements, the system was manually cut into a hall bar nanoarrays after the growth process. The measurement results demonstrated that the QAHE was first achieved with low mobility without any magnetic field in this intrinsic magnetic topological insulator system and this system that was fabricated via the MBE paved a path for the further development of the low-power-consumption, topological quantum electronic and spintronic devices. (Bi1−y Sby )2 Te3 and Zn1−x Crx Te were chosen as a topological insulator and ferromagnetic insulator respectively to form a Zn1−x Crx Te/(Bi1−y Sby )2 Te3 /Zn1−x Crx Te sandwich heterostructure via MBE method [141]. The detailed process was as following: InP(111)A substrate was first installed to the chamber to experience an annealing process at 350 ∘ C in a vacuum. Then, a 2 nm thick ZnTe buffer layer was grown on semi-insulating InP(111)A substrate. Subsequently, Zn1−x Crx Te (10 nm)/(Bi1−y Sby )2 Te3 (8 nm)/Zn1−x Crx Te (10 nm) sandwich heterostructure was gown on the substrate. Finally, a 2 nm thick ZnTe buffer layer was adopted to improve the crystallinity of the Zn1−x Crx Te layer. After the growth process, and AlOx protecting layer with a thickness of approximately 5 nm was deposited via atomic layer deposition at room temperature to prevent deterioration of the sample. In this sandwich heterostructure, QAHE was driven by the magnetic proximity coupling. Nowadays, the discovery of the QAHE makes a promise for the construction of the spintronic devices with lower dissipation as a consequence of the dissipationless nature of the hall currents and chiral propagation. The spintronic application based on the domain wall is a prototypical example, owing to its non-volatile nature. In this application, racetrack memory has been expected to be the next new generation spintronics memory device [139] due to a prominent advantage that the device does not need electricity to keep information. However, the challenge is that a current-driven device of ferromagnetic metals suffers from Joule heating. Fortunately, the QAHE or Weyl semimetal states of kagome layered materials were proposed to overcome this barrier. It is reported that in these materials, domain walls can be driven by the electric field and a large electric current that causes Joule heating is not necessary [142–144]. A quantum anomalous Holzer insulator with the amagnetic configuration of a domain wall was proposed [142]. In this system, a magnetically doped topological insulator with the configuration of a domain wall was connected by two voltage-controlled electrical leads. The opposite QAHEs in two domains were generated due to the domain-wall configuration and the domain wall can be removed in the presence of an external magnetic field. This device can provide an efficient scheme to reconfigure the domain-wall chiral interconnects for possible memory and can also be used as a magnetoelectric piston when the insulator was contacted by electrical reservoirs.

3.6 Summary and Perspectives Carrying information in both the charge and spin of an electron potentially offers devices with a greater diversity of functionality, and with the development of the

Acknowledgments

spintronic, novel magnetotransport phenomena appear due to the GMR, CMR, TMR, and QAHE. In this chapter, we summarized the main fabrication methods of composition and microstructure defined magnetic thin films for these four quantum effects. Their novel electron transport control were also discussed according to their unique signal sensing applications for the integrated information technology. (1) GMR that exists in multilayered films is a significant change in the electrical resistance depending on whether the magnetization of adjacent ferromagnetic layers is in a parallel or an antiparallel alignment in the presence of an external magnetic field. Among the fabrication methods, the MBE method, magnetron sputtering method, and electron beam deposition method are three main methods to fabricate the samples. GMR has been widely used in the emerging applications including eddy current sensing probe, ultra-low pressure measurement, biological magnetic sensors, some novel technologies in water pollution detection, wireless charging for electric vehicles, and even current monitoring for smart power grids. (2) CMR effect mostly based on manganese-based perovskite oxides demonstrates many orders of magnitude exceeding GMR of multilayered and granulated systems. The devices with CMR effect can be fabricated via magnetron sputtering method, PLD, MBE and other methods such as solid-state method. CMR can be used as a phonon-mediated CMR magnetic field sensor applied for position sensing and pulsed magnetic field measurement. (3) TMR that exists in magnetic tunnel junctions consisting of ferromagnets/insulator/ferromagnets system is the key to developing magnetoresistive random-access-memory, novel programmable logic devices, and magnetic sensor that poses a bright perspective in micro- and nano-scale strain sensing technologies. Magnetron sputtering, MBE, PLD, cluster deposition system, and CVD technique can be used to fabricate TMR devices. (4) QAHE that exists in graphene-like honeycomb materials and magnetically doped topological insulators is a consequence of the interplay between spin–orbit coupling and reduction of the time-reversal symmetry due to intrinsic magnetization. The discovery of QAHE in intrinsic magnetic topological insulator system without an external magnetic field paves a path for the further development of the low-power consumption, topological quantum electronic and spintronic devices. (5) The development and advanced application of controllable preparation methods of quantum effect, structure and composition in thin films, and how to combine with optical communication and sensing technology to break through the bottleneck of molar effect and build the next generation of high-speed and high-throughput quantum communication technology are prospected.

Acknowledgments This chapter is supported by the NSFC–BRICS STI Framework Program (No. 51861145309), the National Natural Science Foundation of China (No. 51971029),

113

114

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

the National S&T Major Project (No. 2018ZX10301201), the Postdoctor Research Foundation of Shunde Graduate School of University of Science and Technology Beijing (No. 2020BH005), the Project funded by China Postdoctoral Science Foundation (No. 2020M680336), the “All English teaching demonstration course construction project of University of Science and Technology Beijing” (No. KC2015QYW06, 2016), the “1125” Zhihui Zhengzhou Talent project of Henan province (Fund No. in USTB: 39080070), the “100 talent plan” fund of Fujian province (Fund No. in USTB: 39080067), and the development of a high sensitive magneto-optical biomolecular sensor experimental prototype (Fund No. in USTB: 2019-0649) by Hangzhou Ruidi Biotechnology Co. Ltd.

List of Abbreviation and Symbol AMR GMR CMR TMR QAHE MBE MRAM PLD DC RF PMMA 2D

anisotropic magnetoresistance giant magnetoresistance colossal magnetoresistance tunneling magnetoresistance quantum anomalous Holzer effect molecular beam epitaxy magnetoresistive random access memory pulsed laser deposition direct current radio frequency polymethyl methacrylate two-dimensional

References 1 Inoue, J.I. (2014). GMR, TMR, BMR, and related phenomena. In: Nanomagnetism and Spintronics, 2e (ed. T. Shinjo), 15–106. Oxford: Elsevier. 2 Mattis, D.C. and Butrymowicz, D.B. (1965). The theory of magnetism. Physics Today 18 (11): 66–66. 3 Slade, H.D. and Vetter, J.K. (1956). Studies on Streptococcus pyogenes. II. Observations on the microscopical and biological aspects of the disintegration and solubilization of a type 6 strain by shaking with glass beads. Journal of Bacteriology 72 (1): 27–33. 4 Gerlach, W. (1922). Der experimentelle nachweis der richtungsquantelung im magnetfeld. Zeitschrift für Physik 9 (1): 349–352. 5 Zhang, W., Wang, Q., Zhao, C., and Song, Y. (2019). The optical cavity enhanced magneto-optical Kerr effect signals of AAO/Al-based CoFeB nanostructure arrays. Optics Communications 437: 44–49. 6 Zhang, W.W., Liu, Z.H., Belotelov, V.I. et al. (2018). The magnetic properties of CoFeB and CoFeB/Ag nanodot arrays fabricated by a template transfer imprinting method. Thin Solid Films 660: 301–305.

References

7 Tian, Q., Zhang, W., Belotelov, V., and Song, Y. (2018). Tunable magnetic properties of the nanoporous hybrid multilayer arrays. Modern Physics Letters B 32 (17): 1850191. 8 Zhang, W., Li, J., Ding, X. et al. (2017). Tunable magneto-optical Kerr effects of nanoporous thin films. Scientific Reports 7 (1): 2888. 9 Wang, J. and Song, Y. (2017). Microfluidic synthesis of nanohybrids. Small 13 (18): 1604084. 10 Ma, J., Wang, J., Zhong, X. et al. (2016). Synthesis of Sn(1−x) Fex @Fey Sn(1−y) Oz nanohybrids via a simple programmed microfluidic process. RSC Advances 6 (87): 84255–84261. 11 Wang, J., Zhao, K., Shen, X. et al. (2015). Microfluidic synthesis of ultra-small magnetic nanohybrids for enhanced magnetic resonance imaging. Journal of Materials Chemistry C 3 (48): 12418–12429. 12 Song, Y., Wang, R., Rong, R. et al. (2011). Synthesis of well-dispersed aqueous-phase magnetite nanoparticles and their metabolism as an MRI contrast agent for the reticuloendothelial system. European Journal of Inorganic Chemistry (22): 3303–3313. 13 Li, J., Zhang, W., Song, Y. et al. (2016). Template transfer nanoimprint for uniform nanopores and nanopoles. Journal of Nanomaterials 2016: 1–7. 14 Parkin, S., Xin, J., Kaiser, C. et al. (2003). Magnetically engineered spintronic sensors and memory. Proceedings of the IEEE 91 (5): 661–680. 15 Baibich, M.N., Broto, J.M., Fert, A. et al. (1988). Giant magnetoresistance of (001)Fe/(001)Cr magnetic superlattices. Physical Review Letters 61 (21): 2472–2475. 16 Cros, V., Fert, A., Seneor, P., and Petroff, F. (2009). The 2007 Nobel Prize in physics: Albert Fert and Peter Grunberg. The Spin 55: 147–157. 17 Candid, R., María-Dolores, C.B., and Diego Ramírez, M.O. (2009). Magnetic field sensors based on giant magnetoresistance (GMR) technology: applications in electrical current sensing. Sensors 9 (10): 7919–7942. 18 Grunberg, P., Schreiber, R., Pang, Y. et al. (1987). Layered magnetic structures: evidence for antiferromagnetic coupling of Fe layers across Cr interlayers. Physical Review Letters 61 (8): 3750–3752. 19 Alvarado, S.F. and Carbone, C. (1988). Magnetism and epitaxy of Fe/Cr(001) multilayers. Physica B+C 149 (1–3): 43–48. 20 Barthelemy, A., Fert, A., Baibich, M.N. et al. (1990). Magnetic and transport properties of Fe/Cr superlattices. Journal of Applied Physics 67 (9): 5908–5913. 21 Parkin, S.S.P., More, N., and Roche, K.P. (1990). Oscillations in exchange coupling and magnetoresistance in metallic superlattice structures: Co/Ru, Co/Cr, and Fe/Cr. Physical Review Letters 64 (19): 2304–2307. 22 Parkin, S.S.P., Li, Z.G., and Smith, D.J. (1991). Giant magnetoresistance in antiferromagnetic Co/Cu multilayers. Applied Physics Letters 58 (23): 2710–2712. 23 Araki, S., Yasui, K., and Narumiya, Y. (1991). Giant magnetoresistance and oscillatory magnetic coupling of evaporated and MBE-grown Co/Ag multilayers. Journal of the Physical Society of Japan 60 (9): 2827–2830.

115

116

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

24 Greig, D., Hall, M.J., Hammond, C. et al. (1992). The giant magnetoresistance of Co/Cu superlattices grown by MBE. Journal of Magnetism and Magnetic Materials 110 (3): L239–L246. 25 Hall, M.J., Hickey, B.J., Howson, M.A. et al. (1992). Interface scattering and the giant magnetoresistance of MBE-grown Co/Cu superlattices. Journal of Physics: Condensed Matter 4 (36): L495–L502. 26 Avdi, E., Hickey, B.J., Greig, D. et al. (1993). Giant magnetothermopower and giant magnetoresistance in molecular beam epitaxy grown Co/Cu(111) superlattices. Journal of Applied Physics 73 (10): 5521–5523. 27 Hall, M.J., Hickey, B.J., Howson, M.A. et al. (1993). Magnetoresistance of Co/Cu superlattices grown by molecular beam epitaxy. Physical Review B: Condensed Matter 47 (19): 12785–12793. 28 Howson, M.A., Hickey, B.J., Xu, J. et al. (1993). Oscillations in the exchange coupling for (111)-oriented Co/Cu magnetic multilayers grown by molecular-beam epitaxy. Physical Review B: Condensed Matter 48 (2): 1322–1324. 29 Howson, M.A., Hickey, B.J., Xu, J., and Greig, D. (1993). Magnetisation and giant magnetoresistance of Co/Cu(111)-oriented MBE-grown magnetic multilayers: antiferromagnetic coupling and in-plane anisotropy. Journal of Magnetism and Magnetic Materials 126 (1–3): 416–418. 30 Kobayashi, Y., Aoki, Y., Sato, H. et al. (1993). Giant magnetoresistance in MBE-grown Co/Cu multilayers. Journal of Magnetism and Magnetic Materials 126 (1–3): 501–503. 31 Clarke, R., Barlett, D., Tsui, F. et al. (1994). An alternate route to giant magnetoresistance in MBE-grown Co–Cu superlattices. Journal of Applied Physics 75 (10): 6174–6177. 32 Kobayashi, Y., Sato, H., Aoki, Y., and Kamijo, A. (1994). The giant magnetoresistance and the anomalous Hall effect in molecular-beam-epitaxy grown Co/Cu superlattices. Journal of Physics: Condensed Matter 6 (36): 7255–7267. 33 Kyeong-Ik, M., Hoo-San, L., and Seung-Ki, J. (1994). Effects of low temperature annealing on the giant magnetoresistance phenomenon in Co/Cu artificial superlattices prepared by sputtering. Korean Applied Physics 7 (5): 393–398. 34 Soliman, A., Schreiber, F., Meckenstock, R. et al. (1994). Ferromagnetic resonance studies of anisotropy in Co/Cu(111) multilayers. Physica Status Solidi B 186 (2): 437–441. 35 Thomson, T., Riedi, P.C., and Greig, D. (1994). Interfacial quality and giant magnetoresistance in MBE-grown Co/Cu(111) superlattices. Physical Review B: Condensed Matter 50 (14): 10319–10322. 36 Tsui, F., Baoxing, C., Barlett, D. et al. (1994). Scaling behavior of giant magnetotransport effects in Co/Cu superlattices. Physical Review Letters 72 (5): 740–743. 37 Wellock, K.P., Hickey, B.J., Greig, D. et al. (1994). The effect of Au impurities at the interfaces on the magnetoresistance of MBE-grown Co/Cu multilayers. Journal of Applied Physics 75 (10): 7055–7057.

References

38 Pollard, R.J., Wilson, M.J., and Grundy, P.J. (1995). Giant magnetoresistance and microstructure in highly oriented Co/Cu multilayers sputter-deposited on (111) silicon. Journal of Magnetism and Magnetic Materials 151 (1–2): 139–144. 39 Wellock, K.P., Hickey, B.J., Xu, J. et al. (1995). The role of interface impurities in Co/Cu giant magnetoresistance multilayers. Journal of Magnetism and Magnetic Materials 140–144: 585–586. 40 Wellock, K., Theeuwen, S.J., Caro, J. et al. Giant magnetoresistance of magnetic multilayer point contacts. Physical Review B: Condensed Matter 1999, 60 (14): 10291–10301. 41 Dupuis, V., Maurer, M., Piecuch, M. et al. (1993). Structure and magnetism of Mn/Ru and Mn/Fe/Mn/Ru superlattices prepared by molecular-beam epitaxy. Physical Review B: Condensed Matter 48 (8): 5585–5593. 42 Chaiken, A., Gutierrez, C.J., Krebs, J.J., and Prinz, G.A. (1993). Composition dependence of giant magnetoresistance in Fe/Ag/Cox Fe1−x sandwiches. Journal of Magnetism and Magnetic Materials 125 (1–2): 228–238. 43 Parkin, S.S.P., Fan, S., More, N., and Roche, K.P. (1990). Giant magnetoresistance in sputtered magnetic superlattices. Journal of Applied Physics 67 (9): 5931–5931. 44 Parkin, S.S.P. (1991). Giant magnetoresistance and oscillatory interlayer exchange coupling in copper based multilayers. MRS Proceedings 231: –211. 45 Parkin, S.S.P. (1992). Dramatic enhancement of interlayer exchange coupling and giant magnetoresistance in Ni81 Fe19 /Cu multilayers by addition of thin Co interface layers. Applied Physics Letters 61 (11): 1358–1360. 46 Jimbo, M., Kanda, T., Goto, S. et al. (1992). Giant magnetoresistance in FeNiCo/Cu multilayers. Japanese Journal of Applied Physics, Part 2: Letters 31 (9B): L1348–L1350. 47 Highmore, R.J., Shih, W.C., Somekh, R.E., and Evetts, J.E. (1992). Sputter deposition of giant magnetoresistive Co–Cu multilayers. Journal of Magnetism and Magnetic Materials 116 (1–2): 249–258. 48 Tan, M., Barnard, J.A., Parker, M.R., and Seale, D. (1993). Giant magnetoresistance in Co/Ag multilayers[J]. MRS Proceedings 313: 191. 49 Gang, X., Jian Qing, W., and Peng, X. (1993). Giant magnetoresistance and its evolution in the granular Fex Ag100−x system (0 ≤ x ≤ 100). Applied Physics Letters 62 (4): 420–422. 50 Highmore, R.J., Somekh, R.E., Shih, W.C. et al. (1993). Studies of giant magnetoresistive Co–Cu multilayers. Applied Surface Science 66 (1–4): 124–130. 51 Parkin, S.S.P. and York, B.R. (1993). Influence of deposition temperature on giant magnetoresistance of Fe/Cr multilayers. Applied Physics Letters 62 (15): 1842–1844. 52 Liou, S.H., Malhotra, S.S., and Yung, S.W. (1993). Magnetoresistance in Co/Ag and Fe/Ag nanocomposite films. Journal of Applied Physics 73 (10): 5539–5539. 53 Jimbo, M., Tsunashima, S., Kanda, T. et al. (1993). Giant magnetoresistance in soft magnetic NiFeCo/Cu multilayers. Journal of Applied Physics 74 (5): 3341–3344.

117

118

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

54 Honda, S., Ohmoto, S., Imada, R., and Nawate, M. (1993). Giant magnetoresistance in Co/Cu multilayers sputter-deposited on glass substrates. Journal of Magnetism and Magnetic Materials 126 (1–3): 419–421. 55 Kanda, T., Jimbo, M., Tsunashima, S. et al. (1994). Giant magnetoresistance in NiFeCo/Cu multilayers. IEEE Translation Journal on Magnetics in Japan 9 (1): 103–109. 56 Beach, R.S., Rao, D., Carey, M.J. et al. (1994). Giant magnetoresistance in heterogeneous (CoFe)x Ag1−x films. Journal of Applied Physics 75 (10): 6933. 57 Kyeong-Ik, M. and Seung-Ki, J. (1994). Enhancement of magnetoresistance sensitivity in a Co/Cu artificial superlattice. Journal of Applied Physics 75 (9): 4632–4635. 58 Kubota, H. and Miyazaki, T. (1994). Composition dependence of the magnetoresistance for Ni–Co/Cu and Co–Fe/Cu multilayer films. IEEE Translation Journal on Magnetics in Japan 9 (6): 291–297. 59 Milyaev, M., Naumova, L., Proglyado, V. et al. (2019). High GMR effect and perfect microstructure in CoFe/Cu multilayers. IEEE Transactions on Magnetics 55 (4): 1–4. 60 Shkurdoda, Y.O., Dekhtyaruk, L.V., Basov, A.G. et al. (2019). The giant magnetoresistance effect in Co/Cu/Co three-layer films. Journal of Magnetism and Magnetic Materials 477: 88–91. 61 Li, G., Shen, H., Shen, Q. et al. (2000). Influence of Si buffer layer on the giant magnetoresistance effect in Co/Cu/Co sandwiches. Science China Technological Sciences 43 (3): 225–231. 62 Maji, N., Kar, U., and Nath, T.K. (2019). GMR based spin valve like magnetic diode exploiting negative JMR in Heusler alloy thin film/semiconductor heterostructures. Materials Today: Proceedings 11: 673–678. 63 Kumar, A. and Srivastava, P.C. (2015). Magnetic, morphological and structural investigations of CoFe/Si interfacial structures. Journal of Experimental Nanoscience 10 (10): 803–818. 64 Vyatkin, A.F., Matveev, V.N., Volkov, V.T. et al. (2013). Multilayer nanostructures with the effect of a giant magnetoresistance. Journal of Nano and Microsystem Technologies 2013, 11: 26–29. 65 Jepu, I., Porosnicu, C., Lungu, C.P. et al. (2014). Combinatorial Fe–Co thin film magnetic structures obtained by thermionic vacuum arc method. Surface and Coatings Technology 240 (7): 344–352. 66 Henini, M. (2000). Handbook of Thin-Film Deposition Processes and Techniques (ed. K.K. Schuegraph), 219. Noyes, ISBN: 0-8155-1153-1. Microelectronics Journal 2000, 31 (3). 67 Postolache, O., Ramos, H.G., Ribeiro, A.L., and Alegria, F.C. (2009). GMR based eddy current sensing probe for weld zone testing. Sensors: 73–78. 68 Borole, U.P., Subramaniam, S., Kulkarni, I.R. et al. (2018). Highly sensitive giant magnetoresistance (GMR) based ultra low differential pressure sensor. Sensors and Actuators A: Physical 280 (2018): 125–131. 69 Gaster, R.S., Hall, D.A., Nielsen, C.H. et al. (2009). Matrix-insensitive protein assays push the limits of biosensors in medicine. Nature Medicine 15 (11): 1327–1332.

References

70 Lee, J.R., Haddon, D.J., Wand, H.E. et al. (2016). Multiplex giant magnetoresistive biosensor microarrays identify interferon-associated autoantibodies in systemic lupus erythematosus. Scientific Reports 6 (1): 27623. 71 Ng, E., Nadeau, K.C., and Wang, S.X. (2016). Giant magnetoresistive sensor array for sensitive and specific multiplexed food allergen detection. Biosensors and Bioelectronics 80: 359–365. 72 Xu, L., Yu, H., Akhras, M.S. et al. (2008). Giant magnetoresistive biochip for DNA detection and HPV genotyping. Biosensors and Bioelectronics 24 (1): 99–103. 73 Rizzi, G., Lee, J.-R., Dahl, C. et al. (2017). Simultaneous profiling of DNA mutation and methylation by melting analysis using magnetoresistive biosensor array. ACS Nano 11 (9): 8864–8870. 74 Zhang, L., Zhang, Y., Ma, L. et al. (2019). Research about the water pollution detection technologies based on GMR sensor. In: ADVANCES IN MATERIALS, MACHINERY, ELECTRONICS III: 3rd International Conference on Advances in Materials, Machinery, Electronics (AMME 2019). 75 Ouyang, Y., Wang, Z.X., Zhao, G. et al. (2019). Current sensors based on GMR effect for smart grid applications. Sensors and Actuators A: Physical 294: 8–16. 76 Li, Y., Wang, Y., Klein, T., and Wang, J.-P. External-field-free magnetic biosensor. Applied Physics Letters 104 (12): 122401. 77 Daughton, J.M. (1999). GMR applications. Journal of Magnetism and Magnetic Materials 192 (2): 334–342. 78 Ramirez, A.P. (1997). Colossal magnetoresistance. Journal of Physics: Condensed Matter 9 (39): 8171–8199. 79 Rodriguez-Martinez, L.M. and Attfield, J.P. Cation disorder and size effects in magnetoresistive manganese oxide perovskites. Physical Review B: Condensed Matter 54 (22): R15622–R15625. 80 Baldini, M., Muramatsu, T., Sherafati, M. et al. (2015). Origin of colossal magnetoresistance in LaMnO3 manganite. Proceedings of the National Academy of Sciences 112 (35): 10869–10872. 81 Lee, K.R., Chung, Y.J., Lee, J.H. et al. (2003). Electrical and magnetic properties of La0.7 Ca0.3 MnO3 films prepared by RF magnetron sputtering method for colossal magnetoresistance applications. Thin Solid Films 426 (1–2): 205–210. ˇ ˙ K.; Cerniuk ˙ I.; Butkute, ˙ R.; Lisauskas, V.; Maneikis, A. 82 Vengalis, B.; Šliužiene, e, (2008). Preparation and properties of hybrid bilayer structures based on organic Alq3, ferromagnetic La2/3 Sr1/3 MnO3 and Fe3 O4 . Proceedings of Spie the International Society for Optical Engineering, 7142. 83 Li, H.Q., Fang, Q.F., and Zhu, Z.G. (2002). Preparation and colossal magnetoresistance in a trilayer La0.67 Sr0.33 MnO3 /La0.75 MnO3 /La0.67 Sr0.33 MnO3 device by DC magnetron sputtering. Materials Research Bulletin 37 (5): 859–866. 84 Prokhorov, V.G., Kaminsky, G.G., Kim, J.M. et al. Origin of an enhanced colossal magnetoresistance effect in epitaxial Nd0.52 Sr0.48 MnO3 thin films. Low Temperature Physics 37 (4): 305. 85 Srivastava, M.K., Kaur, A., and Singh, H.K. (2010). Huge anisotropic magnetoresistance in epitaxial Sm0.53 Sr0.47 MnO3 thin films. The 55th DAE Solid State Physics Symposium. AIP Conference Proceedings 2011, 1349 (1), 703–704.

119

120

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

86 Domínguez, C., Ordoñez, J.E., Gómez, M.E., and Lopera, W. (2014). Study of exchange bias effect in bilayers based on La2/3Ca1/3MnO3and BiFeO3. Journal of Physics: Conference Series 480: 012029. 87 Zhang, L., Singh, V., Rajapitamahuni, A., and Hong, X. (2014). Anisotropic magnetoresistance in colossal magnetoresistive oxide La1−x Sr x MnO3 thin films. Journal of Applied Physics 102: 103901. 88 Hong, X. (2006). Ferroelectric field effect studies of the colossal magnetoresistive oxide lanthanum(1−x) strontium(x) manganese oxide. Yale University. 89 Broussard, P.R., Cestone, V.C., and Qadri, S.B. Characterization of off-axis sputtered La0.67 Ca0.33 MnO3 films and La0.67 Ca0.33 MnO3 /YBa2 Cu3 O7−δ bilayers. MRS Proceedings 474: 235. 90 Prada, R.D., Golod, T., Bernhard, C., and Krasnov, V.M. (2019). Growth and nanofabrication of all-perovskite superconducting/ferromagnetic/ superconducting junctions. Journal of Superconductivity and Novel Magnetism 32 (9): 2721–2726. 91 Holzapfel, B., Hutten, A., Eckert, D. et al. (1997). Colossal magnetoresistance in La1−x Pbx MnO3 thin films prepared by pulsed laser deposition. Journal De Physique IV 7 (C1): 637–638. 92 Wei, Z., Boyd, I.W., Elliott, M., and Herrenden–Harkerand, W. (1997). Colossal magnetoresistance behavior in LaCaMnO/CeO2 /Si(100) multilayers grown by pulsed laser deposition. Journal of Applied Physics 81 (8): 5759. 93 Chao, H., Duan, P., Jiyou, W. et al. (2011). Colossal magnetoresistance effect in perovskite La0.9 Sb0.1 MnO3 films prepared by pulsed laser deposition. In: 2011 International Conference on Multimedia Technology, 2044–2046. 94 Wang, H.O., Chu, Z., Su, K.P. et al. (2016). Colossal magnetoresistance of Pr0.7 Sr0.3 MnO3 layer grown on charge-ordered La0.5 Ca0.5 MnO3 manganite layer. Journal of Alloys and Compounds 689: 69–74. 95 Wang, H.O., Su, K.P., Huang, S. et al. (2017). Colossal magnetoresistance at wide temperature range in Pr0.7 Sr0.3 MnO3 film grown on (0001) sapphire. Journal of Materials Science: Materials in Electronics 28 (15): 11275–11278. 96 Kumary, T.G., Amaladass, E.P., Nithya, R., and Mani, A. (2016). Strain-enhanced colossal magnetoresistance in Pr0.6 Sr0.4 MnO3 thin films. Journal of Superconductivity and Novel Magnetism 29 (10): 2685–2690. 97 Odonnell, J., Onellion, M., Rzchowski, M.S. et al. (1997). Anisotropic properties of molecular beam epitaxy-grown colossal magnetoresistance manganite thin films. Journal of Applied Physics 81 (8): 4961–4963. 98 O’Donnell, J., Andrus, A.E., Oh, S. et al. (2001). Growth of “colossal” magnetoresistance heterostructures by molecular beam epitaxy. In: Magnetoresistive Oxides and Related Materials, vol. 602 (eds. M.S. Rzchowski, M. Kawasaki, A.J. Millis, et al.), 9–14. 99 O’Donnell, J., Andrus, A.E., Oh, S. et al. (2000). Colossal magnetoresistance magnetic tunnel junctions grown by molecular-beam epitaxy. Applied Physics Letters 76 (14): 1914–1916. 100 Adamo, C., Mechin, L., Heeg, T. et al. (2015). Enhanced electrical and magnetic properties in La0.7 Sr0.3 MnO3 thin films deposited on CaTiO3 -buffered silicon substrates. APL Materials 3 (6): 10.

References

101 Galceran, R., Balcells, L., Pomar, A. et al. (2016). Tunneling anisotropic magnetoresistance in La2 /3 Sr1 /3 MnO3 /LaAlO3 /Pt tunnel junctions. AIP Advances 6 (4): 143–236. 102 Mercone, S., Belmeguenai, M., Malo, S. et al. (2017). Investigation of ferromagnetic heterogeneities in La0.7 Sr0.3 MnO3 thin films. Journal of Physics D: Applied Physics 50 (4): 045001. 103 Niu, W., Gao, M., Wang, X. et al. (2016). Evidence of weak localization in quantum interference effects observed in epitaxial La0.7 Sr0.3 MnO3 ultrathin films. Scientific Reports 6 (1): 26081. 104 Niu, W., Wang, X., Gao, M. et al. (2017). Unsaturated magnetoconductance of epitaxial La0.7 Sr0.3 MnO3 thin films in pulsed magnetic fields up to 60 T. AIP Advances 7 (5): 056404. 105 Jin, K.J., Lu, H.B., Zhou, Q.L. et al. (2005). Positive colossal magnetoresistance from interface effect in p−n junction of La0.9 Sr0.1 MnO3 and SrNb0.01 Ti0.99 O3 . Physical Review B 71 (18): 184428. 106 Lu, H.B., Yang, G.Z., Chen, Z.H. et al. (2004). Positive colossal magnetoresistance in a multilayer p–n heterostructure of Sr-doped LaMnO3 and Nb-doped SrTiO3 . Applied Physics Letters 84 (24): 5007–5009. 107 Eom, C.B., Sun, J.Z., Yamamoto, K. et al. (1989). In situ grown YBa2 Cu3 O7−d thin films from single-target magnetron sputtering. Applied Physics Letters 55 (6): 595–597. 108 Tsui, F., Smoak, M.C., Nath, T.K., and Eom, C.B. (2000). Strain-dependent magnetic phase diagram of epitaxial La0.67 Sr0.33 MnO3 thin films. Applied Physics Letters 76 (17): 2421–2423. 109 Lunney, J.G. Pulsed laser deposition of metal and metal multilayer films. Applied Surface Science 86 (1–4): 79–85. 110 Berndt, L.M., Balbarin, V., and Suzuki, Y. (2000). Magnetic anisotropy and strain states of (001) and (110) colossal magnetoresistance thin films. Applied Physics Letters 77 (18): 2903–2905. 111 Liu, Y., Yudhistira, I., Yang, M. et al. (2018). Phonon-mediated colossal magnetoresistance in graphene/black phosphorus heterostructures. Nano Letters 18 (6): 3377–3383. 112 Lukose, R., Zurauskiene, N., Balevicius, S. et al. (2019). Hybrid graphene-manganite thin film structure for magnetoresistive sensor application. Nanotechnology 30 (35): 355503. 113 Stankevic, T., Medisauskas, L., Stankevic, V. et al. (2014). Pulsed magnetic field measurement system based on colossal magnetoresistance-B-scalar sensors for railgun investigation. Review of Scientific Instruments 85 (4): 5. 114 Mathon, J. and Umerski, A. Theory of tunneling magnetoresistance of an epitaxial Fe/MgO/Fe(001) junction. Physical Review B 63 (22): 220403. 115 Han, X., Mi, W., and Wang, X. (2019). Large magnetoresistance and spin-polarized photocurrent in La2 /3 Sr1 /3 MnO3 (Co)/quaterthiophene/La2 /3 Sr1 /3 MnO3 organic magnetic tunnel junctions. Journal of Materials Chemistry C 7 (14): 4079–4088. 116 Mathon, J. and Umerski, A. (2005). Theory of resonant tunneling in an epitaxial Fe/Au/MgO/Au/Fe(001) junction. Physical Review B 71 (22): 402.

121

122

3 Multilayered Magnetic Thin Films for Electron Transport Control and Signal Sensing

117 Arakawa, T., Sekiguchi, K., Nakamura, S. et al. (2011). Sub-Poissonian shot noise in CoFeB/MgO/CoFeB-based magnetic tunneling junctions. Applied Physics Letters 98 (20): 031001. 118 Almasi, H., Hickey, D.R., Newhouse-Illige, T. et al. (2015). Enhanced tunneling magnetoresistance and perpendicular magnetic anisotropy in Mo/CoFeB/MgO magnetic tunnel junctions. Applied Physics Letters 106 (18): 182406. 119 Choi, J.Y., Lee, D.G., Baek, J.U., and Park, J.G. (2018). Double MgO-based perpendicular magnetic-tunnel-junction spin-valve structure with a top Co2 Fe6 B2 free layer using a single SyAF [Co/Pt]n layer. Scientific Reports 8 (1): 2139. 120 Feng, J.F., Chen, J.Y., Kurt, H., and Coey, J.M.D. Very low 1/f barrier noise in sputtered MgO magnetic tunnel junctions with high tunneling magnetoresistance. Journal of Applied Physics 112 (12): 123907. 121 Iqbal, M.Z., Siddique, S., Hussain, G., and Iqbal, M.W. (2016). Room temperature spin valve effect in the NiFe/Gr–hBN/Co magnetic tunnel junction. Journal of Materials Chemistry C 4 (37): 8711–8715. 122 Moodera, J.S., Kinder, L.R., Wong, T.M., and Meservey, R. Large magnetoresistance at room temperature in ferromagnetic thin film tunnel junctions. Physical Review Letters 74 (16): 3273–3276. 123 Inomata, K., Saito, Y., Nakajima, K., and Sagoi, M. Double tunnel junctions for magnetic random access memory devices. Journal of Applied Physics 87 (9): 60–64. 124 Min, B.C., Shin, I.J., Choi, G.M. et al. (2010). MgO-based magnetic tunnel junctions for spin-transfer-torque random access memory. In: 10th IEEE International Conference on Nanotechnology, 17-20 Aug. 2010, 144–147. 125 Kimura, H. (2005). TMR-based logic-in-memory circuit for low-power VLSI. ITC-CSCC Digest of Technical Paper, 2004 8 (6): 1408–1415. 126 Wolf, S.A., Awschalom, D.D., Buhrman, R.A. et al. (2001). Spintronics: a spin-based electronics vision for the future. Science 294 (5546): 1488–1495. 127 Ney, A., Pampuch, C., Koch, R., and Ploog, K.H. Programmable computing with a single magnetoresistive element. Nature 425 (6957): 485–487. 128 Wang, M., Peng, L., and Ye, C. (2019). Magnetocardiography Measurements with a High Sensitivity Tunnel Magnetoresistance Sensor. In: 2019 IEEE International Conference on Sensors and Nanotechnology, 24–25 July 2019, 1–4. IEEE. 129 Nakano, T., Oogane, M., Furuichi, T., and Ando, Y. Magnetic tunnel junctions using perpendicularly magnetized synthetic antiferromagnetic reference layer for wide-dynamic-range magnetic sensors. Applied Physics Letters 110 (1): 012401. 130 Tavassolizadeh, A., Rott, K., Meier, T. et al. (2016). Tunnel magnetoresistance sensors with magnetostrictive electrodes: strain sensors. Sensors 16 (11): 1902. 131 Chang, C., Zhang, J., Feng, X. et al. (2013). Experimental observation of the quantum anomalous Hall effect in a magnetic topological insulator. Science 340 (6129): 167–170.

References

132 Hu, J., Zhu, Z., and Wu, R. (2015). Chern half metals: a new class of topological materials to realize the quantum anomalous hall effect. Nano Letters 15 (3): 2074–2078. ´ B.K. (2017). Monolayer of the 5d transition 133 Sheng, X.-L. and Nikolic, metal trichloride OsCl3 : a playground for two-dimensional magnetism, room-temperature quantum anomalous Hall effect, and topological phase transitions. Physical Review B 95 (20): 201402. 134 Qiao, Z., Yang, S.A., Feng, W. et al. (2010). Quantum anomalous Hall effect in graphene from Rashba and exchange effects. Physical Review B 82 (16): 161414. 135 Onoda, M. and Nagaosa, N. (2003). Quantized anomalous hall effect in two-dimensional ferromagnets: quantum hall effect in metals. Physical Review Letters 90 (20): 06601. 136 Weng, H., Yu, R., Hu, X. et al. (2015). Quantum anomalous Hall effect and related topological electronic states. Advances in Physics 64 (3): 227–282. 137 Liu, C.-X., Qi, X.-L., Dai, X. et al. (2008). Quantum anomalous hall effect in Hg1−y Mny Te quantum wells. Physical Review Letters 101 (14): 146802. 138 Lan, T.-B., Xu, Y., Tan, H. et al. (2019). Quantum anomalous Hall effect with Landau levels in nonuniformly strained silicene. Journal of Applied Physics 126: 104303. 139 Kobayashi, K., Takagaki, M., and Nomura, K. (2019). Robust magnetotransport in disordered ferromagnetic kagome layers with quantum anomalous Hall effect. Physical Review B 100 (16): 161301. 140 Kovalev, V.M. and Savenko, I.G. (2019). Quantum anomalous valley Hall effect for bosons. Physical Review B 100 (12): 121405(R). 141 Watanabe, R., Yoshimi, R., Kawamura, M. et al. (2019). Quantum anomalous Hall effect driven by magnetic proximity coupling in all-telluride based heterostructure. Applied Physics Letters 115 (10): 102403. 142 Upadhyaya, P. and Tserkovnyak, Y. (2016). Domain wall in a quantum anomalous Hall insulator as a magnetoelectric piston. Physical Review B 94 (2): 020411. 143 Araki, Y., Yoshida, A., and Nomura, K. (2016). Universal charge and current on magnetic domain walls in Weyl semimetals. Physical Review B 94 (11): 115312. 144 Kurebayashi, D. and Nomura, K. (2019). Theory for spin torque in Weyl semimetal with magnetic texture. Scientific Reports 9 (1): 5365.

123

125

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics: Part A – Fabrication Methods and Microstructure Property Study Andrey A. Voronov 1,2 , T. Mikhailova 3 , Olga V. Borovkova 1,2,4 , Alexander N. Shaposhnikov 3 , Vladimir N. Berzhansky 3 , and Vladimir I. Belotelov 1,2,3 1 Lomonosov Moscow State University, Faculty of Physics, Department of Photonics and Microwaves Structures, Leninskie Gory, Moscow 119991, Russia 2 Russian Quantum Center, 45, Skolkovskoye Shosse, Moscow 121353, Russia 3 V.I. Vernadsky Crimean Federal University, Physics and Technology Institute, Department of Experimental Physics, Vernadsky Avenue 4, Simferopol 295007, Russia 4 Lomonosov Moscow State University, Faculty of Physics, Department of Physics of Oscillations, Leninskie Gory, Moscow 119991, Russia

4.1 Introduction Iron garnets, whose properties were comprehensively studied in the middle of the twentieth century [1], can be attributed to ferrites – magnetically ordered substances with low conductivity, in which the magnetic moments of the atoms of different sublattices are oriented antiparallel, but the moments of the sublattices are not equal, which leads to a spontaneous nonzero magnetization. Due to the combination of a sufficiently large Faraday effect [2, 3] with high transparency (compared with metallic ferromagnets), this material has found a wide application in magneto-optics. One should additionally note a great interest in iron garnet due to the inverse Faraday effect and high-quality spin waves excited in this material [4], spintronics [5], and Bose–Einstein condensation of magnons [6]. The general structural formula of iron garnets is {R3 }3+ [Fe2 ]3+ (Fe3 )3+ (O12 )2− where R is a rare-earth element or yttrium (Y, Gd, Tb, Dy, Ho, Er, Sm, or Eu). In general case iron can also be partially replaced by ions of other metals. The iron garnet is a cubic body-centered syngony, and is formed by oxygen ions O2− , in the voids between which there are ions of rare-earth elements and iron. There are three types of voids in the garnet structure: tetrahedral, where the metal cation is surrounded by four oxygen ions; octahedral, by six oxygen ions; and dodecahedral, by eight oxygen ions. As a result, Fe3+ iron cations usually occupy tetrahedral and octahedral positions, and rare-earth cations R3 + are dodecahedral [7]. The iron garnet fabrication as epitaxial films allows one to vary the chemical composition, and the presence of three cationic positions enables to introduce Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications, First Edition. Edited by Yujun Song. © 2021 WILEY-VCH GmbH. Published 2021 by WILEY-VCH GmbH.

126

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

more than half of the chemical elements of the periodic table, which contributes to imparting certain physical properties to certain samples and setting magnetic and optical parameters. The presence of three magnetic sublattices connected by a ferromagnetic interaction and induced magnetic anisotropy allows one to vary the size of the domains (from 10−7 to 10−3 m), the saturation magnetization (from 0 to 1.5 × 102 kA/m), the uniaxial anisotropy constant (from −104 to +104 J/m3 ), and the Hilbert attenuation parameter (from 10−5 to >1). Such a wide range of property changes contributes to the suitability of iron garnet for several applications [8]. Iron garnets are characterized by high transparency in the near-infrared (IR) region of the spectrum in the wavelength range of 1300–5500 nm; at 𝜆 > 5500 nm there is an increase in optical absorption associated with the absorption of photons by the crystal lattice, and absorption at 𝜆 < 1500 nm is associated with the edge of the electron band, whose center is approximately 900 nm. The optical absorption coefficient κ in the transparency region is less than 0.1 cm−1 . At the electromagnetic wavelength range from 100 μm and up to the microwaves, iron garnets have high transparency, which is also important for applied problems [7]. However, for the problems of magneto-optics, the visible and near-IR spectral regions are of interest. The maximal (up to 3 deg/μm) Faraday rotation (FR) is characteristic of yttrium iron garnet [3] and its Bi-containing varieties [8]. It is established that the introduction of Bi into the composition of iron garnet contributes to a significant increase in the magneto-optical (MO) qualities of the material. In this chapter we will describe the fabrication methods of the bismuth-substituted iron garnet (BiIG) films and some types of magnetophotonic crystals (MPCs) with transparent BiIG layers, MO properties of such MPCs, and MO response enhancement in the ultrathin BiIG films with plasmonic covering as Part A. The unique MO properties and large values of MO effects in heterostructures based on iron garnet films open up prospects for the use of such materials for novel sensors. In particular, we will further consider some typical applications of these iron garnet films and iron garnet-based metamaterials for magnetic field sensing, biosensing, and eddy current flaw detection as Part B in the following chapter.

4.2 Fabrication Methods 4.2.1 Synthesis Technology and Conditions of Bismuth-substituted Iron Garnet Films BiIG films are often used as basic elements of one-dimensional magneto-photonic crystals (1D-MPCs) as they exhibit high transmittance and large specific Faraday rotation (SFR) in the visible and near-IR optical spectrum regions [7, 9–12]. Depending on the 1D-MPC optical spectrum operating range, the thicknesses of these films can vary from a few tens to several hundred nanometers. The structural quality of the magneto-active layer and characteristics of film−substrate transitional interface can significantly affect the 1D-MPC parameters such as the FR magnitude, transmittance, and, as a result, MO figure of merit (FM). Thus the investigation of the

4.2 Fabrication Methods

Figure 4.1 The scheme of synthesis of ultrathin BiIG films, deposited by RIBS (at the left) and crystallized by annealing (at the right). Source: Okuda et al. [9] Li et al. [20]; Parkin et al. [21]; Baibich et al. [22]; Candid et al. [23]; Grunberg et al. [24]; Alvarado and Carbone [25].

Furnace

Argon Oxygen

Target

Ion gun Substrate

Air sample

Sputtered material

optical and MO properties of ultrathin BiIG films, as a function of the substrate surface and other synthesis conditions, is of great interest. The understanding of the kinetics of film growth is crucial to the development of the new generation of MO devices. To clarify the influence of different grown factors, we discuss the results on the optimization of synthesis conditions of ultrathin BiIG garnet films, deposited by reactive ion-beam sputtering (RIBS) [9, 13–19]. RIBS technology consists in the processes of bombardment of ceramic target of certain composition by ion beam and the subsequent deposition of material onto a substrate (Figure 4.1) [9, 13–19]. To obtain high-quality films, it is necessary to satisfy the optimum synthesis conditions. It is assumed that the chemical composition of the garnet phase is closest to the target composition; specific FR has a maximum value, and optical absorption and surface roughness are minimal at optimum synthesis conditions. At the first stage of the technological process, the seven targets of the following compositions were used for BiIG sputtering: ● ● ● ● ● ● ●

M1: Bi1.0 Y0.5 Gd1.5 Fe4.2 Al0.8 O12 , aM1 = 1.2444 nm. M2: Bi2.8 Y0.2 Fe5 O12 , aM2 = 1.2600 nm. M3: Bi1.5 Gd1.5 Fe4.5 Ga0.5 O12 , aM3 = 1.2535 nm. M4: Bi2.5 Gd0.5 Fe3.8 Al1.2 O12 , aM4 = 1.2502 nm. M5: Bi2.5 Y0.5 Fe5 O12 , aM5 = 1.2576 nm. M6: Bi0.9 Gd1.4 Lu0.7 Fe4.1 Al0.9 O12 , aM6 = 1.2404 nm. M7: Bi1.5 Gd1.5 Fe4.5 Al0.5 O12 , aM7 = 1.2506 nm.

Here, the lattice constants a are shown for BiIG corresponding to the target composition. The constants were calculated using Vegard’s law. The lattice constant of yttrium iron garnet a YIG = 1.2376 nm was taken as the basic parameter. The selection of target chemical compositions was determined by optimization of film characteristics such as FR angle value ΘF , saturation magnetization H S , coercive force H c , compensation T comp , Curie temperatures T c , transmittance K t , and squareness ratio of MO Faraday hysteresis loops (FHLs) K g . The chemical compositions of the synthesized targets and films were studied on an REM-106 scanning electron microscope equipped with an EDS-1 energy-dispersive spectrometer. The lattice parameters of the synthesized films were determined with a DRON-3 diffractometer. The FR ΘF , the coercivity H c , compensation T comp , and Curie temperatures T c were determined from FHLs using Faraday magneto-polarimeter operating at 𝜆 = 655 nm in 20–150 ∘ C temperature range.

127

128

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

The targets were prepared by conventional ceramic technique [26] in several stages: (i) Cold pressing of homogenized mixtures at the pressure of 300 kg/cm2 . (ii) First solid-phase synthesis in air at atmosphere pressure and temperatures 800–1000 ∘ S for 8 or 12 hours. (iii) Grinding, repeated homogenization, and cold pressing of mixtures. (iv) Second solid-phase synthesis at the same conditions. According to X-ray phase analysis, two phases of Bi2 Fe4 O9 and BiFeO3 are the main constituents of M2 target after second solid-phase synthesis. Mass fractions of the phases were 22% and 78%, respectively. In other targets, a little shift of those phase peaks as compared to the M2 target proved the presence phases comprising Gd, Lu, and Al. The targets were fabricated in the form of disks with a diameter of 120 mm. Single-crystal plates of gadolinium–gallium garnet (GGG), calcium–niobium– gallium garnet (CNGG), and calcium–manganese–zirconium–gadolinium–gallium garnet (CMZGG) of orientation (111) with lattice parameters aGGG = 1.2383 nm, aCNGG = 1.2507 nm, and aCMZGG = 1.2495 nm were used as substrates, respectively. In addition, fused silica KU-1 and heat-resistant optical glass were applied. The synthesis of BiIG films was carried out at a URM 3-279.014 setup in oxygen–argon mixture (Figure 4.1) using an ion-beam Kholodok-1 source. The films were grown by two different methods: (i) During deposition on “hot” substrate (in situ). (ii) Sputtering to a “cold” substrate (radiative heating to 80 ∘ C) and subsequent annealing in a vacuum chamber (post-annealing crystallization). In the second method, the amorphous layer deposited on a “cold” substrate was subjected to annealing in a furnace (Figure 4.1). The main variable parameters for RIBS method are the substrate temperature T S , partial pressure of working gases (argon PAr and oxygen PO2 ), residual pressure Pb , and accelerating voltage U at the anode of ion source, ion-beam current J, and the “target−substrate” distance. These parameters determine the target sputtering r S and deposition r D rates. The rate r D of the layers on a “hot” substrate was relatively low, from 1.5 to 2.5 nm/min. The rate r D on a “cold” substrate varied from 4.8 to 10 nm/min. The rate r S of films with a high Bi content was higher than a rate rS of films with a lower Bi content. The rate r D was determined by the deposition time 𝜏 S and the thickness of already synthesized films h. The film thickness was determined using a Biolar PI polarization-interference microscope and MII-4 Linnik micro-interferometer or was calculated from the transmission spectra of the films. Subsequently, the rate r D was used to calculate the thickness of films of various series obtained at the same conditions and from the same target. At the post-annealing crystallization, the choice of annealing temperature T a , annealing time 𝜏 a , and heating rate during the annealing is decisive. Table 4.1 shows the optimal synthesis conditions of BiIG films on different substrates.

4.2 Fabrication Methods

According to microanalysis data, the sputtering and annealing regimes at low value of T S and T a for films of high bismuth content are optimal, which is associated with the substantial difference in the condensation temperature of bismuth-containing vapor in comparison to other garnet-forming elements. In addition, the bismuth evaporation during annealing is influenced. A detailed description of the technological process and the characteristics of BiIG films of different compositions is given [15–19]. The BiIG films grown by RIBS with post-annealing crystallizations have the best optical and MO performance for multilayer structures [16]. The temperature dependencies of FR and FHLs at the some T a for M1 and M2 films are shown in Figure 4.2a. For M2 films on GGG, increasing of T a from 500 to 680 ∘ S occurs with FR increasing from zero to maximum value −5.5 deg/μm at T a = 660 ∘ C. M2-garnet phase is completely destroyed at T a > 950 ∘ C (Figure 4.2a, curve 1). The maximum value of ΘF = −1.9 deg/μm for M1 films on GGG is achieved at the range of T a from 720 to 820 ∘ C (Figure 4.2a, curve 2). The maximum value of ΘF = −0.9 deg/μm for the M1 films on SiO2 is reached at T a range from 680 to 700 ∘ C. An interesting peculiarity of crystallization processes of the M1 films on SiO2 is the inversion of FR sign at T a > 900 ∘ C (Figure 4.2a, curve 3). This can be attributed to the thermo-activated redistribution of garnet constitutive elements and bismuth reevaporation at high temperature and garnet formation of another composition [15]. Thus, at the temperature region from 660 to 680 ∘ C, FR has maximum values for all the films. Thus, we crystallized M2 and M1 films at optimal for both compositions T a = 680 ∘ C The lattice parameter for M2 films is 1.2620 nm and these films have monocrystalline quality. M1 films on SiO2 have polycrystals with different anisotropies as confirmed by our ferromagnetic resonance (FMR) experiments [27]. Presented in Figure 4.2b,c are resonance fields as functions of azimuthal angle at rotation in plane of M1 layer (b) and bilayer M1/M2 (c) on GGG substrates. These dependencies are evidence of monocrystalline quality of M2 films on GGG, which are typical for cubic symmetry with (111) plane. The high quality of M2 films is proved by the maximum value of ΘF = −5.5 deg/μm. The cause of some resonance characteristic asymmetry in film plane is possible off-orientation of (111) plane relative to the film normal. It is interesting that the signal symmetry of structure GGG/M1/M2 differs from the signal symmetry of M2 film on GGG. Such signal symmetry of M2 film on sub-layer M1 is characteristic for the crystallographic plane (100). The FMR line width ΔH is the structural sensitivity parameter, which for M2 film on GGG is ΔH = 600 Oe and on sublayer M1 is ΔH = 100 Oe. This is because the lattice mismatch of M2 film and GGG is more than one of M2 and M1 films. In the result, the stresses in structure GGG/M2 are more than in the M1/M2 bilayer. Using atomic force microscopy data on the surface topography [19, 28], we also found that high (up to 80%) oxygen content in the gas mixture at sputtering and low heating rate r (∼2 deg/min) during crystallization annealing provided a more low surface roughness 𝜎 rms without changing the value of MO effects.

129

Table 4.1

Optimal synthesis conditions of BiIG films.

Substrate

Target

Sputtering and crystallization conditions U (kV)

J (mA)

P b (Torr)

P Ar (Torr)

PO2 (Torr)

T s (∘ S)

5 × 10−4

T a (∘ S)

𝝉 a (min)

In situ

GGG

0.5 × 10−4

520–540





M3

0.6 × 10−4

590





M4

0.5 × 10−4

560





M5

0.5 × 10−4

560





M6

0.6 × 10−4

650





0.5 × 10−4

80

M2

5

100

5 × 10−6

Post-annealing crystallization (air, atmospheric pressure) GGG

M1

5

5 × 10−6

5 × 10−4

680–690

15–20

650–680

15–20

M3

650–680

15–20

770

180

M5

730

180

M6

780

180

680–690

15–20

540–550

15–20

M4

Silica/SiO2 Quartz/SiO2

160

M2 100

M1



0.5 × 10−4

160

Post-annealing crystallizations (vacuum) GGG

M2

5

90–100

5 × 10−6

5 × 10−4

0.5–0.8

80

M5

0.7

520

M7

0.7

650

4.2 Fabrication Methods Hres(Oe)

1 ΘF(º/μm)

3680

3 0 –1

500

600

700

800

900

1000 Ta (ºC)

3640

–2

3620

2 (b)

–3

4800 4700 4600 4500 4400 4300

–4 –5 –6 (a)

3660

1 (c)

0 60 120 180 240 300 360 𝜑 (º) Hres(Oe)

0

60 120 180 240 300 360 𝜑 (º)

Figure 4.2 (a) Specific FR ΘF as a function on the annealing temperature T a for the films: M2 on GGG (1), M1 on GGG (2), and M1 on SiO2 (3). FHLs at some T a are shown in the insets. FHL scale for (2) and (3) is multiplied by three in comparison with (1). Azimuthal dependencies of resonance field Hres for M2 layer (b) and bilayer M1/M2 (c) on GGG. Source: Berzhansky et al. [16].

As expected, during optimization of synthesis conditions arise the technological difficulties of deposition and crystallization of garnet films with a high bismuth content (for example, M2 and M3) on the SiO2 layers, silica, or quartz substrate. To effective growth of garnet phases with high bismuth content, a two-step synthesis technology has been proposed [26, 29–31]. The sub-layer with a lower bismuth content (less than 1 at./f.u.) was deposited on SiO2 layer, silica, or quartz substrate. Then, after crystallization annealing of sub-layer, the main magneto-active layer with a higher Bi content (more than 1.5 at./f.u.) was deposited and crystallized by annealing. Due to the proposed method of crystallization, new record values of the FR angle of the microcavity and Tamm 1D-MPCs in optical wavelength range were achieved [31–33] (see also hereinafter). The state of the substrate surface determines the crystallization processes of the films and the surface roughness, structure, and properties of “substrate–film” transition layer. To determine the effect of substrate pretreatment on the properties of synthesized M2 films, the substrates were processed before the film deposition by argon and oxygen ions with varying energies and currents [18, 19]. The duration of ion pretreatment was five minutes. Some of the GGG substrates were also pre-annealed in air at the atmospheric pressure. The sputtered amorphous films were crystallized by annealing at the temperature T = 650 ∘ C for 20 minutes. According to the substrate pretreatment regimes, the films were divided into four types (Table 4.2). An ion source with a cold cathode “Radical” II-4-015 was used for substrate pretreatment and thermal activation. The dependencies of the specific FR on the film thicknesses h for three types of BiIG M2 films are shown in Figure 4.3. The corresponding FHLs are schematically shown in the insets. At room temperature, FR drops to zero for all film types when h is below 6 nm. I-type films demonstrate a negative FR and have an easy-plane or close to an easy-plane magnetic

131

132

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

Table 4.2

The regimes of the substrate pretreatment.

Film types

Gas

Ion energy

Ion current density (mA/cm2 )

Substrate pre-annealing, T a and 𝝉 a

I

Ar

Low-energy plasma

1.0



II

Ar

Ion beam, 1.0 keV

2.5



III

Ar

Ion beam, 4.0 keV

5.0



IV

O

Low-energy plasma

1.0

In air, 800 ∘ S , 60 min

ΘF (°/μm) 2.0 1.0 0 ‒1.0

h(nm) 10

20

30

40

50 90

120

150 180

‒2.0 ‒3.0 ‒4.0 ‒5.0 ‒6.0

II III I

Figure 4.3 Thickness dependencies of FR and FHLs for M2 films of types I, II, and III. Source: Shaposhnikov et al. [19].

anisotropy for all thicknesses in the studied range, which is typical for the thick BiIG films prepared using conventional techniques. However, the saturation magnetic field of ultrathin I-type films (H S is approximately 4.5 kOe) is much higher than that of the thick ones (H S is approximately 1.0 kOe). The FR angle approaches the bulk material value for h, above 12 nm, which is approximately 10 lattice periods. It reaches the lower limit, ΘF = −5.8 deg/μm, for h ≥ 100 nm. There is no compensation point in this type of films. The measured Curie temperature is T C = 350 ∘ C. A quite different behavior was observed for II- and III-type films. Here, there is an interval of thicknesses where FR is positive. The FR angle becomes negative only above a critical thickness hcr (10.6 nm for II-type films and 8.3 nm for III-type ones). The films display perpendicular anisotropy for all thicknesses in the [hmin , hcr ] interval. The increase of Ar+ ion energy as well as the ion-beam current density leads to the narrowing of the interval while the maximum value of FR becomes larger. Above hcr , the FR angle is negative. The observed FR sign inversion in II- and III-type films is related to the variation of T comp with the film thickness. Our measurements of T comp for II-type films show that it drops from 70 to 28 ∘ C when h increases from 9.7 to 11.2 nm. Thus, for thin films with h < hcr , the compensation point is above the room temperature, and the magnetization of their octahedral sublattices exceed the magnetization of the tetrahedral ones. Above the critical thickness, the situation

II type h = 8.7 nm h = 5.8 nm

5

A, 103 cm–1

A, 103 cm–1

10 I type

Phase diagram

4.2 Fabrication Methods

0

10

FIM I

I type II type

Ma + Mc

PM

FIM II

FIM I Md

Ma + M c

Md External H

I type

5 0

–5

II type

–5 h = 8.7 nm h = 5.8 nm

–10 400

300

500

300 400 λ, nm

(a) I type

A, 103 cm–1

4

–10 500

600

(b)

5

10

15 h, nm

20

25

II type

300 K

2

250 K 150 K 110 K 50 K 8K

0

–2

(c)

300

400

500 λ, nm

600

700

Figure 4.4 (a) MCD spectra of I- and II-type films at 300 K. (b) Thickness dependencies of amplitude of MCD long-wave peak A at 300 K (bottom) and phase diagram of films (at the top). (c) MCD spectra of 2.9 nm I-type film at 300 K and II-type film at different temperatures. The value of MCD signal is twofold increased for film of II-type at temperature 250 K and is halved for film of I-type. Source: Berzhansky et al. [18].

is reversed. A similar change of T comp with thickness is observed in III-type films. Here, the compensation point drops below room temperature at a lower thickness (of about 8 nm). The highest values of ΘF = −7.8 deg/μm at 𝜆 = 655 nm were obtained for 100-nm-thick IV-type film. The substrate pretreatment for these films were practically analogous to I-type films except for the fact that the oxygen plasma was used for the processing instead of argon plasma. As a result, the thickness dependence of FR for IV-type films is similar to I-type one. The increase of FR in these films is most probably related to lower destruction of the substrate surface at oxygen plasma treatment (decrease of breaking bond number on the surface and its amorphization). The results were confirmed by magnetic circular dichroism (MCD) measurements of the films [18, 19]. MCD spectral measurements were performed using a Jobin–Yvon dichrograph at H = 5.5 kOe in 270–850 nm wavelength range with the step 1 nm. All MCD spectra of I- and II-type films in the investigated range of thickness (at least for h > 5.8 nm) have the form that is typical for spectra of BiIG with high content of bismuth: two peaks of opposite signs and the point of intersection with the wavelength axis (Figure 4.4a) [34–36]. The positions and amplitudes of short- and long-wave peaks as well as the value of wavelength for zero point depend on the resonant frequency, half-width, and intensity (the density of active ions and oscillator strength) of MO transitions [7, 35].

133

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

Since in the case of II-type films the substrate surface pretreatment leads to the substrate surface destruction and amorphization, the film thickness is reduced (below 15 nm), MCD spectra of II-type films demonstrate features of Ga and Gd dilution: the “blue shift” and change in the intensity of MO transitions. MCD and FR measurements at 300 K show that as II-type film thickness decreases, the series of magnetic phase transition appear (Figure 4.4b). Inversion of signs observed at the critical thickness of hcr = 11 nm suggests a concentrational spin-orientation phase transition from ferrimagnetic of I-type (I-type FIM) to ferrimagnetic of II-type (II-type FIM). According to MCD spectra measurements at different temperature for 2.9 nm II-type films, both long-range magnetic order and MCD spectra typical for iron garnet are formed at temperatures less than 130 K (Figure 4.4c). As a result, when thickness approaches 5 nm, the second phase transition from the ferrimagnetic phase II-type FIM to paramagnetic (PM) occurs. The distribution of Ga is not uniform along the film thickness, and the thickness of film–substrate transition layer of II-type films is 15 nm. The influence of the substrate surface pretreatment on the properties of I-type films is much less pronounced. MO activity at 300 K is present through all thickness ranges, even for the films of thicknesses 1.5 and 2.9 nm. Inverse effects are absent and “blue shift” is negligible. However, decreases of the intensity of MO transitions is still observed at thicknesses less than hcr = 11 nm. To clarify the reasons of inversion, the next experiment was realized [17]. At the beginning, a series of the films of II-type with h = 8.2 nm was synthesized. Then, the nanolayers of the thickness from 0.5 to 3.0 nm were deposited on these films from target of the same composition M2 and separately crystallized under the same conditions as bottom layers. The results of investigation of these bilayer nanostructures are shown in Figure 4.5. Here FHLs of the film with h = 8.2 nm (a) and bilayer nanostructures (8.2 + 0.5) nm (b), (8.2 + 1.5) nm (c), and (8.2 + 3.0) nm (d) are presented. As one can see, the film with h = 8.2 nm is characterized by the positive sign of FR and the so-called “left” FHL. The top BiIG film of thickness 0.5 nm essentially (fourfold) reduces the value of FR without leading to a change in the sign of effect (Figure 4.5b). The top BiIG films of thicknesses 1.5 nm and above considerably change the shapes of FHL (and, consequently, the type of magnetic anisotropy) and the sign of effect (Figure 4.5c,d). The reason for the sign inversion of MO effects (or MO transitions) in the vicinity of hcr is reorientation of sublattice magnetization 0.08

(a) h = 8.2 nm

(b) h = (8.2 + 0) nm

(c) h = (8.2 + 1.5) nm

(d) h = (8.2 + 3.0) nm

0.04

ΘF (°)

134

0.00

–0.04 –0.08 –2

–1

0

1

2 –2

–1

0

1

2 –2

H (kOe)

–1

0

1

2 –2

–1

0

1

Figure 4.5 FHLs of bottom 8.2 nm film (a) and bilayer structures: (8.2 + 0.5) nm (b), (8.2 + 1.5) nm (c), and (8.2 + 3.0) nm (d). Signal of substrate was substracted. Source: Berzhansky et al. [17].

2

4.2 Fabrication Methods

respectively to direction of external field, i.e. so-called spin-orientation phase transition that was discussed hereinabove (Figure 4.4b) [37]. T comp ≈ 40 ∘ S. But at small fields (up to 300 Oe), T comp is much greater. This allows controlling the compensation point by value of external magnetic field, and it is the evidence of possible existence of spin-flop phase in bilayer nanostructure in intermediate state, in which MO effects disappear and its dependencies pass through zero at hcr . The Curie temperature T S of such nanostructure is achieved at 152 ∘ S. High value of N s for nanostructure of the thickness (8.2 + 1.5) nm (Figure 4.5c) may be due to the closeness of its T comp to temperature of measurements. The compensation temperature in the bottom film of composition M2 is absent, and T S ≈ 360 ∘ S. Thus, the top film of thickness just 0.5 nm with negative sign of FR substantially changes temperature and field dependencies of initial film of thickness 8.2 nm with the positive sign of FR. Possible mechanisms of the spin-reorientation phase transition observed in the bilayer structures are as follows: (i) Concentration mechanism. After the top BiIG layer deposition, the integral content of Bi3+ and Fe3+ ions in the structure increases due to the interdiffusion of layers and exceeds threshold value at critical thickness of the top layer ∼1.5 nm and above. This results to appearance of compensation point in the investigated temperature interval and the sign inversion of FR. Stratification of composition in thickness in this case is created artificially and can be controlled. (ii) The mechanism of exchange interaction between layers. During crystallization annealing from the sputtered top film of thicknesses 1.5 nm and above, a monolayer of iron garnet of composition M2 with negative FR is synthesized on a layer of mixed composition (Bi,Gd,Y)3 (Fe,Ga)5 O12 with positive FR. Antiferromagnetic exchange interaction between the layers appears. As a result, the magnetic moment of bottom layer is reoriented, and summary FR of bilayer nanoscale structure became a negative.

4.2.2

Fabrication of Fabry–Perot 1D-MPC with BiIG Bilayer

Fabry–Perot one-dimensional magnetophotonic crystals (FP-1D-MPCs) with transparent BiIG layers are the first structures of MPCs proposed to control optical response by MO effects [38, 39]. According to experimental results obtained by various groups [39–41], all FP-1D-MPCs based on non-garnet Bragg mirrors (BMs) showed the values of MO quality factor Q not higher than 6∘ and the values of figure of merit F not higher than 4% in the range of wavelengths from 550 to 850 nm. Here MO quality factor Q and figure of merit F of 1D-MPC are determined according to [39–42] as Q = −2|ΘF |∕ ln(T) (∘ )

(4.1)

F = T ⋅ sin(2ΘF ) (%)

(4.2)

where ΘF and T are FR and transmittance of 1D-MPC, respectively. These characteristics evaluate the success of practical implementation of 1D-MPCs with MO layers.

135

136

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

Figure 4.6 Schematic representation of FP-1D-MPCs on the basis of BiIG with m = 5.

MO layer

SiO2 TiO2

Substrate

The results of numerical calculations for configuration [Ta2 O5 /SiO2 ]m /Bi:YIG/ [SiO2 /Ta2 O5 ]m with different repetition number m potentially showed F = 12.5% [42]. In order to increase the characteristics Q and F of FP-1D-MPCs based on non-garnet BM, we used of a BiIG bilayer that allow to overcome technological difficulties of deposition and crystallization of garnet films with a high bismuth content on SiO2 layers [31]. FP-1D-MPCs with general formula [TiO2 /SiO2 ]m /M/[SiO2 /TiO2 ]m were proposed, where M is a bilayer MO defect and m is the repetition number of layer pairs in BM. The structure is shown schematic in Figure 4.6. The fabrication process of the whole structure included the following three stages [32]: (i) The synthesis of the bottom BM on the substrate by electron beam evaporation with SiO2 and TiO2 layers. (ii) The formation of magnetic films (with different thicknesses in different FP-1D-MPCs) on the bottom BM by RIBS and annealing. (iii) The synthesis of the top BM on crystallized bilayer by the technique used for the growth of the first BM.

4.2.3

Fabrication of Tamm 1D-MPC with BiIG Bilayer

Application of BiIG bilayer allows us to increase the MO effect of Tamm one-dimensional magnetophotonic crystals (T-1D-MPCs) [33, 39]. New original T-1D-MPCs with single layer and bilayer were modeled and studied. Proposed T-1D-MPCs consist of a seven-pair dielectric BM on the surface of which BiIG single layer or bilayer, SiO2 buffer layer, and Au layer are placed successively. The top Au layer has the gradient thickness. Formulas of structures are listed below: ● ●

T-1D-MPC-1: GGG/[TiO2 /SiO2 ]7 /M1/SiO2 /Au. T-1D-MPC-2: GGG/[TiO2 /SiO2 ]7 /M1/M3/SiO2 /Au.

4.2 Fabrication Methods

Au

Au

MO layer M3

MO layer M1

(a)

MO layer M1

SiO2

SiO2

TiO2

TiO2

Substrate

Substrate

(b)

Figure 4.7 Schematic illustration of synthesized T-1D-MPCs with BiIG single-layer (a) and bi-layer (b).

Schematic diagrams of configurations are shown in Figure 4.7. The buffer layer of SiO2 was used to reduce the absorption of light at the magnetic layer–metal interface. The parameters of the structures were previously calculated. The following thicknesses were used in modeling: ● ● ●

hTiO2 = 73 nm and hSiO2 = 115 nm for BM layers. hM1 = 108 nm and hbSiO2 = 140 nm for M1 and SiO2 buffer layers of T-1D-MPC-1. hM1 = 55 nm, hM3 = 178 nm and hbSiO2 = 80 nm for M1, M3, and SiO2 buffer layers of T-1D-MPC-2.

The structures with single layer and bilayer were modeled to form a Tamm plasmon polariton (TPP) mode at the center of photonic band gap (PBG) 𝜆0 = 655 nm. The values of layers component used in simulation are listed in Table 4.3. Synthesis of the structures occurred in several stages. Dielectric BM were synthesized by electron-beam evaporation. The thickness of layers was optically controlled during deposition. Iron garnet layers were fabricated by RIBS of corresponding ceramic targets in argon–oxygen mixture and crystallized in the annealing process at the air. Detailed description of the methods and conditions of garnet synthesis can be found hereinabove or in [13–16]. Au film with a gradient of thickness hAu from 0 to 70 nm along the chosen direction on the sample surface was deposited by thermal evaporation in vacuum using a technique described in [43]. The length of sample along the gradient of Au thickness was 12 mm. Therefore, the gradient of Au thickness was 5.8 nm/mm. Investigation of transmittance was carried out using an automated spectrophotometer KFK-3. Measurements of FR were performed using handmade computer-controlled spectropolarimeter by compensation method in saturation fields. The beam aperture and the gradient of Au thickness at the scale of beam aperture were, respectively, 0.1 and 0.6 nm.

137

Table 4.3

Permittivity tensor components for layers of model 1D-MPCs.

Wavelength 𝝀 (nm)

600 650 700 750

M1𝜺1xx , 𝜺1xy

M2𝜺2xx , 𝜺2xy

7.039 + 0.121 • i,

8.35 + 0.177 • i

−0.014 + 2.656 • 10−3 • i

−0.072 + 0.019 • i

6.591 + 0.066 • i,

7.863 + 0.063 • i

−9.187 • 10−3 + 2.702 • 10−3 • i

−0.037 + 0.01 • i

6.277 + 0.038 • i,

7.518 + 0.044 • i

−7.37 • 10−3 + 2.813 • 10−3 • i

−0.024 + 7.801 • 10−3 • i

6.06 + 0.021 • i,

7.279 + 0.041 • i

−6.619 • 10−3 + 2.95 • 10−3 • i

−0.018 + 7.886 • 10−3 • i

M3𝜺3xx , 𝜺3xy

SiO2 and KU-1𝜺SiO2

TiO2 𝜺TiO2

2.125

5.409

−9.34 + 1.47 • i

2.12

5.282

−12.864 + 1.198 • i

2.117

5.182

−16.519 + 1.113 • i

2.114

5.102

−20.22 + 1.122 • i

Au𝜺Au

−0.050 + 5.305 • 10−4 • i −0.027 + 1.081 • 10−4 • i −0.018 + 5.253 • 10−5 • i −0.015 + 4.191 • 10−5 • i

4.3 Properties of the Structures

4.3 Properties of the Structures 4.3.1

Magneto-optical Properties of FP-1D-MPCs

Before synthesis of structures, to define the effectiveness of FP-1D-MPCs based on a BiIG bilayer, the modeling of different configurations has been carried out. We compared the FP structures with BiIG single layer or bilayer schematically illustrated in Figure 4.8(a, b): ● ● ●

M1 with optical thickness of lM = 𝜆0 /2 (MPC1). M1/M2 of lM = 𝜆0 /2 (MPC2). M1/M2 of lM = 𝜆0 (MPC3).

Simulation results are presented in Figure 4.8c–e. In the figure, amplification factor of Faraday effect of MO layer in Fabry–Perot structure t is also shown. The value t calculated as (4.3)

t = ΘF ∕𝛼F (times)

where 𝛼 F is FR of defect layer of a certain thickness, estimated in our case from specific FR angle spectra of M1 and M2 films. The amplification factor t shows how many times FR of FP-1D-MPC exceeds FR of a BiIG bilayer. The light propagation in proposed FP-1D-MPCs was modeled by computational solution of Maxwell’s equations [44, 45]. It was considered that transverse magnetic (TM) polarized light falls perpendicularly to the structure surface. We used the antisymmetric permittivity tensor with a complex off-diagonal component for MO layers and the tensors with a diagonal component for BM layers [32]. The spectral dependencies of the permittivity components were taken into account. The components of permittivity tensor were calculated using the refractive indexes and extinction coefficients evaluated from the measured transmittance spectra of single BM [TiO2 /SiO2 ]m , M1 and M2 films [28, 32]. To calculate non-diagonal elements, the

15

(c)

(d)

TiO2

Substrate

ΘF, (arbitrary units)

SiO2

1

(e)

80 60

10

40

5

20 0

0

100

t

20

MO layer M2 or M3 MO layer M1

TiO2

15

40 5

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 m m

SiO2

20

80 60

10

0

MO layer M1

100

t

Q (°) ; F (%)

20

Q (°) ; F (%)

(b)

(a)

1 2 3 4 5 6 7 8 9 10 m

0

(f)

0

Substrate

–1

2

1 –2

–1

0 1 H (kOe)

2–2 –1

0 1 H (kOe)

2 –2 –1

3 0 1 H (kOe)

2

Figure 4.8 Schematic representation of FP-1D-MPCs on the basis of single layer (a) and bilayer (b) BiIG with m = 5. MO quality factor Q, figure of merit F, and FR amplification factor t of model structures MPC1 (c), MPC2 (d), and MPC3 (e) as a function of repetition number m. (f) FHLs of synthesized FP-1D-MPCs [TiO2 /SiO2 ]5 /M/[SiO2 /TiO2 ]5 with layers M1 (1), M1/M2 (2), and M1/M3 (3). Source: Berzhansky et al. [31].

139

140

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

estimated optical parameters and experimental spectra of specific FR and MCD were used. The components are listed in Table 4.3. The design resonant wavelength is 𝜆R = 690 nm, that is, PBG center 𝜆0 , 𝜆0 = 𝜆R . The presence of substrate KU-1 with thickness of 0.5 mm was taken into account. The layer thicknesses in FP-1D-MPCs were determined as hTiO2 = 𝜆R ∕4nTiO2 , hSiO2 = 𝜆R ∕4nSiO2 , nM1 hM1 = 𝜆R ∕2, hM1∕M2 = nM1 hM1 + nM2 hM2 = 𝜆R ∕4 + 𝜆R ∕4 = 𝜆R ∕2 or hM1∕M2 = nM1 hM1 + nM2 hM2 = 𝜆R ∕4 + 3𝜆R ∕4 = 𝜆R . In Figure 4.8c–e, the dependencies of modeled Q and F versus m have analogue i correspond to optimal repetition numtype. The maximum values of Qimax and Fmax i bers of mopt (i is the index of FP-1D-MPC: i = 1 for MPC1, i = 2 for MPC2, i = 3 3 1 2 for MPC3) and satisfy to relations Q1max < Q2max < Q3max and Fmax < Fmax < Fmax . Increasing of optical thickness of MO layer from 𝜆0 /2 to 𝜆0 results in decreasing of miopt without of reduction of values of Qmax or F max . It was also shown [18] that efficiency of implementation of BiIG bilayer depends on MO figure of merit of MO layers f = QM1 /QM2 (or QM1 /QM3 ). Q and F of FP-1D-MPCs with BiIG bilayer are reduced in the case of f < 1, and using bilayer is not profitable. Q and F of configurations with m ≤ mopt are low as the substrate was taken into account. The values of Q and F for FP-1D-MPCs with m ≥ mopt are not influenced by the presence of substrate. To compare the properties of FP-1D-MPCs with different MO layers, the three structures were fabricated with m = 5 at different resonant wavelengths 𝜆R . The resonant wavelengths for FP-1D-MPCs with bilayer M1/M2 and M1/M3 are 680 and 635 nm, respectively. FP-1D-MPC only with M1 layer was made for comparison with 𝜆R = 655 nm. FP-1D-MPC 𝜆R are defined by the total thickness of MO layers: hM1 = 150 nm, hM1/M2 = 305 nm, and hM1/M3 = 260 nm. The thicknesses of dielectric layers are hTiO2 = 71 nm and hSiO2 = 103 nm for all MPCs. All layers of FP-1D-MPCs were synthesized by RIBS of corresponding targets in argon–oxygen mixture on GGG substrates. FHLs are shown in Figure 4.8f. FHLs indicate that FP-1D-MPCs of various types of the magnetic anisotropy can be implemented depending on the composition of the magnetic layer. Synthesized FP-1D-MPC with an M1 layer, characterized by an “easy-axis” anisotropy, has the maximum squareness ratio K S = 0.83 and low saturation field H S = 0.5 kOe. FP-1D-MPC with a M1/M2 layer has “easy-plane” anisotropy, K S = 0.19, H S = 1.6 kOe. FP-1D-MPC with a M1/M3 layer has intermediate values of magnetic characteristics, K S = 0.41, H S = 0.7 kOe. These magnetic properties of FP-1D-MPCs allow the use of such crystals in different devices. The best MO quality demonstrates the structure with M1/M2 layer. The characteristics of FP-1D-MPCs with M1 and M1/M2 layers are Q = 5.3∘ , F = 2.6%, t = 21 and Q = 11∘ , F = 4.3%, t = 13, respectively. The characteristics of FP-1D-MPCs with MO bilayer can be improved using the configurations with MO layer of optical thickness of 𝜆0 /2 < lM < 𝜆0 or 𝜆0 < lM < 3𝜆0 /2

4.3 Properties of the Structures

and with two resonances inside PBG at 𝜆R1 and 𝜆R2 (𝜆0 ≠ 𝜆R ). We optimized of FP-1D-MPCs with bilayer of composition M1/M2 [32]. The structure design was performed by changing the number of layer pairs in BM m and the optical thickness of BiIG bilayer lM to achieve high values of MO characteristics. We used the following parameters of structure during optimization: (i) The fixed wavelength of PBG center 𝜆0 = 690 nm. (ii) The fixed thicknesses of nonmagnetic layers hSiO2 = 116 nm and hTiO2 = 75 nm, which are close to the optical thickness l = 𝜆0 /4. (iii) The thicknesses of BiIG bilayer hM in the range from 97 to 383 nm that corresponds to the optical thickness lM ranging from (0.74⋅𝜆0 /2) to (3⋅𝜆0 /2). The defect thickness was increased by changing the thickness of M2 layer hM2 from 29 nm (0.24⋅𝜆0 /2) to 315 nm (2.5⋅𝜆0 /2). The thickness of M1 layer hM1 was fixed and chosen such that in the experiment, quality deterioration of the garnet bilayer on SiO2 did not occur (lM1 = 𝜆0 /4, hM1 = 68 nm) [27, 31]. The results of calculation are shown in Figure 4.9. From the figures, we notice that for the structures with the optical thickness of magnetic layer in the vicinity of (𝜆0 /2), the largest FR enhancement and specific FR are observed for the first I peak. For the structures with the optical thickness of magnetic layer in the vicinity of (2.5⋅𝜆0 /2), the second II and third III peaks coexist; and high values of its Q and F can be achieved at the same time. Optimum numbers m, at which FP-1D-MPC have high Q and F at resonance wavelength for investigated thickness range, are mopt = 4 for the first I and second II resonance peaks and mopt = 3 for the third III peak. Therefore, the structures with maximum values of F and Q for resonance peaks I and II and the optimum number mopt = 4 and the structures with high t, non-zero value of F, and m = 7 were experimentally implemented. FP-1D-MPCs with bilayer M1/M2 and m = 4 and m = 7 on the quartz substrate were fabricated. Optical thicknesses of the bilayer were (0.8⋅𝜆0 /2), (1.2⋅𝜆0 /2), (1.8⋅𝜆0 /2), and (2.5⋅𝜆0 /2) for m = 4 and (0.8⋅𝜆0 /2), (1.2⋅𝜆0 /2), (1.6⋅𝜆0 /2), and (2.5⋅𝜆0 /2) for m = 7 (𝜆0 = 690 nm). Transmittance and FR spectra of synthesized FP-1D-MPCs with M1/M2 bilayer are presented in Figure 4.10. Measurements were carried out at room temperature by a spectral device based on a KSVU-6 system with a double-diffraction MDR-6 monochromator [32]. The magnetic field applied during FR measurements was significantly higher than a saturation field of FP-1D-MPC with bilayer of composition used (H s = 1.6 kOe). Evidently, as the defect thickness increases from (0.8⋅𝜆0 /2) to (3⋅𝜆0 /2), the three resonance peaks corresponding to the resonance conditions (𝜆R /2), first I peak; (𝜆R ), second II peak; and (3⋅𝜆R /2); third III peak pass the PBG. The following maximum values of MO quality factor 15.1∘ for 624 nm, specific FR −113 deg/μm (that exceeds in 62 times the FR of bilayer film) for 654 nm, and absolute FR −20.6∘ for 626 nm were achieved for three different configurations of the structure with m = 4, lM = (2.5⋅𝜆0 /2); m = 7, lM = (0.8⋅𝜆0 /2); and m = 4, lM = (2.5⋅𝜆0 /2), respectively. According to calculations of electric field intensity inside FP-1D-MPCs with m = 4, lM = (0.8⋅𝜆0 /2) at resonance wavelengths, the asymmetry of distribution

141

-15 -10 -5 0 80

(c)

60

20 0

3

7 4 3 1

I peak

7 5 4 3 1

7 5 4 3 1

5 4 3 1 III peak

7 5 4 3 1

15 12 (d) 9 5 6 6 3 7 0 21 18 (e) 15 12 5 6 9 7 6 0.6 0.8

I peak 4 3 2 1

I peak 4

3 2 1

1.0

1.2

1.4

1.6

1.8

lM *(λ0/2)

1.5 1.8 2.1 2.4 2.7 3.0 lM *(λ0/2)

15 II peak 4 12 (f) 9 5 6 6 3 7 0 21 II peak 4 18 (g) 5 15 6 12 9 7 6 1.4 1.6 1.8 2.0 2.2 2.4

3 2 1

3 2 1 2.6

2.8

lM *(λ0/2)

Q (°)

F (%)

0.6 0.9 1.2

II peak

4 5 7 III peak 7

Q (°)

5 7 II peak 5

7 5 4 3 1

40

3 4

F (%)

T (%) ΘF (°)

-20

1.5 1.8 2.1 2.4 2.7 3.0 II peak III peak 1 1

Q (°)

0.6 0.9 1.2 I peak (a) 80 1 3 60 4 40 5 20 7 0 -30 (b) I peak -20

100

F (%)

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

t

142

15 3 III peak 12 (h) 2 9 4 6 1 5 3 6 7 0 21 III peak 3 2 18 (i) 4 1 15 5 12 6 9 7 6 2.2 2.4 2.6 2.8 3.0 3.2 Optical thickness of MA defect layer lM *(λ0/2)

Figure 4.9 Calculated transmittance T (a), FR angle ΘF (b), FR enhancement factor t (c), MO figure of merit F (d, g, i), and MO quality factor Q (e, f, h) of FP-1D-MPCs with M1/M2 bilayer on fused quartz substrate as a function of the repetition numbers m = 1, 3, 4, 5, 7 and the optical thickness of bilayer lM for three resonance peaks corresponding to conditions (𝜆R /2), first I peak; (𝜆R ), second II peak; and (3𝜆R /2), third III peak. The arrows indicate the positions at which the peaks pass the PBG center 𝜆0 . Source: Mikhailova et al. [32].

occurs (Figure 4.11). The effects of light localization in the structure reduce the influence of magnetic layer absorption at resonance. The most influence on the properties of structures is provided by “SiO2 –M1” and “M2–SiO2 ” interfaces. The obtained values of MO quality factor Q and figure of merit F of proposed FP-1D-MPCs exceed approximately in two times the same characteristics of known FP-1D-MPCs based on non-garnet BM in the wavelength range from 550 to 850 nm [39]. New FP-1D-MPCs provide not only high FR angles but also essentially lower MO quality factor compared to those of all-garnet structures [40, 41]. MO efficiency of FP-1D-MPCs can be modified by the utilization of gold coatings, gratings, or embedded nanoparticles, leading to interplay of optical resonances of different nature [39, 47, 48].

4.3 Properties of the Structures -20

100 m=4

(a) 50

-10

0 100

2.5

0 -10 ΘF (°)

0 100

-10

1.2

50

1.8

0

0 100

IM*(l0/2)

IM*(l0/2)

1.8

50 T (%)

m=4

(b)

2.5

1.2

0

50

-10

0.8

0.8 0

0 400

500

600 700 𝜆 (nm)

800

500

600 700 𝜆 (nm)

800

100 50

-20

m=7

(c)

2.5

0 100

m=7 2.5

0 -20

50

1.8

50

1.2

0 100

-10 ΘF (°)

0 100

1.8

0 -20 -10

IM*(l0/2)

IM*(l0/2)

T (%)

(d)

-10

1.2

0 -20

50

0.8

0 400

-10

0.8

0 500

600

𝜆 (nm)

700

800

500

600

𝜆 (nm)

700

800

Figure 4.10 Transmittance (a, c) and FR (b, d) spectra of synthesized FP-1D-MPCs on KU-1 with M1/M2 bilayer. Source: Mikhailova et al. [32].

4.3.2

Magneto-optical Properties of T-1D-MPCs with BiIG Bilayer

A characteristic feature of the spectra of T-1D-MPC-1 and T-1D-MPC-2 presented on Figure 4.7 is the presence of a pronounced resonant peak (Figure 4.12), the position of which depends on the thickness of Au layer. In the calculated spectra with increasing of Au thickness hAu from 6.1 to 65.2 nm, the blue shift of TPP mode takes place. Nevertheless, the red and blue shifts of resonance occur in the experiment at 6.1 nm < hAu < 30 nm and 30 nm < hAu < 65.2 nm, respectively. The red shift of the mode results from the changes of granularity, surface roughness, and continuity of Au coating (structural changes) with increase of its thickness, which leads to a change of optical properties of coating. Blue shift of TPP modes is caused by the change in the coating thickness itself, since in the simulation we assume smooth boundaries, continuity, and non-gradient thickness of Au layer. The surface roughness of layers, discontinuity, and thickness gradient of Au coating in the beam aperture in the experiments lead to increasing of half-width of the resonant peaks. Experimental and calculated spectra of the T-1D-MPC-2 without Au coating are identical and have the peak at 675 nm. The maximum resonant FR value were achieved for T-1D-MPC-2 −12.3∘ at 645 nm for configuration with hAu = 65.2 nm. Calculated spatial distribution of electric field intensity in T-1D-MPC-2 with Au coating of different thickness, T-1D-MPC-2 without Au coating, BM with magnetic

143

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

Electric field intensity (relative units) 0 2 4 6 8 10 12 14 16 18 20 1600

TiO2 SiO2

1400 Structure depth (nm)

144

1200 1000

TiO2 SiO2 TiO2 SiO2 TiO2

600

SiO2 M1 M2 SiO2 TiO2

400

SiO2 TiO2

800

SiO2

200

TiO2 SiO2 TiO2

Figure 4.11 Calculated distribution of electric field intensity inside FP-1D-MPC KU-1/[TiO2 /SiO2 ]m /M1/M2/[SiO2 /TiO2 ]m with m = 4 and lM = (0.8⋅𝜆0 /2) at the first I resonance peak. Source: Berzhansky et al. [46].

layers, and single BM is shown in Figure 4.12i. The distribution is classical for TPP mode. Distribution in magnetic bilayer is characterized by asymmetry as in the case of FP-1D-MPCs. The maximum intensity is observed in the layers adjacent to the plasmonic coating and for configuration with thickness of Au layer of 28.3 nm. This is in good agreement with the experimental results and indicates that a coating thickness is optimal for such type of structure. Consequently, the resonances on TPP have the maximum optical quality factor and transmittance at the vicinity of Au thickness of 30 nm for both structures T-1D-MPC-1 and T-1D-MPC-2. MO quality factor Q of T-1D-MPC-1 and T-1D-MPC-2 increases from 0∘ to 0.96∘ and from 0.55∘ to 5.01∘ as the thickness of Au layer increases from 0 to 65.2 nm. Obtained values of Q are higher than Q of the first magnetophotonic Tamm structure [SiO2 /BiIG]5 /Au (Q = 0.58∘ ) [39]. So the optimal synthesis conditions of ultrathin BiIG garnet films, deposited by RIBS, were observed. It was shown that temperature of crystallization annealing and the regimes of substrate pretreatment had substantial effects on the MO properties of films. To effective growth of garnet phases with high bismuth content, two-step synthesis technology has been proposed. The sub-layer with a lower bismuth content (less than 1 at./f.u.) was deposited on SiO2 layer, silica, or quartz substrate. Then, after crystallization annealing of sub-layer, the main magneto-active layer with a higher Bi content (more than 1.5 at./f.u.) was deposited and crystallized by annealing. Due to the proposed method of crystallization, new record values of the FR angle of the Fabry–Perot and Tamm 1D-MPCs in optical wavelength range were achieved. MO effect of synthesized structures exceeds of the same values of known analogue.

4.3 Properties of the Structures 100

hAu = 0 nm

hAu = 65.2 nm

hAu = 8.6 nm

0

Bragg mirrors

–0.5

ΘF (°)

TPP mode

2.0

60 T (%)

1.5

–1.0

0.0

600

650 700 λ (nm)

12.5 times

400

800

1200

1600

0

750

8 4

600

700

800

hAu = 65.2 nm

hAu = 6.1 nm

T (%)

3

0

λ (nm)

8 4

hAu = 65.2 nm hAu = 28.3 nm

–1.0

hAu = 58.6 nm

2

hAu = 6.1 nm

1 0

700

–0.5

TPP mode

4

20

680

Bragg mirrors

5

60

660

and hSiO2 = 0 nm

ΘF (°)

80

640

0.0

hAu = 0 nm

hAu = 8.6 nm

620

(b) hAu = 0 nm

hAu = 28.3 nm

40

600

900

λ (nm)

(a) 100

T (%)

1600

–2.0

500

hAu = 0 nm

–1.5 600

650 700 λ (nm)

15 times

hAu = 0 nm

750

and hSiO2 = 0 nm

–2.0

0 500

600

700

800

600

900

100

hAu = 65.2 nm

hAu = 0 nm

hAu = 28.3 nm

hAu = 0 nm

hAu = 6.1 nm

1.5

650 700 λ (nm)

750

700 λ (nm) hAu = 65.2 nm

0

hAu = 8.6 nm

5

TPP mode

–4

hAu = 65.2 nm hAu = 28.3 nm

–6

hAu = 58.6 nm hAu = 6.1 nm

9 times

2

hAu = 0 nm

–8 600

700

0

650 700 λ (nm)

750

hAu = 0 nm and hSiO2 = 0 nm

–10

0 500

680

660

λ (nm)

–2

3

0

640

hAu = 0 nm

1

20

0

Bragg mirrors

4 T (%)

40

620

(h)

and hSiO2 = 0 nm

hAu = 6.1 nm

60

Air

Structure depth (nm)

600

900

hAu = 0 nm

hAu = 28.3 nm

80

800

ΘF (°)

600

0

4

–12

500

4

hAu = 6.1 nm hAu = 0 nm

–10

0

100

8

8

hAu = 58.6 nm

10 times

(e)

0

0

hAu = 28.3 nm

–6 –8

600

4

4 hAu = 65.2 nm

1.0

0.0

700

8

8

–4

0.5

20

680

0

Bragg mirrors

TPP mode

2.0 T (%)

40

660

λ (nm)

–2

ΘF (°)

2.5

60

640

0

and hSiO2 = 0 nm

hAu = 8.6 nm

80

620

(d)

λ (nm)

(c)

T (%)

hAu = 6.1 nm

–1.5

0

T (%)

1200

hAu = 58.6 nm

0.5

20

800

4

hAu = 28.3 nm

1.0

40

400

(i)

8 hAu = 65.2 nm

Electric field intensity (relative units)

T (%)

and hSiO2 = 0 nm

hAu = 6.1 nm

2.5

0.0

hAu = 0 nm

hAu = 28.3 nm

80

–12 600

700 λ (nm)

800

900

600

620

640

660

680

700

λ (nm)

Figure 4.12 Measured (a, b, e, f) and calculated (c, d, g, h) transmittance and FR spectra of synthesized BM, BM with BiIG layers, and T-1D-MPC-1 (a, b, c, d) and T-1D-MPC-2 (e, f, d, h) as a function of the thickness of Au layer. (i) Calculated spatial distribution of electric field intensity in T-1D-MPC-2 with different Au thickness, T-1D-MPC-2 without Au coating, BM with BiIG layers, and a BM. Source: Mikhailova et al. [33]. ©2018, EDP Sciences.

4.3.3 An increase of the Magneto-optical Response in the Ultrathin Films A study of the MO response in ultrathin films has drawn attention of researchers just recently. In the pioneer work [49] published in 2015, the authors demonstrated the monotonic increase in SFR with the reduction of the iron garnet film thickness. The effect appears when the film thickness is about 100 nm and shows the dramatic growth when it reduces up to 20 nm. As soon as the SFR directly depends on the MO response in the material, the SFR growth in the ultrathin films indicates that the MO response also arises in them. Besides the Faraday effect, one can address another MO effects to explore the dependence of the MO properties of the film on its thickness. For instance, the plasmon-enhanced transverse magneto-optical Kerr

145

146

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

effect (TMOKE) [50] is known to be very sensitive to the properties of the material. So this effect allows to measure and analyze the MO response of even a few nanometer-thick iron garnet films. On the contrary, the Faraday effect is cumulative, and in such extremely thin layers, it turns out to be very small and is hardly detectable. However, MO analysis by means of the TMOKE requires a fabrication of the plasmonic grating on top of the film. Here we review the investigation of the MO properties by means of both techniques, plasmon-enhanced TMOKE and Faraday effect. In Ref. [51], the MO response in ultrathin film was studied by means of the plasmon-enhanced TMOKE in bismuth-substituted iron garnet films of thickness less than 60 nm with a one-dimensional gold grating cover. It was shown that the differences of the TMOKE values appearing when the film thickness is decreased can be explained by a variation of the non-diagonal element of the dielectric tensor and by localization of the surface plasmon polaritons (SPPs) at the metal–dielectric interface. For the experimental studies, Bi0.8 Gd0.2 Lu2 Fe5 O12 films of thicknesses 19, 46, and 60 nm were used. These samples were grown by a liquid-phase epitaxy (LPE) on (100) GGG (Gd3 Ga5 O12 ) substrates. The films exhibit planar magnetic anisotropy and saturation magnetization of 4𝜋M s = 1800 G. All the samples were made from the same original magnetic film by a sequential etching in an ortho-phosphoric acid bath with a slow rotation rate to ensure uniform thickness. Thickness uniformity was verified by probing different points of the sample surface via transmission electron microscopy (TEM). To fabricate the magnetoplasmonic crystal (Figure 4.13), an 80-nm-thick gold layer was deposited on the magnetic films by the magnetron sputtering. A resistive mask was then formed on top via electron-beam lithography, and gold-layer patterning was performed by ion etching in an argon-ion plasma single-frequency discharge. The pattern is a periodic array of slits, and its period was determined by preliminary numerical modeling to provide the extrema of the TMOKE resonances at the same frequencies for the different samples. This allows one to avoid the dielectric dispersion impact and to isolate the influence of the magnetic garnet film thickness on the MO response. The fabricated plasmonic gratings had the following periods and widths of the slits, respectively: d = 324 nm, r = 85 nm for the 60-nm-thick film; d = 322 nm, r = 75 nm for the 46-nm-thick film; and d = 347 nm, r = 60 nm for the 19-nm-thick film. All three samples demonstrate pronounced TMOKE resonances in the same wavelength range. The addressed MO effects were measured over a wide range of wavelengths (visible and near-IR ranges) and incidence angles. The sample was placed in a uniform external magnetic field of 2000 Oe along y-axis in Figure 4.13 generated by the electromagnet. The applied magnetic field exceeded the magnetic field required to saturate the magnetization of the iron garnet films under consideration and guarantees the reproducibility of the results. The light was collimated after it exited the sample and detected with the spectrometer. A 2D charge-coupled device (CCD) camera in the spectrometer was used to observe the spectral decomposition along one axis

4.3 Properties of the Structures

Figure 4.13 The scheme of the plasmonic crystal. Ultrathin LPE-grown bismuth-substituted iron garnet film on GGG substrate and covered by sub-wavelength gold grating is illuminated by obliquely incident p-polarized light.

p-Polarized light

z

y x

Gold grating

Ultrathin iron garnet film

SPP

M

GGG substrate

and the incidence angle decomposition along the perpendicular axis. Therefore, the angular and wavelength resolved transmission spectra of all samples for two opposite directions of the magnetic field were measured. Each measurement with alternating opposite directions of the magnetic field was repeated 200 times, and then these results were averaged. This regime provided reproducibility of the measurements, with a signal-to-noise ratio exceeding three orders of magnitude in the spectral range of our interest. Based on these spectra, one can find 𝛿, the value of the TMOKE, as a relative change of the transmitted light intensity T(M) when the structure is re-magnetized [7]: 𝛿=2

T(M) − T(−M) T(M) + T(−M)

(4.4)

The measured wavelength- and angular-resolved transmission (left column) and TMOKE (right column) spectra of the three samples are given in Figure 4.14. Excitation of SPPs at the [gold]/[ferromagnetic dielectric] interface leads to the dips in the transmission spectra from the calculation of the SPP dispersion based on the phase synchronism condition (dashed lines in Figure 4.14) [52]. The corresponding numerical simulations performed by the rigorous coupled-wave analysis (RCWA) [53, 54] yield the same experimental results with high accuracy. In Figure 4.14a, secondand third-band SPPs are observed. The transverse in-plane magnetic field spectrally shifts the transmission dips either toward lower or higher frequencies depending on the direction of the magnetic field with respect to the normal to the sample surface and SPP wavevector. As a result, near the frequencies of SPP, excitation resonances in the TMOKE spectra are noted with an S-shape with positive and negative maxima where the TMOKE reaches 0.04. The TMOKE spectra is antisymmetric with respect to the normal incidence when TMOKE vanishes due to symmetry reasons. Although the thicknesses of the magnetic garnet films in the plasmonic crystals under consideration are different from each other, the resonance positions in all three cases are almost the same due to the proper choice of the grating periods.

147

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics 950

60-nm-thick film transmission

0.10

60-nm-thick film TMOKE

950

0.040

0.09 900

0.032

900 0.08

0.024 850

0.07 0.06

800

0.05 750

0.04 0.03

700

Wavelength (nm)

Wavelength (nm)

850

0.016 0.008

800

0.000 750

–0.008 –0.016

700

–0.024

0.02 650 600 –5

0.00 0

5

10 15 Angle (°)

20

600

25

–0.032 –5

–0.040 0

5

10 15 Angle (°)

20

25

(b) 46-nm-thick film transmission

46-nm-thick film TMOKE

0.80

900

0.040

900

0.72

850

0.032

0.64

0.024

850

0.56 800 0.48 750

0.40 0.32

700

0.24 650

0.08 5

10 15 Angle (°)

20

0.000

700

–0.008 –0.016 –0.024 –0.032

600 –5

25

–0.040 0

5 10 Angle (°)

15

20

25

(d)

(c) 950

0.008 750

0.00 0

0.016

800

650

0.16

600 –5

Wavelength (nm)

Wavelength (nm)

650

0.01

(a)

19-nm-thick film transmission

0.40

950

19-nm-thick film transmission

0.030

0.36

900

0.024 900

0.32 850

0.018 850

0.28 0.24

800

0.20 750

0.16 0.12

700

Wavelength (nm)

Wavelength (nm)

148

0.012 0.006

800

0.000 750

–0.006 –0.012

700

–0.018

0.08 650

650

0.04 600 –5

(e)

0.00 0

5

10 15 Angle (°)

20

600 –5

25

–0.024 0

5 10 Angle (°)

15

20

25

–0.030

(f)

Figure 4.14 Wavelength and angular resolved transmission (a, c, e) and TMOKE (b, d, f) spectra of the 60-nm-thick (a, b), 46-nm-thick (c, d), and 19-nm-thick bismuth-substituted iron garnet films (e, f). White dashed lines show the SPP dispersion calculated in the simplified model of smooth interfaces. Gold grating periods are 324 nm (60-nm-thick film), 322 nm (46-nm-thick film), and 347 nm (19-nm-thick film). Source: Borovkova et al. [51].

4.3 Properties of the Structures

If the magnetic garnet films are thick enough, then the TMOKE hardly depends on the film thickness (compare the TMOKE spectra for the 60-nm- and 46-nm-thick films in Figure 4.14b,d). However, for thinner films, the TMOKE tends to decrease. Nevertheless, the decrease is not very pronounced so that the TMOKE maximum drops to 0.03 for the 19-nm-thick film (Figure 4.14f). Such behavior in the TMOKE can be due to the changes in the magnetic garnet film permittivity tensor and modifications of the SPP modes caused by the decrease in film thickness. The physical origins of the observed TMOKE were revealed by an electromagnetic modeling based on RCWA. The refractive indices of gold, GGG, and BiIG were taken from [40, 55]. Geometrical parameters of the gold grating were measured by TEM imaging. There remained only one unknown parameter: the off-diagonal components of the magnetic garnet film permittivity tensor, 𝜀1 . Therefore, 𝜀1 were found by matching the calculated and experimentally measured transmission and TMOKE spectra. In Figure 4.15 (dots), the spectral range between 0.7 and 0.9 μm in wavelength is shown in detail where two TMOKE S-shaped resonances are present. These are caused by the SPPs propagating in opposite directions that makes their signs opposite to each other. At 10∘ incidence, the two resonances are relatively close so that a slight overlap between them takes place. For 20∘ incidence, the resonances are sufficiently separated and do not interfere with one another. It should be noted that 𝜀1 found for all three samples have similar dispersion but are different in values. Particularly, at 𝜆 = 0.754 μm, close to one of the TMOKE resonances, 𝜀1 = 0.0084 for the 60-nm-thick film, 𝜀1 = 0.0088 for the 45-nm-thick film, and 𝜀1 = 0.0102 for the 19-nm-thick film. Therefore, TMOKE measurements indicate a slight growth in the MO gyrotropy parameter of the film as the thickness decreases. This agrees with the previous results obtained from Faraday effect measurements [49]. Interestingly, in spite of the growth of 𝜀1 , the TMOKE decreases for thinner films. As TMOKE is related to the plasmonic resonances, this phenomenon can be understood by analyzing the SPP modes in these structures. The spatial distribution of the electromagnetic field at the SPP resonance (Figure 4.16) indicates that the penetration depth of the SPP wave is comparable with the thickness of the magnetic garnet films in the samples. The SPP field amplitude decreases by a factor of e at the depth of about 40 nm. In plasmonic crystals with 60-nm-thick and 46-nm-thick films, SPP field is localized mostly inside the ultrathin magnetic garnet film. However, in the 19-nm-thick film, a significant part of the SPP field penetrates into the nonmagnetic GGG substrate. This diminishes the influence of the magnetic field on the SPP in the 19-nm-thick film sample, the SPP frequency is shifted by a smaller value, and, consequently, the TMOKE decreases even though 𝜀1 is a bit larger. If 𝜀1 is assumed to be independent of the film thickness, then the TMOKE decrease in the thinner films is more pronounced from the calculated curve in Figure 4.16d. Further decrease in film thickness makes the TMOKE smaller, but it has relatively large values even for the nanometer-thick films. For example, for a 2-nm-thick film sample, the TMOKE is 1.8 × 10−3 that can be easily detectable. It should be noted that the Faraday effect for such films is very small (2.7 × 10−6 deg for the 2-nm-thick film)

149

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

0.06

60-nm-thick film, incident angle 20°

0 –0.02 –0.04 0.65

Experiment Model

0.02 0 –0.02

0.7

0.75 0.8 0.85 Wavelength (µm)

–0.04 0.65

0.9

(a)

0.7

0.75 0.8 0.85 Wavelength (µm)

0.9

(b)

0.04

46-nm-thick film, incident angle 20°

46-nm-thick film, incident angle 10° 0.04

Experiment Model

0.03 0.02

0.02

0.01

0.01

0 –0.01

0 –0.01

–0.02

–0.02

–0.03

–0.03

–0.04 0.65

0.7

0.75

0.8

0.85

0.9

Experiment Model

0.03

TMOKE

TMOKE

60-nm-thick film, incident angle 10°

0.04 TMOKE

0.02

TMOKE

0.06

Experiment Model

0.04

–0.04 0.65

0.95

0.7

Wavelength (µm)

0.75

0.8

0.85

0.9

0.95

Wavelength (µm)

(c)

(d) 0.03

19-nm-thick film, incident angle 20°

0.03

Experiment Model

0.02

19-nm-thick film, incident angle 10° Experiment Model

0.02 0.01 TMOKE

0.01

TMOKE

150

0

–0.01

–0.01 –0.02 0.65

0

–0.02 0.7

0.75

0.8

0.85

0.9

0.95

–0.03 0.65

(e)

0.7

0.75

0.8

0.85

0.9

0.95

Wavelength (µm)

Wavelength (µm)

(f)

Figure 4.15 Theoretical (solid curves) and experimental (dots) TMOKE spectra for 60-nm-thick films (a, b), 46-nm-thick films (c, d), and 19-nm-thick films (e, f) for 20∘ and 10∘ incident angles. Gold grating periods are 324 nm (60-nm-thick film), 322 nm (46-nm-thick film), and 347 nm (19-nm-thick film). Source: Borovkova et al. [51].

and hardly measurable. However, for the ultrathin films, more than 10-nm-thick Faraday effect provides high-precision tool along with the TMOKE. So far, an increase of the MO response with the decrease of the iron garnet film thickness has been considered in assumption of the constant distribution of the MO effect inside the ultrathin film. To refine the model and find out the most optimal distribution of the MO response inside the ultrathin film, two hypotheses have been proposed. The proposed models were verified by means of the compositional

4.3 Properties of the Structures 60-nm-thick film, field intensity

0

46-nm-thick, field intensity

9

9

0

8

8

7

0.05

7

0.05

5

0.1

4

6 z (μm)

z (μm)

6

5

0.1

4

3 0.15

3 0.15

2

2

1 0.2

0

0.1

0.2 x (μm)

0.3

1 0.2

0

(a) 0

0.2 x (μm)

0.3

0

0.035

9 8

0.03

6 5

0.1

4 3

0.15

max(TMOKE)

7

0.05 z (μm)

0.1

(b) 19-nm-thick film, field intensity

0.2

0.025 0.02 0.015 0.01

2 0.005

1

(c)

0

0

0.1

0.2 x (μm)

0.3

0

0

0

10°. incidence λ = 0.754 μm 10 20 30 50 60 40 Magnetic garnet filim thickness (nm)

(d)

Figure 4.16 The SPP wave field distribution in the plasmonic crystal with (a) 60-nm-thick, (b) 46-nm-thick, and (c) 19-nm-thick bismuth-substituted iron garnet film. White dashed lines mark the contours of the gold grating and the ferromagnetic film. Gold grating periods are 324 nm (60-nm-thick film), 322 nm (46-nm-thick film), and 347 nm (19-nm-thick film). Incidence angle is 20∘ . (d) Calculated dependence of the TMOKE maximum value at 𝜆 = 0.754 μm on the magnetic film thickness. Source: Borovkova et al. [51].

analysis and experimental measurements of Faraday effect and compared with the simplest model of constant gyrotropy in the film. The addressed theoretical models are given in Figure 4.17. In Model 1 (green line), the MO response of the ultrathin film is supposed to be constant as it was assumed above. Model 2 (blue line) takes into account that the dielectric properties do not change instantly at the [Bi-substituted iron garnet]/[GGG substrate] interface. Actually, the compositional analysis (see Figure 4.18) reveals that there is a thin (about 30-nm-thick) transitional layer, where the distribution of the elements typical for the iron garnet changes gradually to the substrate composition. Model 3 (red line) contains two peculiarities, namely, the gradual decrease of the dielectric properties at the [Bi-substituted iron garnet]/[GGG substrate] interface and the sharp increase of the gyrotropy at [air]/[Bi-substituted iron garnet] interface. There were fabricated two sets of the samples. These films with Bi0.8 Gd0.2 Lu2 Fe5 O12 and Bi0.7 Gd0.4 Lu1.9 Fe4.1 Ga0.9 O12 per formula unit (pfu) were grown by LPE on GGG (100)-oriented substrates. The films for any set of measurements were all taken from the same wafer and wet-etch-thinned-down sequentially in ortho-phosphoric acid, thus avoiding possible composition differences due to

151

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

g BiREIG

15 nm

hˈ/2

hˈ/2

15 nm

Air

Model 1

GGG

g=const

Model 3

Model 2

0

h (nm)

Figure 4.17 Schematic depiction of the gyrotropy parameter g (vertical axis) distribution across the film (horizontal axis) in the three models. BiREIG stands for Bi-rare-earth-substituted iron garnet, and GGG stands for gadolinium–gallium garnet substrate. Source: Levy et al. [56]. ©2019, The Optical Society.

5.0

Ga Substrate

Film

(pfu)

(a.u.)

152

Bi

(a)

0

20 40 60 Nanometers

Film LU

1.0 0.0

‒40 ‒20

Fe

Ga 4.0 Substrate 3.0 Gd 2.0

80

Bi ‒100

0

100 Nanometers

200

300

(b)

Figure 4.18 (a) Bi and Ga concentrations in Bi0.8 Gd0.2 Lu2.0 Fe4.5 O12 (Sample 1), arbitrary units, determined by SIMS on a 60-nm-thick film. Bi concentration is shown multiplied by factor 10 for the sake of clarity. (b) Elemental concentrations (pfu) for Bi, Lu, Gd, Ga, and Fe in Bi0.7 Gd0.4 Lu1.9 Fe4.1 Ga0.9 O12 (Samples 2 and 3) determined by cross-sectional S-TEM EDX. Both show uniform concentrations as a function of position above (film) and below (substrate) the transient layer. The zero of horizontal coordinates is set at the film–substrate interface. Source: Levy et al. [56]. ©2019, The Optical Society.

growth conditions. The bulk magneto-optic gyrotropy parameter g, obtained from measurements in 2-μm-thick films, is different for the two different composition. For Bi0.7 Gd0.4 Lu1.9 Fe4.1 Ga0.9 O12 pfu films (Samples 2 and 3), the bulk is g = 0.029 + 0.003i, and for Bi0.8 Gd0.2 Lu2 Fe5 O12 (Sample 1), it is g = 0.038 + 0.004i. Compositional secondary-ion mass spectroscopy (SIMS) measurements were performed with O2 + bombardment and positive ion detection mode using an IMS-7f (CAMECA) microanalyzer. The primary beam was rastered over an area of 125 × 125 μm2 , and the secondary ions were collected from the central part of this area (diameter of 33 μm). In order to avoid charging effect, the sample was coated with a layer of gold (50 nm), and electron flooding was employed using a normal incidence e-gun. Electron micrographs and energy dispersive X-ray (EDX) spectroscopy maps were obtained on Bi0.7 Gd0.4 Lu1.9 Fe4.1 Ga0.9 O12 (Samples 2 and 3) using an FEI Titan Themis aberration-corrected S-TEM operated at 200 kV. The point resolution in this aberration-corrected mode is 0.08 nm. The microscope is fitted with a Super-XTM

4.3 Properties of the Structures

X-ray detector, which is a combination of four detectors for fast X-ray mapping in S-TEM mode. For the present experiment, 1-nm resolution EDX maps were taken with an average beam current of 100 pA. The size of the maps was 512 × 512 pixels, and 50 μs/pixel dwell time was used for collecting the signal. All maps are generated by summing over 10 frames. Drift correction during data collection and subsequent analysis were performed using Velox software. These results are shown in Figure 4.18b, with a standard deviation of ±0.05 pfu for Bi and ±0.11 pfu for Fe. In Figure 4.18a, one can see how the concentrations of Bi and Gd depends on the depth of the nanostructure. These elements have been chosen as soon as they are key elements and allow to distinguish the magnetic layer and nonmagnetic substrate. From left to the right, one can observe how the concentration of Gd decreases to 0 and at the same time the concentration of Bi grows until the saturation value. The layer where the atoms of Gd disappear and atoms of Bi appear can be considered as a transitional layer. Its thickness is about 30 nm. Based on these data, Model 2 (Figure 4.17) has been proposed. The Faraday effect has been measured experimentally in the addressed ultrathin films. Light from a continuous-wave (CW) laser source was used to probe the FR at 532 nm. Measurements normal to the film surface were conducted in a rotating-polarizer configuration. Faraday-rotation hysteresis loops were recorded for each sample, and the FR at saturation was chosen to characterize the response of each film. The paramagnetic signal of the GGG substrate was subtracted out from the overall FR signal. The experimental setup is described in [49]. Film thickness was measured via ellipsometry, and FR per unit length is plotted in Figure 4.19 for both sets of samples as a function of thickness. For the ferrimagnetic films thinner than 50 nm, the SFR grows dramatically with a decrease in film thickness. One can see that the SFR reaches 10 ∘ μm−1 in the 20-nm-thick film, almost 2.5 times greater than SFR in the same thick film. For the thin films we observe oscillations due to interference of the emerging waves because of multiple reflections inside the thin film. The obtained experimental results presented in Figure 4.19 have been analyzed by the theoretical model based on classical electrodynamics (see also [56]). Linearly polarized light impinges on the magnetic film at normal incidence. The magnetization is directed along the propagation direction, perpendicular to the film surface. Refractive indices and film thickness are as shown in the inset of Figure 4.19. The magneto-optic gyration and the refractive index in the iron garnet film are initially assumed to be uniform inside the film and parameterized by the gyrotropy parameter g. Multiple reflections are included in the analysis, giving rise to an oscillatory behavior in FR as a function of film thickness h. The FR angle 𝜃 is given by [7] ( ) √ √ √ ⎡ ⎤ n3 ⎢ (1 + n3 ) cos(k0 𝜀2 − gh) − i √𝜀2 −g + 𝜀2 − g sin(k0 𝜀2 − gh) ⎥ 1 ⎥ arg ⎢ ( ) ⎢ ⎥ 2 √ √ √ n3 ⎢ (1 + n3 ) cos(k0 𝜀2 − gh) − i √𝜀 −g + 𝜀2 − g sin(k0 𝜀2 − gh) ⎥ 2 ⎣ ⎦ (4.5)

153

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics 14

11 Air n1=1

12 11

n2

10

n3 Substrate

Sample 1 Sample 2 Sample 3

532 nm

M

10 9 8

h

7

9

6

8

5

7

4

6

3

5

2

4

1

3

0

0.05

0.1

0.15 0.2 0.25 0.3 Magnetic film thickness (μm)

0.35

0.4

Specific Faraday rotation (°/μm)

Input light

13

Specific Faraday rotation (°/μm)

154

0 0.45

Figure 4.19 Specific Faraday rotation at 532 nm versus film thickness for Bi0.8 Gd0.2 Lu2 Fe5 O12 (Sample 1) and Bi0.7 Gd0.4 Lu1.9 Fe4.1 Ga0.9 O12 (Samples 2 and 3), showing a dramatic increase below 50 nm. Inset: Schematic depiction of the samples under consideration. Source: Levy et al. [56]. ©2019, The Optical Society.

where n3 is the refractive index of the substrate. The refractive index of the ferrimag√ netic film can be written as 𝜀2 ± g for positive and negative polarization helicities. k0 = 2π/λ is the wave number in vacuum, and 𝜆 is the wavelength. Equation (4.5) has two asymptotes, for h ≪ 2π/λ ultrathin films and for h ≫ 2π/λ bulk iron garnet material, with SFRs given by { g 2𝜋 − 2n 𝜆 , if h ≫ 2𝜋 𝜃 𝜆 2 (4.6) SFR = = g 2𝜋 − 1+n 2𝜋 , if h ≪ h 𝜆 𝜆 3

From Eq. (4.6), the ratio of the SFR at small h with respect to the SFR at large h is, thus, SFRh→0 2n2 = SFRh→∞ 1 + n3 The refractive indices at wavelength 𝜆 = 532 nm are given by n3 = 1.980 + 0.003i in the GGG substrate and n2 = 2.610 + 0.056i in the Bi-substituted iron garnet films for all samples. Therefore, at 𝜆 = 532 nm this ratio is predicted to be about 1.75, but the experimentally measured ratios are always larger than 2. In particular, SFRh → 0 /SFRh → ∞ = 10.25/5 ≈ 2.05 for Sample 1, and SFRh → 0 /SFRh → ∞ = 9.818/4.226 ≈ 2.32 for Samples 2 and 3. Thus, the analytical expression derived from classical electrodynamics cannot account for all the effects operating in this system. Besides the preceding analysis we also performed an electromagnetic modeling based on an RCWA [53, 54]. The result of the corresponding numerical simulation in the case of constant magneto-optic gyration inside the whole magnetic film is given in Figure 4.20 by the red line (Model 1). One can see that this model does not

Figure 4.20 Theoretical fits to Sample 2 data based on uniform gyrotropy parameter in the film (Model 1), gyrotropy parameter proportional to Bi content in the film (Model 2), and gyrotropy parameter enhancement at the top surface (Model 3). A sevenfold g-value magnification is predicted within 2 nm of the surface (Model 3), and fourfold magnification over the next 2 nm, as compared with bulk g value. Source: Levy et al. [56]. ©2019, The Optical Society.

Specific Faraday rotation (°/μm)

Acknowledgment

10 9

Experiment Model 1 Model 2 Model 3

8 7 6 5 4 3 0

0.1 0.2 0.3 Magnetic film thickness (μm)

0.4

reproduce the pronounced increase in the SFR obtained experimentally (points, squares, and diamonds in Figure 4.19). Therefore, the assumption of constant gyration inside the ferrimagnetic film does not describe the observed abrupt growth in the SFR in ultrathin ferromagnetic films. Assuming a gyration proportional to the Bi content, comprising also its decline in the transient layer (Figure. 4.17, blue line), the numerical simulation produces an even poorer coincidence with the experimental data than Model 1, constant gyration. This is shown by the blue line (Model 2) in Figure 4.20. Notice that Model 1 in effect already assumes a significant increase in magneto-optic gyration in the interfacial region if one considers the decrease in Bi content in the transient layer. Yet the calculated SFR in that case remains well below the experimental data. To achieve coincidence with the experimental data, the gyrotropy parameter g was assumed to increase near the air/film interface. Best fits to the experimental data are produced by a rise in g value taking place over a very thin layer adjacent to the ′ surface, which is h = 4 nm thick. In Model 3, the gyration evinces a near-surface steplike growth, as depicted in Figure 4.5. One can see from Figure 4.20 (Model 3) that this assumption provides a very good agreement with the SFR, including sub-50-nm-thick iron garnet films. Note that this model has a good agreement not only with the experimental data for Sample 2 (Figure 4.20) but also with the measurements of all three samples. All theoretical curves in Figure 4.19 are based on the Model 3 and demonstrate a good agreement with the experiment. These fits predict that the magneto-optic gyrotropy parameter exhibits a sevenfold amplification in magnitude within 2 nm of the surface and a fourfold amplification in the next 2 nm over its bulk value. Please see Chapter 5 for the Devices and Applications based on the thin iron-garnet films.

Acknowledgment This work is financially supported by the Russian Ministry of Education and Science, Megagrant project N 075-15-2019-1934.

155

156

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

List of Abbreviations and Symbols 1D-MPC BiIG BM CMZGG CNGG CW FHLs FM FP-1D-MPC FR GGG LPE MCD MO MPC PBG RCWA RIBS SFR SIMS SPP T-1D-MPC TMOKE TPP

one-dimensional magneto-photonic crystal bismuth-substituted iron garnet Bragg mirror calcium–manganese–zirconium–gadolinium–gallium garnet calcium–niobium–gallium garnet continuous wave Faraday hysteresis loops figure of merit Fabry–Perot one-dimensional magnetophotonic crystal Faraday rotation gadolinium–gallium garnet liquid-phase epitaxy magnetic circular dichroism magneto-optical magnetophotonic crystal photonic band gap rigorous coupled-wave analysis reactive ion-beam sputtering specific Faraday rotation secondary- ion- mass- spectroscopy surface plasmon polariton Tamm one-dimensional magnetophotonic crystal transverse magneto-optical Kerr effect Tamm plasmon polariton

References 1 Geller, S. and Gilleo, M.A. (1957). Structure and ferrimagnetism of yttrium and rare-earth–iron garnets. Acta Crystallographica 10 (3): 239–239. 2 Vasiliev, M., Alam, M.N.-E., Kotov, V.A. et al. (2009). RF magnetron sputtered (BiDy)3 (FeGa)5 O12 :Bi2 O3 composite garnet-oxide materials possessing record magneto-optic quality in the visible spectral region. Optics Express 17 (22): 19519–19535. 3 Wettling, W. et al. (1973). Optical absorption and Faraday rotation in yttrium iron garnet. Physica Status Solidi (b) 59 (1): 63–70. 4 Kajiwara, Y. et al. (2010). Transmission of electrical signals by spin-wave interconversion in a magnetic insulator. Nature 464 (7286): 262. 5 Uchida, K. et al. (2010). Spin Seebeck insulator. Nature Materials 9 (11): 894–897. 6 Bozhko, D.A., Serga, A.A., Clausen, P. et al. (2016). Supercurrent in a room-temperature Bose–Einstein magnon condensate. Nature Physics 12: 1057–1062.

References

7 Zvezdin, A.K. and Kotov, V.A. (1997). Modern Magnetooptics and Magnetooptical Materials. CRC Press. 8 Randoshkin, V.V. and Chervonenkis, A.Ya. (1990). Prikladnaya magnitooptika (Applied Magnetooptics). Moscow: Energoatomizdat. 9 Okuda, T., Koshizuka, N., Hayashi, K. et al. (1988). Synthesis of new magnetooptical material. IEEE Translation Journal on Magnetics in Japan 3: 483–484. 10 Kahl, S. and Grishin, A.M. (2004). Enhanced Faraday rotation in all-garnet magneto-optical photonic crystal. Applied Physics Letters 84: 1438–1440. 11 Fedyanin, A.A., Kobayashi, D., Nishimura, K. et al. (2005). Materials Research Society Symposium Proceedings 834 (J1.5): 1–4. 12 Inoue, M., Uchida, H., Nishimura, K. et al. (2006). Magnetophotonic crystals – a novel magneto-optic material with artificial periodic structures. Journal of Materials Chemistry 16: 678–684. 13 Okuda, T., Katayama, T., Satoh, K. et al. Recent advances in magnetism and magnetic materials. Proceedings of the Fifth Symposium on Magnetism and Magnetic Materials, 61–76. 14 Okuda, T., Koshizuka, N., Hayashi, K. et al. (1987). Faraday rotation in highly Bi-substituted yttrium iron garnet films prepared by ion beam sputtering. IEEE Transactions on Magnetics 23 (5): 3491. 15 Berzhansky, V.N., Karavainikov, A.V., Milyukova, E.T. et al. (2010). Synthesis and properties of substituted ferrite-garnet films for one-dimensional magnetophotonic crystals. Functional Materials 17: 120–126. 16 Berzhansky, V.N., Shaposhnikov, A.N., Prokopov, A.R. et al. (2011). One-dimensional magnetophotonic crystals based on double-layer Bi-substituted iron garnet films. Materialwissenschaft und Werkstofftechnik 42: 19–23. 17 Berzhansky, V.N., Shaposhnikov, A.N., Karavainikov, A.V. et al. (2013). The effect of Faraday rotation enhancement in nanolayered structures of Bi-substituted iron garnets. Solid State Phenomena 200: 233–238. 18 Berzhansky, V., Mikhailova, T., Shaposhnikov, A. et al. (2013). Magneto-optics of nanoscale Bi: YIG films. Applied Optics 52: 6599–6606. 19 Shaposhnikov, A.N., Prokopov, A.R., Karavainikov, A.V. et al. (2014). Modification of Bi: YIG film properties by substrate surface ion pre-treatment. Materials Research Bulletin 55: 19–25. 20 Li, J., Zhang, W., Song, Y. et al. (2016). Template transfer nanoimprint for uniform nanopores and nanopoles. Journal of Nanomaterials 2016: 1–7. 21 Parkin, S., Xin, J., Kaiser, C. et al. (2003). Magnetically engineered spintronic sensors and memory. Proceedings of the IEEE 91 (5): 661–680. 22 Baibich, M.N., Broto, J.M., Fert, A. et al. (1988). Giant magnetoresistance of (001)Fe/(001)Cr magnetic superlattices. Physical Review Letters 61 (21): 2472–2475; Press release. 23 Candid, R., María-Dolores, C.B., and Diego Ramírez, M.O. (2009). Magnetic field sensors based on giant magnetoresistance (GMR) technology: applications in electrical current sensing. Sensors 9 (10): 7919–7942.

157

158

4 Bismuth-Substituted Iron Garnet Films for Magnetophotonics

24 Grunberg, P., Schreiber, R., Pang, Y. et al. (1987). Layered magnetic structures: evidence for antiferromagnetic coupling of Fe layers across Cr interlayers. Physical Review Letters 61 (8): 3750–3752. 25 Alvarado, S.F. and Carbone, C. (1988). Magnetism and epitaxy of Fe/Cr(001) multilayers. Physica B & C 149B+C (1–3): 43–48. 26 Okuda, T., Katayama, T., Satoh, K., and Yamamoto, H. (1991). Preparation of polycrystalline Bi3 Fe5 O12 garnet films. Journal of Applied Physics 69 (8): 4580. 27 Shaposhnikov, A.N., Berzhansky, V.N., Prokopov, A.R. et al. (2010). Scientific Notes of Taurida National University, Series: Physics and Mathematics Sciences, vol. 23, 62, No. 1, 146. 28 Berzhansky, V.N., Shaposhnikov, A.N., Prokopov, A.R. et al. (2016). One-dimensional magnetophotonic crystals with magnetooptical double layers. Journal of Experimental and Theoretical Physics 123 (5): 744–751. 29 Toraya, H. and Okuda, T. (1995). Crystal structure analysis of polycrystalline Bi3 Fe5 O12 thin film by using asymmetric and symmetric diffraction techniques. Journal of Physics and Chemistry of Solids 56 (10): 1317. 30 Okuda, T., Kudox, A., Yoshihara, S. et al. (1997). In situ growth of polycrystalline bismuth-iron-garnet films on quartz glass substrate. Journal de Physique IV 7: C1–C707. 31 Berzhansky, V.N., Mikhailova, T.V., Karavainikov, A.V. et al. (2012). Microcavity one-dimensional magnetophotonic crystals with double layer iron garnet. Journal of the Magnetics Society of Japan 36 (1_2): 42–45. 32 Mikhailova, T.V., Berzhansky, V.N., Shaposhnikov, A.N. et al. (2018). Optimization of one-dimensional photonic crystals with double layer magneto-active defect. Optical Materials 78: 521–530. 33 Mikhailova, T., Shaposhnikov, A., Prokopov, A. et al. (2018). Tamm plasmon-polaritons structures with Bi-substituted garnet layers. EPJ Web of Conferences 185: 02016. 34 Deb, M., Popova, E., Fouchet, A., and Keller, N. (2012). Magnetooptical Faraday spectroscopy of completely bismuth-substituted Bi3 Fe5 O12 garnet thin films. Journal of Physics D: Applied Physics 45: 455001. 35 Dionne, G.F. (2010). Magneto-optical properties. In: Magnetic Oxides, 343–384. Springer Science & Business Media. 36 Wittekoek, S., Popma, T.J.A., Robertson, J.M., and Bongers, P.F. (1975). Magneto-optic spectra and the dielectric tensor elements of bismuth-substituted iron garnets at photon energies between 2.2–5.2 eV. Physical Review B 12: 2777–2788. 37 Belov, K.P., Zvezdin, A.K., Kadomseva, A.M., and Levitin, R.Z. (1979). Orientation Transitions in Rare Earth Magnetics. Moscow: Nauka (In Russian). 38 Lyubchanskii, I.L., Dadoenkova, N.N., Lyubchanskii, M.I. et al. (2003). Magnetic photonic crystals. Journal of Physics D: Applied Physics 36: R277–R287. 39 Inoue, M., Baryshev, A.V., Goto, T. et al. (2013). Magnetophotonic crystals: experimental realization and applications. In: Magnetophotonics (eds. M. Inoue, M. Levy and A.V. Baryshev), 163–190. Berlin Heidelberg: Springer-Verlag.

References

40 Dzibrou, D.O. and Grishin, A.M. (2009). Fitting transmission and Faraday rotation spectra of [Bi3 Fe5 O12 /Sm3 Ga5 O12 ]m magneto-optical photonic crystals. Journal of Applied Physics 106: 043901. 41 Khartsev, S.I. and Grishin, A.M. (2007). High performance [Bi3 Fe5 O12 /Sm3 Ga5 O12 ]m magneto-optical photonic crystals. Journal of Applied Physics 101: 053906. 42 Takahashi, K., Takagi, H., Shin, K.H. et al. (2007). Figures of merit of magneto-optic spatial light modulators with magnetophotonic crystals. Physica Status Solidi (c) 4: 4536–4539. 43 Tomilin, S.V., Berzhansky, V.N., Shaposhnikov, A.N. et al. (2016). Ultrathin and nanostructured Au films with gradient of effective thickness. Optical and plasmonic properties. Journal of Physics: Conference Series 741: 012113. 44 Yin, C.P., Wang, T.B., and Wang, H.Z. (2012). Magneto-optical properties of onedimensional conjugated magnetophotonic crystals heterojunctions. European Physical Journal B 85: 104. 45 Berzhansky, V.N., Shaposhnikov, A.N., Prokopov, A.R. et al. (2016). One-dimensional magnetophotonic crystals with magneto- optical double layers. Journal of Experimental and Theoretical Physics 123: 744–751. 46 Berzhansky, V.N., Karavainikov, A.V., Mikhailova, T.V. et al. (2017). Nano- and micro-scale Bi-substituted iron garnet films for photonics and magneto-optic eddy current defectoscopy. Journal of Magnetism and Magnetic Materials 440: 175–178. 47 Armelles, G., Cebollada, A., Garcia-Martin, A., and Ujue Gonzalez, M. (2013). Magnetoplasmonics: combining magnetic and plasmonic functionalities. Advanced Optical Materials 1: 10–35. 48 Khokhlov, N.E., Prokopov, A.R., Shaposhnikov, A.N. et al. (2015). Photonic crystals with plasmonic patterns: novel type of the heterostructures for enhanced magneto-optical activity. Journal of Physics D: Applied Physics 48: 095001. 49 Levy, M., Chakravarty, A., Huang, H.-C., and Osgood, R.M. Jr., (2015). Large magneto-optic enhancement in ultra-thin liquid-phase-epitaxy iron garnet films. Applied Physics Letters 107: 011104. 50 Belotelov, V.I., Akimov, I.A., Pohl, M. et al. (2011). Nature Nanotechnology 6: 370–376. 51 Borovkova, O.V., Hashim, H., Kozhaev, M.A. et al. (2018). TMOKE as efficient tool for the magneto-optic analysis of ultra-thin magnetic films. Applied Physics Letters 112: 063101. 52 Maier, S.A. (2007). Plasmonics: Fundamentals and Applications. Springer-Verlag US. 53 Moharam, M.G., Grann, E.B., Pommet, D.A., and Gaylord, T.K. (1995). Journal of the Optical Society of America A 12: 1068. 54 Li, L. (2003). Journal of Optics A: Pure and Applied Optics 5: 345. 55 Palik, E.D. (1998). Handbook of Optical Constants of Solids. Academic Press. 56 Levy, M., Borovkova, O.V., Sheidler, C. et al. (2019). Faraday rotation in iron garnet films beyond elemental substitutions. Optica 6 (5): 642–646.

159

161

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications Andrey A. Voronov 1,2 , Daria O. Ignatyeva 1,2 , Nikolay A. Gusev 2 , Petr M. Vetoshko 2,3 , Nazar V. Lugovskoy 4 , Yujun Song 5,6 , Vladimir N. Berzhansky 4 , and Vladimir I. Belotelov 1,2 1 Lomonosov Moscow State University, Faculty of Physics, Department of Photonics and Microwaves Structures, Leninskie Gory, Moscow 119991, Russia 2 Russian Quantum Center, 45, Skolkovskoye shosse, Moscow, 121353, Russia 3 Kotelnikov Institute of Radioengineering and Electronics, Mokhovaya 11-7, Moscow, 125009, Russia 4 V.I. Vernadsky Crimean Federal University, Physics and Technology Institute, Department of Experimental Physics, Vernadsky Avenue 4, Simferopol, 295007, Russia 5 University of Science and Technology Beijing, Center for Modern Physics Technology, Applied Physics Department, School of Mathematics and Physics, 30 Xueyuan Road, Beijing 100083, China 6 Zhejiang Key Laboratory for Pulsed Power Technology Translational Medicine, Hangzhou Ruidi Biotechnology Company, Hangzhou 310000, China

5.1 Device Assemble and Application of Iron Garnet Films for Ultrasensitive Magnetometry One of the applications of iron garnet films is magnetometry, which is associated with the increase in the sensitivity of flux-gate magnetometers [1], based on the electrodynamics principle of operation, to the level of quantum magnetometers – the SQUIDs [2] and the optically pumped magnetometers (OPMs) [3]. Flux-gate sensors were proposed in 1936 [4] and subsequently proved to be very good in searching the mines during the Second World War. The flux-gate principle for measuring magnetic fields is based on the periodic magnetization and demagnetization of a ferromagnetic core using an excitation coil [1]. When the external (measured) magnetic field effects on such a system, the magnetization reversal frequency is transformed, and additional signals appear at even harmonic frequencies. Usually, a second harmonic is used to record the magnetic field. Flux-gate magnetometers are characterized by maximal interaction energy with the measured magnetic field and have the sensitivity up to 100 pT/Hz1/2 . They are cheap and easy to handle, consume low power, and do not require cooling or heating unlike quantum magnetometers. However, the sensitivity of flux-gate magnetometers is limited due to the magnetic cores magnetic energy fluctuations appearing in the process of magnetization reversal. Such fluctuations are often many orders of magnitude higher than the Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications, First Edition. Edited by Yujun Song. © 2021 WILEY-VCH GmbH. Published 2021 by WILEY-VCH GmbH.

162

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

interaction energy of the core with an external magnetic field [5]. This circumstance corresponds the fact that the magnetization reversal is a nonstationary process accompanied by transformation of domain structure and the formation of dynamic instabilities of magnetization under the action of the own demagnetizing field of the core [6]. As a conventional method, the use of magnetic cores with high magnetization and small anisotropy based on iron–nickel alloys allows one to reduce the size and increase the number of domains in the magnetic sample, resulting in smoother averaging of the demagnetizing curve. However, the random character of appearance and disappearance of domains in the process of demagnetization does not change, and the magnetization curve dependence on the applied magnetic field consists of a number of shots related with processes of domain restructuring – Barkhausen’s jumps [5]. Thus, a saturated – single domain – state of the magnetic core element is the basic condition for achieving ultimate sensitivity of the flux-gate magnetic field sensor [7]. A single domain state of the magnet is attainable by applying external saturation field H s = 4𝜋M s . However, a rather high magnetic susceptibility of the core is required in order that the sensor electronics noise will not exceed the intrinsic noise of the core [8]. The magnetic susceptibility, in turn, is inversely proportional to saturation field H s . Therefore, it is necessary to look for ways to reduce the saturation fields of magnetic sensor cores. At a known ratio between thickness and diameter, the minimal saturation field have the magnets in the form of an oblate ellipsoid. However, the use of a special form that provides a minimal saturation field and a single-domain state is a necessary but insufficient method to achieve the ultimate sensitivity of a flux-gate magnetometer. One proposed to use epitaxial iron garnet films as the material for the approximation steps. The epitaxial bismuth iron garnet (BiIG) films with nominal compositions of Re3−x Bix Fe5−y Mey O12 (0 < x < 2, 0.3 < y < 0.7), where Re is a rare-earth element and Me is an iron-substituting element. As shown in [9], films of such composition have an order of magnitude lower constant of the cubic anisotropy than undiluted iron garnet. Moreover, the effective anisotropy of the films in the (111) plane is lower by two orders of magnitude. In addition to the described advantages, the iron garnet films are characterized by low number of defects per unit volume due to large lattice constant and low (compared with ferromagnetic metals) parameter of the Hilbert dissipation. The abovementioned facts provide the necessary material properties for magnetic sensors operating in the homogeneous magnetization rotation regime. In addition, it should be noted that setting up the elliptical thickness profile is a very complex technological problem for iron garnet films. To overcome these difficulties, a special technological approach was developed, which concludes in a stepped profile (Figure 5.1) approximation by multilayer photolithography [10]. The above approaches allowed one to overcome the limitations of the traditional flux-gate technology and obtain a noise level of 100 fT/Hz1/2 , which was experimentally demonstrated [11]. As a sensitive element, a round iron garnet film with a three-step thickness profile was used (Figure 5.2), winded by two orthogonal coils. The coils receive a harmonic signal from the generator and create a rotating control field. Information about the measured field can be obtained by recalculating

5.1 Device Assemble and Application of Iron Garnet Films for Ultrasensitive Magnetometry

Figure 5.1

The thickness profile approximation by the steps.

X-coil Y-coil Iron garnet film (a)

(b)

(c)

Figure 5.2 The sensitive element: (a) the photograph of the three-step round iron garnet film, (b) the winding element, and (c) the winding configuration: X-coil corresponds to the x-component of the measured field and Y-coil corresponds to the y-component. Source: Nikolay A. Gusev, Petr M. Vetoshko.

Figure 5.3 Animal measurement equipment: the sensor, the board, the plate, and the shield. Source: Nikolay A. Gusev, Petr M. Vetoshko.

the amplitude of the response signal arriving form the coils at a double frequency. Another advantage of such device was the synchronous measurements of both magnetic field vector components lying in the plane of the iron garnet film because the electromagnetic filter (EMF) signals from the coils of the film winding, carrying information about the magnetic field, are fed to the measuring system with a certain phase shifts. Each signal at the frequency of the second harmonic carries information about the corresponding vector field component. Operation of the proposed magnetic field sensor is demonstrated by measuring the magnetocardiography (MCG) signals – magnetic field of healthy rat heart. A lab rat was positioned on a textolite plate fixed above the sensor element inserted into the wooden board. In order to suppress external magnetic noise, the board with a rat and sensor was placed inside a magnetic shield comprising four permalloy cylinders characterized by attenuation factor of about 1500 (Figure 5.3). In this configuration, the sensor measures magnetic field components in plane of the rat thoracic cage. The directions were determined by axes of the detecting magnetic coils. The MCG measurements were performed in two periods: real-time monitoring and time averaging for 10 seconds.

163

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

Figure 5.4 Obtained MCG signal of measured rat without averaging.

15 10 H (pT)

164

5 0 ‒5 300

400

500

600 700 t (ms)

800

900

According to the noise characteristic of the magnetometer, the sensor noise in a frequency interval of 1–100 Hz is about 100 fT/Hz1/2 (Figure 5.2b). The peak observed at 50 Hz is related to the AC mains source frequency and amounts to about 400 fT/Hz1/2 . According to the √ Van der Ziel criterion [12], the minimal level of observable signal is Hmin = k ⟨Hf2 ⟩Δf , (where k ∼ 6 and Δf is the signal frequency interval), one can obtain an estimate of H min = 6 pT for MCG measurements in the 1–100 Hz interval. Figure 5.4 shows the obtained MCG signal of measured rat without averaging. The signal period was about 165 ms. The main R peak with a magnitude of about 15 pT is well recognized on the noise background. It should be emphasized that the analogous peak intensity in MCG patterns of humans is 10–30 times higher [13]. The experimentally measured noise level about 4 pT, in agreement with the Van der Ziel criterion. Combining the flux-gate magnetometry with the magneto-optical scanning leads to new possibilities: vector measurements [14, 15] and imaging and micrometer spatial resolution [16]. In the absence of the reading coils in such magnetometers, the Schottky noise can be avoided. Moreover, magneto-optical measurements are contactless, allowing one to use the magnetic sensors in conjunction with optical fibers, which opens new horizons for invasive and probing magnetometry [17, 18]. The following work [14] describes the development of vector measurements of the magnetic field through reading the magnetization state of an iron garnet film by the Faraday rotation together with the technique of magnetization reversal by homogeneous rotation. The control magnetic field H rotates the film magnetization M in the film plane (Figure 5.5) with the rotation frequency 𝜔. Magnetocrystalline anisotropy of the film deviates the magnetization from the film plane and provides the out-of-plane component. This component can be detected with the linearly polarized light beam due to the magneto-optical Faraday effect. A magnetic film of cubic crystal lattice and crystallographic axis orientation (111) is fully symmetric with respect to the film normal. Since the magnetocrystalline anisotropy leads the magnetization out of the film plane, the Faraday signal becomes sensitive to both the in-plane and out-of plane components of the measured magnetic field. This important property establishes a possibility for the vector magnetometry with the iron garnet films.

5.1 Device Assemble and Application of Iron Garnet Films for Ultrasensitive Magnetometry

S

P Z[111] M

𝜔 F

θ 𝜔t

H Y[011]

X[211] A H sin(𝜔t)

H cos(𝜔t)

D

Figure 5.5 The principle scheme of the vector magneto-optical magnetic sensor. “F” is a magnetic film of cubic magnetic anisotropy, “S” is a light source, “P” is a polarizer, “A” is an analyzer, and “D” is a photodetector. The control rotating magnetic field H is generated by two coils.

If the incident beam impinges upon the sample normally, the observed Faraday angle Ψ(t) senses the out-of-plane magnetization component and Ψ(t) ∼ 𝜃(t), where 𝜃 is the polar angle of the magnetization. Since the azimuthal angle of the control magnetic field is 𝜙H = 𝜔t, one can directly measure the azimuthal dependence of 𝜃, 𝜃(𝜙H ). If there is no external magnetic field, then 𝜃(𝜙H ) is a periodic function with a period of 𝜋/3. In the presence of the monitored magnetic field h (h ≪ H), the symmetry is broken, and even Fourier harmonics occur. The magnetometer measures the Fourier amplitudes An and Bn in the Fourier ( ) ∑∞ series 𝜃 𝜙H = n=0 An cos n𝜙H + Bn sin n𝜙H . It can be shown that 𝜃(𝜙H ) = c1 hx sin 2𝜙H + c2 hy cos 2𝜙H + c3 sin 3𝜙H + c4 hx sin 4𝜙H + c5 hy cos 4𝜙H + c6 hz cos 6𝜙H , where ci is the coefficient depends on the magnetic parameters of the film, including the saturation magnetization, M s , and the uniaxial and cubic anisotropy constants, K U and K 1 , respectively. Therefore, all three components of h can be found from A2 , B2 , and A6 normalized by B3 hx =

c3 B2 cA cA , h = 3 2 , and hz = 3 6 c1 B3 y c2 B3 c6 B3

(5.1)

In general case, expressions for coefficients ci versus M s , K U , and K 1 are quite K H ≪ 1, 4𝜋M1 2 ≪ 1. It cumbersome. However, for iron garnet is usually satisfied 4𝜋M s

allows the significant simplifications: √ ) c3 c3 2 2( 2 c3 2𝜋Ms2 − KU H = = H, = c1 c2 3 c6 3K1

s

(5.2)

165

166

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

Equation (5.2) shows that

c3 c1

and

c3 c2

are linear in field H and do not depend on c

anisotropy in these assumptions. On the contrary, c3 strongly depends on K 1 , K u , 6 and M s . In any case, the decrease in H increases the sensitivity. However, H should be high enough to saturate the magnetic film. For experimental demonstration of the proposed concept, a series of the rare-earth iron garnet films of nominal composition (BiLuPrTmGd)3 (FeAlGa)5 O12 were used. The results obtained for an 11.7–μm-thick film of composition Bi1.1 Lu1.45 Pr0.2 Tm0.2 Gd0.05 Fe3.5 Al0.8 Ga0.7 O12 are presented below. The film has a low value of the saturation magnetization 4𝜋M s (75 G), and the out-of-plane magnetic field saturation is H ⟂ = 166 Oe. The latter is important to increase the amplitudes of 𝜃. The magnetic film was placed inside three mutually orthogonal Helmholtz coils (Figure 5.5). Two of them are used to generate the rotating control magnetic field H. The amplitude of the control field is chosen as H = 20 Oe to saturate the sample. The external magnetic field to be measured was generated by the Helmholtz coils as well. The sample was illuminated with a linearly polarized light from a laser diode “S” at wavelength 𝜆 = 630 nm. The whole experimental setup was placed in the magnetic shielded room with the attenuation coefficient of 1500 at frequencies below 1 kHz. It allows one to eliminate Earth’s magnetic field along with 50 Hz AC magnetic fields generated by wire. In the absence of the monitored magnetic field (h = 0), 𝜃 (𝜙H ) has only the third harmonic (black curves in Figure 5.6a). The presence of a small additional constant external magnetic field h causes some distortion of the signal (Figure 5.6), and the other harmonics appear (Figure 5.6b–d). Experimental data show that A2 is linear in hy only (Figure 5.6b), while B2 is linear exclusively in hx (Figure 5.6c). At the same time, A6 is linear in hz (Figure 5.6d). This is in full agreement with Eq. (5). Note that in a relatively large field hz > 0.6 Oe, the dependency A6 (hz ) becomes nonlinear (Figure 5.6d). Moreover, each Fourier amplitude depends nonlinearly on two other components of h (A2 depends nonlinearly on hx and hz , A6 – on hx and hy , and B2 – on hy and hz ). This does not contradict to Eq. (5) since only linear in h terms was sustained in them. These nonlinear contributions do not influence the sensor functionality since the corresponding derivatives vanish for h → 0. The linear dependence of A2 , B2 , and A6 on different components of the external magnetic field allows one to use the proposed structure as a vector magnetometer. At first, the magnetometer should be calibrated by applying the external magnetic field in three orthogonal directions. It allows one to find c3 /c1 , c3 /c2 , and c3 /c6 from the slopes of the straight line fits of the corresponding dependencies on h (dashed lines in Figure 5.6b–d) and from the amplitude of the third harmonic. As c1 = c2 , they can be found by averaging the slopes of A2 (hy ) and B2 (hx ) divided by A3 . In our case, c3 /c1 = c3 /c2 = 15.01 and c3 /c6 = 38.34. The magnetometer detection threshold is determined by the total fluctuations arising of the magnetic material during the magnetization reversal and the noise of the optical tract appearing due to fluctuations of the light source intensity and photon shot noise. The last two factors are determined by the optical resolution of the

x[211] 0.012

45+

0.006

x[211]

z[111] 315+

x[211]

45+

315+

45+

315+

Faraday angle, red

0.000 –0.006 –0.012 –0.000

270+ 90+

90+

270+ 90+

270+

–0.006 0.000 0.006

225+

135+

225+

135+

225+

135+

0.012

(a)

0.06 0.05

180+

180+

hx

hy

hz

A2/A3(hx) A2/A3(hy) A2/A3(hz)

0.07 0.06

Linear fit of A2/A3(hy)

0.05

0.03

0.020

Linear fit of B2/A3(hx)

0.03

0.012 0.008

0.01

0.01

0.004

0.00

0.00

–0.01 0.0 (b)

0.2

0.4

0.6 hi (Oe)

0.8

1.0

–0.01 0.0 (c)

Linear fit of A6/A3(hz)

0.016

0.02

0.02

A6/A3(hx) A6/A3(hy) A6/A3(hz)

0.024

0.04 B2/A3

A2/A3

0.04

B2/A3(hx) B2/A3(hy) B2/A3(hz)

A6/A3

0.07

180+

0.2

0.4

0.6 hi (Oe)

0.8

1.0

0.000 –0.01 0.0 (d)

0.2

0.4

0.6

0.8

hi (Oe)

Figure 5.6 (a) The variation of the signal Ψ(𝜙H ) for three orthogonal directions (x, y, z) of the external field h – 0 Oe (black), 3 Oe (red), and 6 Oe (blue) and Fourier amplitudes A2 (b), B2 (c), and A6 (d) versus magnetic fields hx (black), hy (red), and hz (blue) normalized to A3 . The linear in hi harmonics can be used to measure all the three projections of the external magnetic field.

168

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

√ device. In modern balancing detectors, this value is ⟨ΔI 2 ⟩∕I ∼ 10−9 , in where I and ΔI are incident intensity and its fluctuation, respectively [19]. Taking into account the conversion factor of 1.5 × 10−3 rad/Oe, one obtains the sensitivity of the order of 100 pT in a 1 Hz bandwidth. Since the induction reading circuit of the magnetometer above gives 100 fT/Hz1/2 , one can conclude that the sensitivity of the discussed vector magneto-optical magnetometer is limited by the optical parameters of the detection scheme, which could be improved by utilizing magneto-optical intensity effects instead of Faraday effect. In general, the in-plane magnetization of the transparent magnetic films is not straightforwardly detectable by light. For technical reasons it is preferable to observe magnetic films at normal or slightly oblique incidence; otherwise special correcting optics or prisms are required that make the sensing element bulky and introduce excess noise. Therefore, one should exploit the magneto-optical effects that are sensitive to the magnetization component orthogonal to the light wave vector. Among such effects, the widely used transverse and longitudinal magneto-optical Kerr effects require oblique incidence at large angles, and in addition the transverse effect is negligibly small for transparent media. In addition to the Kerr effects is the Voigt effect, which is usually quite small for transparent media and may be insufficient for employment in magnetometry. It is well known that a plasmonic cover of the transparent films significantly modifies the magneto-optical response, and the Faraday and transverse Kerr effects are resonantly enhanced by several orders of magnitude [20–22]. The transverse Kerr effect might not be an optimal option compared to other magneto-optical effects arising in the magnetoplasmonic crystals. In the case of a magnetic film covered with a one-dimensional gold grating of subwavelength slits, the longitudinal magnetophotonic intensity effect (LMPIE) arises [20]. Importantly, the LMPIE takes place for normal light incidence, which perfectly fits the requirements of the magnetometry with the in-plane magnetized films. Furthermore, in contrast to the Faraday effect and other bulk magneto-optical effects, to obtain high values of the LMPIE does not require thicker magnetic films. In fact, the LMPIE almost saturates for the 1-μm-thick magnetic films, which provides room for miniaturization. In the next work [23], one proposes and demonstrates a novel type of magneto-optical magnetic field sensor based on the LMPIE in the magnetoplasmonic crystal: a structure of an iron garnet film and a thin gold layer pierced with a periodic slit array. One shows that such a structure provides enhancement of magneto-optical magnetometry and allows using the planar component of magnetization for reading low magnetic fields. The proof-of-concept sample of the magnetoplasmonic sensor is demonstrated in the experimental study a sensitivity level of 2 nT/Hz1/2 . Let us consider, similar to the previous case, an iron garnet film of cubic crystal lattice and crystallographic axis orientation (111) placed into a saturating control magnetic field H, which is rotated in the film plane at a frequency 𝜔. The azimuth angle of the control field varies as 𝜙H = 𝜔t, and then the film magnetization M follows it with some delay in azimuth angle, Δ𝜙 = 𝜙 − 𝜙H , where 𝜙 is azimuth angle of M. The magnetocrystalline anisotropy of the film deviates the magnetization

5.1 Device Assemble and Application of Iron Garnet Films for Ultrasensitive Magnetometry

from the film plane by an angle 𝜃 and provides its out-of-plane component. In the presence of the monitored magnetic field h, the time dependences of both angles Δ𝜙(t) and 𝜃(t) are modified, and analyzing their spectra allows to determine the external magnetic field value and direction. The behavior of Δ𝜙(t) and 𝜃(t) depends on the relation between the control magnetic field and the field of the cubic magnetic anisotropy H C ∼K 1 /M s , where K 1 is the cubic anisotropy constant and M s . is the saturation magnetization. We restrict our consideration to the case of relatively weak field of the cubic magnetic anisotropy H C ≪ H. At this the angle 𝜃 is negligibly small and it is more preferable to detect Δ𝜙(t). Therefore, it is necessary to use a magneto-optical effect that is sensitive to the in-plane magnetization. However, in the case of the transparent magnetic films the Voigt and transverse Kerr effects governed by the in-plane magnetization are quite small and are not suitable. Thus, we deposited on the magnetic film a plasmonic one-dimensional grating to make optical transmittance and reflectance strongly sensitive to the in-plane magnetization as made possible due to the LMPIE. The LMPIE is even in M x and is measured by relative change of the detected light intensity 𝛿 for the remagnetization of the structure from the state with M x = M s to M x = 0. For the sake of mechanical stability, it is more advantageous to detect the transmitted signal. In this case the LMPIE magnitude is given by [20]: ( ) T Ms − T0 (5.3) 𝛿= T0 where T(M s ) and T 0 are transmittance for M x = M s and M x = 0, respectively. Due to the LMPIE, the rotating magnetization causes variations in transmittance that are quadratic in M x [20]: [ ) ] ( Mx (t) 2 𝛿 (5.4) T (t) = T0 1 + Ms where 𝛿 is the LMPIE value measured for the fully saturated magnetic film. Consequently, measuring and analyzing light transmittance through the magnetoplasmonic crystal allows determining M x (t) from Eq. (5.4). To establish a magnetometer scheme on this basis, one needs to relate M x (t) and the monitored magnetic field h. To get the main expressions characterizing the proposed magnetic sensor, one can assume that the monitored magnetic field h varies much slower than H(t) and consequently, it can be considered constant. In this case, the total magnetic field influencing the magnetization is [ ] ] [ Htot (t) = H (t) + h = H cos (𝜔t) + hx ex + H sin (𝜔t) + hy ey + hz ez (5.5) where ei are unitary vectors along the coordinate axes. As the sensor is aimed at the measurement of weak magnetic fields (h ≪ H), we use the linear in h approximation. Neglecting the influence of anisotropy, we assume that magnetization is directed along H tot . Then, taking into the account that the magnetization is fully saturated, for the component of magnetization perpendicular to the slits M x , from Eq. (5.5), one can find } { hy hx hx cos (2𝜔t) − sin (2𝜔t) + (5.6) Mx (t) = Ms cos (𝜔t) − 2H 2H 2H

169

170

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

Substitution of Eq. (5.6) to Eq. (5.4) provides the dependence of the optical transmittance on h, which can be presented as the decomposition into the temporal harmonics: ̃2 (t) + … ̃1 (t) + T ̃0 + T T (t) = T where T k (t) is the combination of terms with cos(k𝜔t) and sin(k𝜔t). In particular, ] [ hy hx ̃3 (t) = −𝛿T0 cos (3𝜔t) + sin (3𝜔t) (5.7) T 2H 2H It means that the third harmonic of T(t) is proportional to the in-plane component h|| = (hx , hy , 0) of the monitored magnetic field. The same is valid also for the first harmonic of T(t). The proposed concept implies restrictions on H and 𝜔. The rotating field H should be strong enough so that its contribution to the free energy of magnetic field is much | | K2 stronger than that of anisotropy, which implies H ≫ || 4M 2𝜋M1 2 −K ||, where K U and | s( s U ) | K 1 are the uniaxial and cubic anisotropy constants, respectively. These constants are material parameters that characterize the energy of the magnetic anisotropy. Therefore, for the measurement of h|| , one can analyze the first or the third harmonic of the modulated transmittance. Since the photodetector current J is equal to the number of electrons knocked out by photons received by the photodetector per unit time: 𝜂e𝜆 IT J= (5.8) 2𝜋cℏ 0 For the amplitude of the third harmonic of the photodetector voltage U 3 , using Eq. (5.8), one can obtain h∥ e R𝜂𝜆I0 T0 𝛿 (5.9) 4𝜋cℏ H where I 0 and 𝜆 are the intensity and wavelength of the incident radiation, R is the photodetector transimpedance amplifier photocurrent-to-voltage conversion ratio, 𝜂 is the quantum yield of the photodetector, e is electron charge, c is speed of light in vacuum, and ℏ is Planck’s constant. One can conclude that the detected signal is proportional to the laser intensity, optical transmittance through the magnetoplasmonic crystal, and the LMPIE magnitude. The plasmic gratings were deposited on two rare-earth iron garnet films of Bi0.9 Gd2.1 Fe4.41 Sc0.59 O12 and Bi0.9 Y1.2 Lu0.9 Fe4.2 Sc0.8 O12 composition. Then, the optical and magneto-optical properties of the magnetoplasmonic crystals were measured, and three magnetoplasmonic samples with the maximal values of LMPIE were selected. The parameters of the samples are shown in Table 5.1. Scheme of the experimental setup for demonstration of the proposed magnetic field sensing method, based on the optical balanced method, is shown in Figure 5.7. Two tunable semiconductor lasers were used: the first with a power of 500 mW (laser-A) and wavelength near 805 nm and the second one with a power of 90 mW and wavelength near 780 nm (laser-B). The optical setup was based on a balanced measurement method. The radiation generated by the laser diode 1 was collected U3 =

5.1 Device Assemble and Application of Iron Garnet Films for Ultrasensitive Magnetometry

Table 5.1

The parameters of the samples used.

Name

Magnetic layer thickness (μm)

Sample 1

Sample 2

Sample 3

1.9

1.9

1.6

Specific Faraday rotation (deg/μm) (at 632 nm)

1.4

1.4

1.3

Grating period (nm)

340

335

350

Grating slit width (nm)

120

200

170

LMPIE

0.28

0.093

0.074

Wavelength 𝜆 (nm)

804.1

781

776.4

U 3 at 10 nT (mV)

0.26

0.04

0.05

I 0 T 0 𝛿 (mW)

16.1

2.7

3.2

Best observed sensitivity (nT)

8.5

2.4

2.9

10

41

42

Sensor noise hn at 𝜈 = 0 (pT/Hz ) 1/2

1

2

3

5

4

6

7

h

h

Figure 5.7 Scheme of the experimental set-up for demonstration of the magnetoplasmonic magnetometer. (1) Diode laser, (2) collecting lens, (3) polarizer, (4) sample with coils, (5) diffuser lens, (6) Wollaston prism, and (7) balanced photodetector. The monitored field is directed along the grating slips (h = {0,hy , 0}).

by an aspherical close-focus lens 2 into a converging beam. The polarization of this beam was set linear by the Glan–Taylor polarizer 3 so that, on the sample 4, which is an iron garnet film with a layer of gold gratings, it was oriented at an angle of 45∘ to the grating lines. Due to the small focal length of the lens 2 equal to 6 mm, the optical waist length reached 1.2 mm, which allowed the sample to be set in the center of the waist and to provide a narrow angular spectrum of optical radiation incident on the plasmonic crystal and high intensity of light at the same time. The transmitted light was modulated by a magnetic field in accordance with the laws described above. Next, the radiation was divided into two beams with TE and TM polarization using a Wollaston 6 prism and was focused by the lens 5 on photo sensors of a balanced photo detector 7 as it shown in Figure 5.7. The photodetector contained the two Si photodiodes and a specially made transimpedance amplifier with its own output noise of about 0.6 μV/Hz1/2 and a current-to-voltage conversion ratio R = 2500 Ω. The photodetector signal was digitized by the data acquisition board and transferred to a PC for further processing. The signal processing is used in obtaining its spectrum by

171

172

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

a fast Fourier transform and extracting the first three harmonics from this spectrum: the second one to estimate the LMPIE value, and the first and third ones to measure the external field h. The sample was placed in a magnetic field H rapidly rotating in the film plane. The field was formed in a small volume, about 10−3 cm3 , by a system of coils without a core to eliminate the influence of its magnetic noise. The frequency of rotation of the magnetic field in the experiment was 126 kHz. The field amplitude did not exceed 4 Oe, which is sufficient to saturate the sample magnetic film providing a single domain state. Optical radiation passing through the sample was modulated by a magnetic field at a doubled frequency of rotation of the magnetic field due to the LMPIE effect. The inhomogeneity of the rate of rotation of the magnetization of the sample caused by the presence of a constant or oscillating external field h with a low frequency led to the appearance of the first U 1 at 126 kHz and third U 3 at 378 kHz harmonics in the spectrum of the photodetector signal. Thus, the output photodetector signal consists in three harmonics: U 1 , U 2 and U 3 , where U 1 and U 3 are proportional to the measured field h. In the experiment the field h was formed by the reference system of Helmholtz coils with a diameter of 100 mm. The inductance of this coil allowed one to generate a magnetic field at the frequencies below 1 kHz. As a result of the magneto-optical modulation, this spectral range was transferred to frequencies near the first (126 kHz) and third (378 kHz) harmonics of the photodetector signal. Since LMPIE is a resonant effect, the lasers were tuned to wavelengths corresponding to the selected LMPIE values (row “wavelength” in the Table 5.1). A monitored magnetic field h was applied parallel to the slits of the magnetoplasmonic crystal and varied from 1 nT to 10 μT, and the time dependence of the transmitted light intensity at the operating wavelengths corresponding to the maxima of the LMPIE were measured. The noise characteristic of the setup measured for the sample-2 at frequencies less than 1 kHz is shown in the inset of Figure 5.8. In this range the noise level almost does not change with frequency, which is essential for the magnetometry of low-frequency magnetic fields. The photodetector signal of the third harmonics by amplitude U 3 is proportional to h for all three samples (see linear fits for the experimental data for samples 1, 2, and 3 shown with circles, triangles, and squares, respectively, in Figure 5.8). This completely agrees with Eq. (5.9). The largest signal is observed for sample 1 (circles, blue line in Figure 5.8). It exceeds U 3 for the other two samples by six and five times, respectively. Actually, in accordance to Eq. (5.9), for a given monitored magnetic field, the signal of the photodetector is proportional to the product of the incident light intensity, transmittance through the grating and the LMPIE magnitude: U 3 ∼I 0 T 0 𝛿. The relative values of this product are in good agreement with the observations: the product for sample 1 is approximately six times larger than for sample 2 and five times larger than for sample 3. The sensitivity of the magnetometer scheme is also related by the noise level, which is U 3c = 0.225 mV for the laser-A (at 𝜆 = 802.9 nm), U 3c = 0.011 mV for the laser-B (at 𝜆 = 781.0 nm), and U 3c = 0.015 mV for the laser-B (at 𝜆 = 776.6 nm). Therefore, the laser-B is much more suitable for the magnetometer measurements.

5.1 Device Assemble and Application of Iron Garnet Films for Ultrasensitive Magnetometry

Sample 1 102

Sample 2 Sample 3

100

100 U3 (mV)

U3 (mV)

101

10–1

10–2

10–1 10–2 10–3 0

200

10–3 1

10

100

1000

400 600 f (Hz)

10 000

800

1000

100 000

hy (nT)

Figure 5.8 Dependence of the third harmonics amplitude, U3, on the monitored magnetic field h = {0, hy , 0} oscillating at 515 Hz. Inset: The noise characteristic of the magnetometer setup measured for the sample 2.

As a result, the lowest monitored field was detected with the sample-2: hmin = 2.4 nT (red asterisk in Figure 5.8). For the sample-3, it is a bit larger: hmin = 2.9 nT. Potentially, sample-1 would give much better results if laser-B could be tuned to its main resonance at 𝜆 = 804.1 nm; in this case the minimum detectable field would be less than 1 nT: hmin = 0.9 nT. In Figure 5.8, the magnetometer has a large dynamic range. The maximal measured field could be considered about 11 μT for sample 1 and 75 μT for samples 2 and 3: above the given values of the monitored fields, the scheme passes in a nonlinear mode, due to the fact that such values are close to the magnitude of the control field. In conclusion, let us give an estimate of the ultimate sensitivity parameters for a proposed magnetometry scheme. The noise value of the third harmonic of the photodetector current in the frequency band Δf is composed of the shot noise arising from the photodetector and the laser intensity fluctuations that are related to the physical processes of photon generation and are caused by fluctuations in the laser pump current. In view of this circumstance, for the magnetometer noise level hn in the Δf band, one can get (√ ) √ 2H 4𝜋cℏ hn = +𝜈 𝛥f (5.10) δ 𝜂𝜆I0 T0 where 𝜈 is the relative amplitude of the noise modulation of the laser intensity.

173

174

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

When the photodetector noise prevails over the noise level of the laser (i.e. if 𝜈 ≪ 4𝜋cℏ/(𝜂𝜆I 0 T 0 )), then the sensitivity increases proportionally to the product √ of 𝛿 I0 T0 . However, if the laser noise is larger than the photodetector noise (i.e. 𝜈 ≫ 4𝜋cℏ/(𝜂𝜆I 0 T 0 )), then the sensitivity increases proportionally to 𝛿. In the experimental setup, the noise was mainly due to the laser noise. However, assuming use of an ideal laser, i.e. 𝜈 = 0, so that the only noise source of the measuring scheme is the shot noise, one could find the detection limit of this magnetometer scheme for three samples (see last row in Table 5.1). The highest sensitivity is expected for the sample 1 and might reach 10 pT/Hz1/2 , which is 10 times higher than the estimate for a smooth magnetic film in the previous case. One of the main advantages of the optical reading approach with respect to the induction method is related to a high spatial resolution, which can be obtained. In the lateral direction, it is limited by the diameter of the focused laser beam. To achieve necessary functionality of the plasmonic grating, at least several periods of the grating must be illuminated by the laser beam, increasing the size of the illuminating beam to a few microns. On the other hand, spatial resolution in the orthogonal to the magnetic film direction is determined by the thickness of the magnetic film. The LMPIE magnitude remains at a rather high level even for magnetic films of 100 nm in thickness that provides submicron resolution in this direction. In the end of our discussion of different types magnetometry schemes with iron garnet films and heterostructures, it should be noted that the noise related to the fluctuations of the magnetization, estimated for the square garnet film with 1 × 1 × 0.0001 cm3 with 4𝜋M s = 1750 G and 𝛼 = 0.0001 at room temperature according to [24], is at the level of 1 fT/Hz1/2 . To reach this level of sensitivity, one should overcome the shot noise and other types of noise coming from the detection scheme.

5.2 Devices Assemble and Application of BiIG Films for Biosensing Plasmonic sensors are one of the most sensitivity nowadays [25, 26]. They are based on the detection of resonances resulting from the excitation of surface plasmon polariton waves, whose properties are very sensitive to the optical characteristics of the surrounding medium. By measuring the optical response of a plasmonic structure (the reflection coefficient) and its variation due to the changes in the refractive index of the medium surrounding the structure, one can determine the concentration of various analytes in liquid solutions and gas mixtures very precisely. Selectivity is ensured by the deposition of special adsorbing coatings on the surface of the sensor. Plasmonic sensors are successfully employed in numerous industries, including food quality control and ecological monitoring. However, the sensitivity of such sensors was insufficient for a number of applications. Thus, two approaches that currently seem the most promising for increasing the sensitivity of magnetoplasmonic sensors were proposed. The first one uses the excitation of long-range modes in plasmonic structures, providing for considerably increased Q-factors of the resonances and the enhanced response of the sensor

5.2 Devices Assemble and Application of BiIG Films for Biosensing

to changes in the refractive index of the surrounding medium [27]. Originally, symmetric dielectric/metal/dielectric structures with a thin metal layer were used for this purpose. However, most applications in biology and chemistry deal with liquids and gases, possessing much lower refractive indices than solid dielectrics. For the excitation of long-range modes at a boundary with these media, it was suggested to use photonic crystal structures, whose effective impedance can be matched to the impedance of the medium under study [28]. The second approach was to use magneto-optical rather than optical measurements. Instead of the reflection spectra, the spectra of the transverse MO Kerr effect (TMOKE) in plasmonic structures are measured in this case [29]. Owing to the presence of metal layers, the magneto-optical response of the structure near plasmon resonances is considerably enhanced [30–32]. Magnetoplasmonic resonances feature higher Q-factors than optical resonances, thus allowing to achieve a severalfold increase in the sensitivity of plasmonic sensors. Both approaches were combined in this paper [33] to make a magnetophotonic plasmonic heterostructure for sensor applications. The heterostructure under investigation contains a photonic crystal required for the occurrence of a long-range mode in an asymmetric structure with a gas-phase analyte, a gold film needed for the excitation of surface plasmon polaritons and a magnetic material providing the magneto-optical response. In [33], bismuth-substituted iron garnet, which is a ferrimagnetic dielectric, was used as the magnetic material. Also, it was demonstrated that instead of magnetic dielectric, ferromagnetic metals can be used, for example, thin layers of cobalt [34]. However, a magnetic dielectric instead of a metal to obtain the magneto-optical response increases considerably the Q-factor of the resonance and the sensitivity of the structure in comparison to magnetoplasmonic sensors based on ferromagnetic metals. A basic feature of photonic crystal structures with a magnetic dielectric layer is that they can support both plasmon modes (when additional metal layers are deposited) and surface electromagnetic waves propagating along the boundary of the magnetic layer. Thus, bearing in mind future sensor applications, a direct comparison of the characteristics obtained using photonic crystals both with and without a thin gold layer has been done. Also, magnetoplasmonic structure with cobalt was added to this comparison to show the advantages of the structures with bismuth-substituted iron garnet layer. The structure with iron garnet under investigation is a photonic crystal formed by 16 layers of tantalum pentoxide (Ta2 O5 , 119.3 nm) alternating with 16 layers of silica (SiO2 , 164.7 nm), a 125-nm-thick layer of bismuth-substituted iron garnet and an 8-nm-thick gold film. However, a part of the sample was left without gold coating in order to study properties of surface electromagnetic waves (Figure 5.9). The excitation of a long-range mode in a magnetophotonic plasmonic structure leads to a resonant enhancement of the transverse magneto-optical Kerr effect. The properties of the obtained enhancement determine the advantages of the proposed sensor (Figure 5.10). The width of the optical resonance was 0.05∘ , which corresponds to a Q-factor of 890. The magneto-optical resonance has a still smaller width of 0.02∘ , corresponding

175

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

Laser

Figure 5.9 Principle scheme of magnetoplasmonic sensor with bismuth-substituted iron garnet magnetic layer.

CMOS matrix Electromagnet

Prism SiO2

Photonic crystal Ta2O5 / SiO2 Garnet layer Au film Gas cell

1.0

0.008 δ 0.004

R (a.u.)

176

0.8 0 0.6

‒0.004

Air

Air Helium

Helium 0.4 44.0

(a)

44.1

44.2 44.3 θ (deg)

44.4

44.5

‒0.008 44.0

(b)

44.1

44.2 44.3 θ (deg)

44.4

44.5

Figure 5.10 (Color online) Angular spectra of the (a) reflectance and (b) transverse Kerr effect for a magnetophotonic plasmonic heterostructure recorded upon the filling of the gas cell with air or helium (black and red curves, respectively). The results of numerical simulations are shown with darker lines.

to a Q-factor of 2200. A variation in the refractive index of the gas under study causes a shift of the resonance by 𝜕𝜃/𝜕n = 39 ∘ /RIU. This yields sensitivities to the refractive-index variation for reflectance and magneto-optical measurements of 𝜕R/𝜕n = 720 RIU−1 for a fixed angle of 𝜃 R = 44.127∘ and 𝜕𝛿/𝜕n = 2.4 × 103 RIU−1 for 𝜃 𝛿 = 44.147∘ , respectively. Calculations indicate that, for an ideal magnetoplasmonic structure with optimum parameters, the width and Q-factor of the reflectance resonance are 0.018∘ and 2.4 × 103 , respectively, while the width and Q-factor of the magneto-optical resonance are 0.005∘ and 8.7 × 103 , respectively. The sensitivity to the refractive-index variation is 𝜕R∕𝜕n = 2.3 × 102 RIU−1 and 𝜕𝛿∕𝜕n = 5 × 103 RIU−1 for reflectance and magneto-optical measurements, respectively. In an ideal structure with no gold

5.2 Devices Assemble and Application of BiIG Films for Biosensing

film, the resonances have smaller widths and higher Q-factors: 𝛥𝜃 SPR = 0.007∘ , 𝛥𝜃 𝛿 = 0.001∘ and QSPR = 6.2 × 103 , Q𝛿 = 4.3 × 104 , respectively. However, the excitation of plasmon polaritons in an ideal sample with a gold film yields a deeper reflectance resonance with a minimum reflectance value of Rmin = 0.17%, in contrast to the structure with no gold film, where Rmin = 1.7%. A greater depth of the optical resonance also gives rise to the enhancement of the transverse Kerr effect in the plasmonic structure: 𝛿 max = 62% and 19% in the structures with and without a gold film, respectively. As a result, in comparison with an ideal magnetoplasmonic structure, the sensitivity of an ideal structure without gold coating appears to be higher for optical measurements (𝜕R𝜕n = 3.7 × 103 RIU−1 ) measured for the fabricated sensor structures along with those calculated for the structures with optimum parameters are listed in the table. Also, the results of theoretical calculations with the ideal structure covered with cobalt and gold layers were added to this table to demonstrate the advantages of applying iron garnet layer. The ideal cobalt structure is one-dimensional photonic crystal consisting of 14 pairs of 118.8-nm-thick Ta2 O5 and 164.7-nm-thick SiO2 . Under the photonic crystal, there is an additional layer of Ta2 O5 with a thickness of 107.5 nm, cobalt and gold film both with a width of 8 nm. Table 5.2 shows that the excitation of long-range plasmon polaritons in a magnetoplasmonic heterostructure leads to a significant increase of magneto-optical sensitivity; as a result one can conclude that samples with a gold layer are appropriate for magneto-optical sensors applications. However, the structure without gold components allows to obtain resonances with higher Q-factor and provides better sensitivity for optical measurements. All previous results were obtained during experiments with air or helium gases. However, all these sensors also can be used for measurements with water chamber. Table 5.2 The excitation of long-range plasmon polaritons in magnetoplasmonic heterostructure. Fabricated structure with a gold film

Fabricated structure without gold film

Ideal structure with a gold film

Ideal structure without a gold film

Ideal structure with cobalt and gold films

Width of the plasmon resonance (deg)

0.13

0.05

0.018

0.007

0.03

Width of the TMOKE resonance (deg)

0.06

0.02

0.005

0.001

0.002

Parameter

Q-Factor

700

2200

8700

43 000

21 000

Sensitivity of the reflectance measurements, RIU−1

250

720

2300

3700

1100

Sensitivity of the TMOKE measurements, RIU−1

18

2400

5000

1500

7500

177

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

1000

Surface concentration, (mm–2)

178

100

10

1

0.1

0

50

100

150

200

250

Radius of nanoparticles (nm)

Figure 5.11 Surface concentration distribution for three types of ideal structures: red line corresponds to the one with cobalt, black line – iron garnet layer with gold cover, and blue line – iron garnet layer without gold.

For a theoretical analysis of the possible results, a calculation was performed for the available sensors. Spherical nanoparticles are deposited on them. From the parameters of nanoparticles and the studied material, the minimum surface concentration of nanoparticles, which can be measured by sensors of each type, was determined. Figure 5.11 shows the dependence of the minimum surface concentration, which can be measured experimentally, on the radius of the studied nanoparticles. All the previous data confirm the wide spectrum of sensor applications for the structures with iron garnet layers. The excitation of surface waves leads to a significant increase of the structure’s sensitivity. The potential of biosensors is rated quite high. The speed of detection, high detection thresholds, and low cost of components are factors to consider for an effective replacement for existing technologies in the long term.

5.3 Devices Assemble and Application of Iron Garnet Films for Magneto-optical Eddy Current Flaw Detection 5.3.1

Introduction

The generation of the eddy currents (ESs) and the accompanying magnetic fields (MFs) are widely used in the technology, and recently in medicine. Eddy current probes are used for the rapid analysis of cardiac activity, to register venous pulse, hyperthermia, etc. [35]. In some cases, it is necessary to know the magnetic field’s distribution of magnetic nanoparticles [36] of various organs of the human [37]. Perspective materials based on epitaxial ferrite garnet (EPFG) films can be used for

5.3 Devices Assemble and Application of Iron Garnet Films

magnetic field sensors. There are various ways of creating magnetic sensors using EPFG. In [8, 38] a high-sensitivity magnetic modulation sensor, which can compete with the SQUID, is used, for example, in magnetocardiography. On the basis of EPFG can be created magneto-optical sensors. The main advantage of such sensors is the possibility of direct observation of the topography of magnetic fields generated by magnetic objects [39–41] or eddy currents excited in the conductive objects [42–46]. In reviews [47, 48], physical principles and various applications of magneto-optical films are considered, and the properties of uniaxial and planar magneto-optical indicator films are described in detail. The authors note that in recent years, the MO imaging has turned into a very useful technology with numerous applications and, despite a number of limitations, continues to develop. This paper presents the results of studies of the application of eddy current magneto-optical imaging for the determination of various defects in magnetic and nonmagnetic objects of control. The principle of operation of the eddy current magneto-optical (MO) introscope is based on the reaction of the magnetization vector or dynamic domain structure (DDS) of the MO sensor to the distribution of magnetic fields generated by eddy currents (EC) in the test sample of conductive material. An inductor of alternating magnetic field excites eddy currents. Visualization of changes in the magnetic system of the magneto-optical sensor is provided by the Faraday effect. Defects in the test object lead to a change in the trajectory of the eddy currents and to a corresponding change in the configuration of the magnetic fields they generated. Transparent magnetic films based on BiIG with maximum values of the Faraday effect are usually used as MO sensors. The optical contrast, size, and quality of the MO image depend on such parameters as the frequency and amplitude of the EC of the exciting field, the bias field, and also the manifestation in certain cases of nonlinear properties of the dynamic domain structure [49, 50].

5.3.2

Experimental Part

The BiIG films with nominal compositions of Re3−x Bix Fe5−y Mey O12 (0 < x < 2, 0.3 < y < 0.7), where Re–Y, Sm, Tm, Gd, Lu, Ca and Me–Al, Ga, Sc, Ge, Si, and Be have been synthesized by means of liquid phase epitaxy (LPE) method on the monocrystalline gadolinium gallium garnet Gd3 Ga5 O12 substrates with orientation (111) [51–53]. The thickness of the films varied in the range from 2 to 12 μm. Depending on the requirements to the magnetic properties of the samples, the necessary chemical elements have been introduced into the composition, in order to minimize the mismatch of the lattice parameters 𝛿 = (a f − a s )/as , where af and as are the lattice parameters of the film and the substrate. The magnetization and anisotropy in films have been adjusted by replacing Fe ions with Al and Ga ions, as well as with Be–Si, Ca–Ge pairs. Films with different types of magnetic crystallographic anisotropy easy plane (EP) and easy axis (EA) have been investigated for magneto-optic eddy current (MOEC) visualization of different defects in nonmagnetic and magnetic metal material. EA films are characterized by a period of equilibrium domain structure

179

180

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

5–32 μm, effective uniaxial anisotropy field 0.5–3 kOe, saturation field H S 22–77 Oe. Out-of-plane magnetic field saturation in EP films H s was 235–328 Oe. The test samples of linear defects of the “through-slit” type with a width of 20–500 μm and defect-free welds of aluminum and stainless steel, as well as defective welds in steel pipes, made by electric and gas welding, have been investigated. The study has been carried out in the frequency range f from 8 to 80 kHz. The intensity of the alternating field of the inductor H i can vary from 2 to 400 Oe, and the bias field H_ from 0 to 40 Oe.

5.3.3

Introscope

Simplified block scheme of the experimental equipment is presented in Figure 5.12. The device consists of optical part for registration in the real-time video in reflected polarized light, changeable inducers, generator (frequency f range from 1 to 80 kHz), LED matrix, synchronizer, and MO EPFG. An alternating current I is generated in the inductor, and the domain structure (DS) of the EPFG responds to the generated field when it increases or decreases. An important aspect of the operational principle of the stroboscopic MO introscope is the correct choice of the exposition moment (strobe phase 𝜃) and duration. The video system or the illuminator (LED) must provide registration of images only at the certain moments. Pulsing mode of operation (in comparison with harmonic one) is conditioned by the simple electronic scheme realization and high efficiency of power amplifier. Triangular waveform of EC excitation is realized because the active resistance of all changeable inducers is much lower than inductive one. A programmable microcontroller is used. It generates signals for feeding the inducer and delivers a short gating pulse to the LED. The controller firmware selects the frequency of current in the inducer, time and duration of the strobe. Inducer with normal and in-plane field orientation to film surface was used. Figure 5.13 presents the picture of the eddy current excitation by a normal field inductor in the vicinity of a linear defect. It can be seen that the distribution of eddy currents changes in the vicinity of the defect. The alternating field N ∼ generates two 6 11

7

10 5

8

9 2

4

3 1

Figure 5.12 Simplified scheme of the introscope: (1) object for inspection; (2) EC inducer; (3) garnet film (MO sensor); (4) generator; (5) synchronizer unit; (6) LED matrix; (7) glass diffuser; (8) polarizer; (9) analyzer; (10) objective lens; and (11) CCD matrix.

5.3 Devices Assemble and Application of Iron Garnet Films

Figure 5.13 Separation of the eddy current contour by linear defect.

H𝛛n H~ H𝛛t

EC contours (Figure 5.13) and each of them forms EC magnetic field with normal H dn and tangential H dt components.

5.3.4

Physical Properties of MO Sensors

Figure 5.14 demonstrates equilibrium domain structure of easy axis garnet film. It is the binary mazelike domain structure with the magnetic moments in opposite directions. If normal oriented bias magnetic field N_ exists, domain structure is modified and at N_ > H S domain structure disappears. In the resulting field H eff = N_ + (H ∼ − H dn ), domain structure “reflects” topology of stray fields from the defect. The H eff magnitude depends on frequency f , inducer field amplitude and N_. The optical and magneto-optical properties of the films have been investigated in the wavelength range of 500–950 nm. The maximum values of the specific Faraday rotation at a wavelength of 520 nm for EA and EP samples are 1.6 deg/μm and 0.74 deg/μm. An important characteristic of MO sensors is the spectral dependence of the magneto-optical Q factor of the sensors. It determines both the sensitivity of the sensors and the type of the most optimal sources and receivers of optical radiation necessary for the functioning of the MO introscope. The magneto-optical quality factor of the sensors is estimated by the formula: Q = 2θf ∕ ln T

(5.11)

where 𝜃 f is the Faraday rotation value, T is the optical transmission. Figure 5.15 shows the spectra of radiation sources and magneto-optical quality factor of one of the EA MO sensors. One can see that the maximum of the Q factor is at the wavelength of 575 nm and the most optimal radiation source is warm LED, whose spectrum has the greatest overlap with the spectrum of MO of quality factor. Figure 5.14 Equilibrium domain structure in EA MO sensor at 100x optical zoom. Source: Nazar V. Lugovskoy.

181

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

100

24

Warm LED

22

Netural LED

20

Relative radiant power (%)

80 70

18

Cold LED

16

60

Halogen lamp

14

50

12 40

10

30

8

20

6

10

4

0

2

Magneto-optical quality factor

90

400 450 500 550 600 650 700 750 800 850 900 950 1000 Wavelength (nm)

Figure 5.15 The spectrum of the Q-factor for EA film and the emission spectra of various light sources.

10 3

8

2

Faraday rotation (deg)

Faraday rotation (deg)

182

1 0

–1

6 4 2 0 –2 –4 –6

–2

–8 –3

–10 –60

–40

(a)

Figure 5.16

–20

0 20 H (Oe)

40

–400

60 (b)

–200

0 H (Oe)

200

400

MO hysteresis loops at the wavelength of 640 nm for EA (a) and EP (b) films.

An important characteristic of films for operation as sensors in the MOEC of flaw detection are the field dependences of magnetization, which largely determine their sensitivity. Figure 5.16 presents similar dependences, measured at the wavelength of 640 nm, which for EA and EP films have hysteresis and hysteresis-free character, respectively. The main parameters that determine the sensitivity of the films are the normalization field H n for the EP films, the saturation field H s , and the coercive force H c for EA sensors.

5.3 Devices Assemble and Application of Iron Garnet Films

The coercive force H c and the type of domain structure are important factors in the operation of the MO introscope because they determine the remagnetization process in alternating magnetic fields. The process of visualization of the distribution of magnetic fields of eddy currents depends on the dynamics of the domain structure in EA sensors. In all EA films, a labyrinth domain structure is observed with a period from 5 to 32 μm. When using EPFG as sensors in the MOEC introscopy, certain conditions are imposed to the roughness and defectiveness of film surface. This is necessary for the rapid remagnetization of the film for EC magnetic field visualization, sinc domain walls are very sensitive to defects in the film structure and can linger on them. The average roughness of the studied films has been measured by atomic force microscopy using an NTEGRA scanning probe microscope and it does not exceed 5 nm [54]. The value of H c is determined by the defectiveness of the films and in the studied EA films will vary from 0.5 to 2.7 Oe. The advantage of EP films is the hysteresis-free nature of its magnetization reversal (Figure 5.16b). The normalization and saturation fields are determined by the values of the effective magnetic anisotropy field: a Heff = Hdem + Hu + Hk

(5.12)

where H dem = 4𝜋 ms , demagnetizing field of the sample form, H u = 2K u /M s is the field of uniaxial anisotropy, and H k = −4/3 K 1 /M s is the field of cubic magnetic anisotropy. The uniaxial anisotropy constant to the greatest extent is determined by the magnetoelastic contribution due to the magnitude and sign of the lattice parameters mismatch 𝛿. The cubic anisotropy constant in Bi-garnets is usually small, so that H k < H dem , H u . To determine the effective magnetic anisotropy fields N a eff , as well as to assess the degree of film homogeneity, the ferromagnetic resonance (FMR) method has been used. FMR spectra for the EA (a) and EP (b) films are presented in Figure 5.17. The effective magnetic fields of anisotropy N a eff are determined from FMR data on the 800 6000

EP

0 90

4000

0 90

600 400

2000 I (a.u.)

I (a.u.)

200 0 –2000

0 –200

–4000 –400 –6000 –600 –8000 –800 0

100

200

300

400

500

0

H (mT)

(a)

100

200

300

400

500

H (mT)

(b)

Figure 5.17 FMR spectra for the EA (a) and EP (b) sensors with excitation field in the normal (0) and in the plane (90) to the sample surface.

600

183

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

base of well-known Kittel formulas for the resonance fields in perpendicular and parallel configurations [55].

5.3.5

The Sensory Properties of the EA Films

5.3.5.1 The Effect of Alternating Field Amplitude

Figure 5.18 demonstrates the MOEC image of linear defects with dimension 20 μm (f = 60 kHz, H ∼ = 0–180 Oe). One can see that in comparatively low fields the defect is presented as a white zone. At fields that are higher than 80 Oe, in the center of this zone, a dark strip appears in accordance with opposite orientation of the normal component of EC field. The dimension of MOEC images increases with increasing of the field magnitude. An accuracy of the white zone determination is lower than of the dark one, contrast of which is higher. The dimension of the 20 μm defect MOEC image can be increased more than 20 times for white zone and 10 times for dark one (Figure 5.19). It means that a MOEC flaw detector can be considered as electromagnetic microscope, in which optical image can be varied by the alternating field tuning. At very high amplitudes of alternating field in the dark zone of image a “thin structure” arises due to the appearance of domains with opposite magnetization directions. This obstacle complicates the determination of real defect configuration. 5.3.5.2 The Effect of Alternating Field Frequency

Figure 5.20 demonstrate the frequency dependence of the pair of linear defects MOEC image at weak H ∼ = 60 Oe and strong H ∼ low value of alternating field. The width of the white and dark zone in MOEC image is decreased when the frequency grows.

(a)

(c)

(b)

(d)

Figure 5.18 Dependence of 20 μm slit MOEC image on the alternating field amplitude at f = 60 kHz, H_ = 0 Oe, H∼ : (a) 20, (b) 60, (c) 100, (d) 180 Oe. Source: Nazar V. Lugovskoy. White Dark

500 400

a (μm)

184

300 200 100 0 20 40 60 80 100 120 140 160 180 200

H~ (Oe)

Figure 5.19 Dependence of the width of white and dark zones of the 20 μm slit MOEC image on the alternating field amplitude H∼ at f = 60 kHz, H_ = 0 Oe.

5.3 Devices Assemble and Application of Iron Garnet Films

(a)

(b)

(c)

(d)

(b)

(c)

(d)

(A)

(a) (B)

Figure 5.20 Dependence of the MOEC image of the line defects on the frequency at weak 60 Oe (A) and strong 120 Oe (B) alternating magnetic field, f : (a) 8, (b) 15, (c) 30, (d) 60 kHz. Source: Nazar V. Lugovskoy. Figure 5.21 Dependence of the width of the white zone of MOEC image on frequency at H∼ = 60 Oe. a (μm)

2400

1800

1200

600

0.1

0.2

0.3

0.4

f –1/2

Figure 5.21 shows the frequency dependence of white zone width in the 20 μm defect MOEC image. One can see that it correlates with frequency dependence of skin-layer depth, which has the same character. The contrast of the MOEC image increases with the amplitude increasing of the alternating field; however, the fine structure of the dark zone (Figure 5.20) is more distorted with decreasing frequency. Chaotic vortex formations appear in the domain structure of the garnet film. In the beginning, they concentrate on the boundaries of the dark zone of the image, and at low frequencies – in the entire field of view. 5.3.5.3 The Effect of Bias Magnetic Field

The effect of an out-off-plane uniform external magnetic field (bias field) H_ on the EC MO image in EA sensors is shown in Figure 5.22. With the increase in the bias field, the width of the dark zone in the MO image of slit increases and the MO image of the defect disappears at large fields (Figure 5.23). The bias field improves the quality of MO visualization of linear defects in the field range H_ = 0–10 Oe. At N _ = 20 Oe, a vortex domain structure appears on the boundaries of the dark zone, which makes it difficult to determine the real width of

185

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

(a)

(b)

(c)

(d)

(e)

(f)

Figure 5.22 Dependence of MOEC image on the bias field at f = 60 kHz, H∼ = 100 Oe, H_: (a) 0, (b) 8, (c) 20, (d) 32, (e) 36, and (f) 40 Oe. Source: Nazar V. Lugovskoy. Figure 5.23 Dependence of the width of the dark zone of the MOEC image of the slit on the bias field (H_) at f = 60 kHz, H∼ = 100 Oe.

1250 1000 a (μm)

186

750 500 250 0

5

10

15

20

25

30

35

Hbias (Oe)

the MOES image. The disappearance of the MOES image of defects at H = 40 Oe is associated with saturation of the film. Such changes are possible because the bias field H_ on the one hand changes the configuration of the equilibrium domain structure in the garnet film, and, on the other hand, it shifts the working point on the static magnetization curve, relative to which the modulation of the magnetic state of the garnet film occurs. Based on these experiments, the best parameters for MOEC introscopy of linear defects in aluminum alloys EA sensors are an inductor current frequency of 60 kHz, an alternating magnetic field of 100 Oe, and a bias magnetic field of 8 Oe. 5.3.5.4 Dynamic Domains in the Garnet Film Sensor Element

To understand the behavior of dynamic domain structure in EA sensor regime of garnet film, it is necessary to investigate its behavior in such regimes but without contact with object of control. The dynamic domain structure in garnet films at low frequencies has been investigated in [47, 48].Formation of different and mostly axis-symmetrical dynamic structures, such as spiral and ring domains, bubble domain lattices depends on the frequency and amplitude of alternating fields. The processes of self-organization in domain structure take place continuously

5.3 Devices Assemble and Application of Iron Garnet Films

(a)

(b)

(c)

(d)

Figure 5.24 Dynamical domain structure in the EA garnet film in the different alternating magnetic fields at f = 15 kHz, H_ = 0 Oe, H∼ : (a) 60, (b) 100, (c) 120, and (d) 140 Oe. Source: Nazar V. Lugovskoy.

(a)

(b)

(c)

(d)

Figure 5.25 Giant dynamic domain structures in the EA garnet film at different frequencies, H∼ = 120 Oe, N _ = 0 Oe, f : (a) 8, (b) 15, (c) 30, and (d) 60 kHz. Source: Nazar V. Lugovskoy. Figure 5.26 Dependence of giant stripe domains width from the frequency at N ≅ 140 Oe, N _ = 0 Oe.

1200 1050

d (μm)

900 750 600 450 300 150 10

20

30

40

50

60

70

80

f (KHz)

at strong alternating magnetic fields, and they depend on excitation conditions, magnetic parameters of garnet films, and their defectiveness. All the processes can be manifested in the MOEC introscopy and make difficult flaw detection. In fact, at the low amplitude of alternating field, the stochastic domain structure (Figure 5.24) transforms in the spiral domain structure with spiral period about 26 μm (Figure 5.24a). With increasing of alternating field up to 120 Oe, the dynamic chaos appears (Figure 5.24c). In the field H ∼ = 140 Oe, this state is replaced by the giant stripe dynamic structure with dimensions 150 μm (Figure 5.24d). The analogous behavior is observed at all frequencies. With frequency growth, the field of domain structure stochastization increases. The images of giant domain structure in the film at different frequencies are clearly seen in Figure 5.25. The size of giant domains increases at frequency decreasing of an alternating magnetic field (Figure 5.26).

187

188

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

(a)

(b)

(e)

(c)

(f)

(d)

(g)

Figure 5.27 Effect of the alternating field (a–d) and frequency (e–g) on the detection efficiency of a 40 μm slits in Al sample by the EP MO sensor. f = 30 kHz, H∼ = 12 (a), 16 (b), 20 (c), 32 Oe (d); bottom row H∼ = 8 Oe, f = 15 (e), 30 (f), 60 kHz (g). Source: Nazar V. Lugovskoy.

Thus, the distortion of the magneto-optical eddy current introscopy image of the defect at large amplitudes of the alternating field, which grows with decreasing frequency (Figure 5.20), is associated with non-threshold dynamic structures that are observed in EA garnet films. These features of EA garnet films impose certain restrictions on their use in magneto-optical eddy current detection of defects.

5.3.6

The Sensory Properties of the EP Films

The results of visualization of slits in the Al sample by EP films differ from the results obtained by EA films (Figure 5.27). At any amplitudes of the alternating field, linear defects are displayed only by the light bands. The sensitivity of the EP sensor increases with increasing excitation field (Figure 5.27a–d) and frequency (Figure 5.27e–g), which can be seen from the increase in contrast when changing these parameters. In EP sensors, the bias magnetic field does not affect the visualization of defects, unlike in EA films, where the bias field modifies the domain structure, which is absent in EP films.

5.3.7

Applications of MOEC: Imaging of Welds

We have investigated the possibility of analyzing the quality of welds in products from magnetic and nonmagnetic metal alloys using MOEC introscopy with EA sensors. Thus, we used alternating field inductors of two configurations, which created a magnetic field either in the plane of the sample (parallel inductor) or perpendicular to it (normal inductor). 5.3.7.1

Nondefective Welds

Model nondefective welds made from both magnetic and nonmagnetic materials have been investigated. Figure 5.28 shows optical and magneto-optical images of

5.3 Devices Assemble and Application of Iron Garnet Films

(a)

(b)

(c)

(d)

(e)

(f)

Figure 5.28 Optical (a,d) and magneto-optical (b, c, e, f) images of welds: aluminum (a–c), stainless steel (d–f); orthogonal (b,e) and longitudinal (c,f) excitation of eddy currents; f = 25 kHz, H∼ = 150 Oe, N _ = 3.6 Oe. Source: Nazar V. Lugovskoy.

welds in aluminum and stainless-steel samples measured at f = 25 kHz and alternating magnetic field H ∼ = 150 Oe. It can be seen that the visually hardly visible (Figure 5.28a,e) welds are clearly visible in the MOEC images for both types of welds. In a stainless-steel sample, the sensitivity of the MO sensor is significantly higher than in aluminum samples. In such a sample, in the field of a normal inductor as in Figure 5.28e, a binary MO EC image is observed, while a dynamic vortex-like domain structure fills the entire space outside the seam. The sensor above the seam in the field H ∼ = 120 Oe is in a saturated state, which produces a clear and contrast image of the weld. In a parallel field in the MO images (Figure 5.28c,f), an “analog picture” of the weld structure is observed, especially on stainless steel, where all the inhomogeneities that are formed during the welding process and are invisible become visible. In the same figures, we see clear boundaries of the weld, represented by a narrow black and white frame of the defect. This indicates the opposite direction of the magnetic field of the eddy currents at the weld boundaries during longitudinal excitation of eddy currents. The influence of the frequency of an alternating magnetic field on the MO visualization of welds is investigated. It was found that the effectiveness of MO imaging decreases with increasing frequency (Figure 5.29) and in order to increase

(a)

(b)

(c)

(d)

Figure 5.29 Magnetooptical images of a weld in a stainless-steel sample with orthogonal excitation of eddy currents, f : (a) 8, (b) 15, (c) 30, and (d) 45 kHz. Source: Nazar V. Lugovskoy.

189

190

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

its efficiency, an increase in the amplitude of an alternating magnetic field is required. Analysis of MO images in magnetic and nonmagnetic samples shows that the magnitude of the EC magnetic fields is much larger in stainless steel samples because these samples have significantly higher values of magnetic permeability. Therefore, saturation of epitaxial ferrite-garnet films at the locations of defect mapping caused by EC magnetic fields occurs in a magnetic sample at lower values of H ∼ = 120 Oe than in a nonmagnetic sample H ∼ = 300 Oe. Such a state leads to a contrasting and clear image of defects. The use of the orthogonal bias field H_ leads to a significant decrease of these values. For example, a bias field, H_ = 24 Oe, reduces the saturation fields up to 80 Oe and 220 Oe for magnetic and nonmagnetic samples, respectively. The effect of the constant bias field and the frequency of the alternating field on the MO imaging of defects in magnetic samples is more pronounced than in the nonmagnetic ones [56].

5.3.7.2

Defective Welds

To test the operation of the MOEC method under real conditions, samples of welds of steel pipes with a diameter of 50 and 70 mm using electric and gas welding were made. MO introscopy was carried out at frequencies from 8 to 60 kHz, in alternating magnetic fields from 40 to 440 Oe. It is shown that the optimal conditions in this case are the following: f = 25 kHz, H ∼ = 80 Oe, the bias magnetic field N_ = 0 Oe. Optical and MO images obtained under these conditions are shown in Figure 5.30. MO images in Figure 5.30 are obtained with equal parameters of alternating and constant fields, the difference is only in the direction of generation of the excitation field. It is clearly seen that the MO image of the defective area contains much more information than a simple optical snapshot. The use of the MOEC method allows detecting defects that lie in the depth of the seam [56]. An interesting result was received at using of ferrite garnet film with an easy-plane anisotropy as a sensor, thus obtaining an “analog picture” of the welds on a stainless-steel sample using an inductor of a normal field.

(a)

(b)

(c)

Figure 5.30 The defects in welds of steel pipes: optical (a) and magneto-optical (b,c) images; orthogonal (b) and longitudinal (c) excitation of eddy currents. Source: Nazar V. Lugovskoy.

5.3 Devices Assemble and Application of Iron Garnet Films

150 120 90 60

2000

Hcr

δ, (MA/M)

30 Dark area

0 –30 –60

H (A/M)

1500

White area

1000

–90 –120 –150

300

0 –1.2

(a)

–4 –1.4 0 0.4 y(mm)

1.2

–2

0

2

4

y (MM) (b)

Figure 5.31 MOEC image of slit and Hz – simulation over a slit (a), the distribution of eddy current in the cross section of the Al sample with a slit (b). Source: Nazar V. Lugovskoy.

5.3.8

Simulation of EC Magnetic Fields in Samples with Defects

The distribution of eddy currents and the magnetic fields induced by them in the vicinity of a linear defect was simulated by the finite element method in the Comsol Multiphysics CAD system [53] and the direct method of integral differential equations relating the density of excited electric charges, eddy currents, and magnetic fields generated by them [56–58]. The magnitude of the magnetic field acting on the magneto-optical sensor is superposition field of the inductor and eddy current of the defect sample. The calculated distribution of the z-component of the resulting alternating magnetic field acting on the EA magneto-optical sensor was analyzed. Black and white images of defect correspond to different signs of alternating magnetic field along normal to defect (y-direction, Figure 5.31). When the amplitude of the alternating field increases, the magneto-optical images are broadened, reflecting a spatial broadening distribution of eddy current magnetic fields [57, 58]. Figure 5.32 shows the model distribution of the normal component of the EC magnetic field in the vicinity of the defect for different frequencies of eddy currents excited by the inductor. It is seen that as the frequency increases, the maximum of Hz -component of the EC field increases, and the width of the distribution decreases. For comparison with the experimental data obtained on easy-plane MO sensors, it is necessary to present the brightness of the MO image of defect (Figure 5.27) as a digital dependence. For processing and analysis of obtained experimental magneto-optical images, an image analysis program written in the MatLab package was compiled. It is based on

191

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

8000 Hz 15000 Hz 30000 Hz 60000 Hz

10 8 6

Bz (mT)

60

MO image width FWHM Bz Maximum Bz

1.6

50

1.4 40

1.2 1.0

30

0.8

20

4 2

0

12.8

25.6

38.4

51.2

64

76.8

0

89.6

10

20

30

40

50

60

Frequency (kHz)

x (mm)

(a)

10

0.6

0

Maximum Bz (mT)

Bz(x)

Width(mm)

192

(b)

Figure 5.32 The frequency dependences of the Bz -component of the EC field distribution near the defect (theory) (a), and the maximum and the width of the Bz (FWHM) distribution (theory) and the width of the EP film MO image (experiment) (b).

the translation of a graphic file into a special matrix, which contains the brightness value of all pixels of the MO image. The electrodynamic simulation of the EC magnetic fields in the sample under study adequately describes the change in the brightness of the MO of the image of the sensor obtained by the EP film with increasing amplitude and frequency of the alternating field. For example, Figure 5.32b shows the dependence of the half-width of the experimental MO image in the frequency range from 8 to 60 kHz and theoretically predicted values of this parameter found by mathematical processing. We can see that they are in good agreement. In accordance with theoretical calculations, an increase in the amplitude of the maximum with a frequency (Figure 5.32b) also leads to an increase in the contrast of the magneto-optical image in the experiment (Figure 5.32) [58]. Figure 5.33 shows the results of modeling a magnetic field above the surface of the welds. When manufacturing a welded seam, a change in the structure of the material inevitably occurs, which leads to a decrease in the conductivity of the conductor in the weld region. In calculations, the conductivity of the weld is assumed to be 1.5 × 107 S/m, which is 40% of the conductivity of pure aluminum. As can be seen, the presence of a weld with a lower conductivity than the conductivity of adjacent regions of the conductor leads to a significant increase in the contrast of the topogram of the magnetic field. The boundaries of the seam are clearly traced. It is natural to expect that the higher the difference in conductivity of the conductors, the greater contrast of the picture of the field, and, consequently, the MO image of variable part of sample will also be more contrast. The presence of ferromagnetic properties in the sample from stainless steel leads to the appearance of magnetization currents. In contrast to eddy currents, the effect of which leads to a decrease in the resulting magnetic field, the magnetization currents tend to increase this field. Therefore, in the case of a non-ferromagnetic sample, the magnetic field is always less than in the case of a ferromagnetic sample, even with a relatively low magnetic permeability, as in stainless steel. The obtained distribution of the normal component of the magnetic field acting on the MO sensor in the vicinity of the welds is in qualitative agreement with the experimental results.

Acknowledgments 1600

420 380

1

340

1200 2 H (A/m)

H (A/m)

300 260 220

800

1 180 140 100

–10

(a)

400 2 0 –6

–2

0 2 y (mm)

6

–10

10

(b)

–6

–2

0 2

6

10

y (mm)

Figure 5.33 Distribution of the normal component of the EC magnetic fields above the sample with a welded seam: (a) non-ferromagnetic sample without (1) and with weld (2); (b) non-ferromagnetic (1) and ferromagnetic (2) samples with welds.

5.4 Conclusions and Perspectives Multicomponent films of Bi garnets with easy plane and easy axis magnetic anisotropy for eddy current magneto-optical flaw detection were synthesized by the LPE method. The spectral dependences of the optical transmittance, the Faraday effects and the magneto-optical Q factor were measured. The magnetic characteristics of the films, the saturation field and the coercive force for films with easy axis and the normalization field for films with easy plane were determined by the magneto-optical method. The effective magnetic anisotropy fields were determined by the FMR method. A comparative analysis was performed regarding the sensory effectiveness of the easy plane and the easy axis films in the detection of various defects by eddy current magneto-optical introscopy in a wide range of frequencies, as well as alternating and bias magnetic fields. Possible research topics and applications can be discussed freely as one unique section.

Acknowledgments This work is financially supported by the Russian Foundation for Basic Research, projects N 18-52-80038 and 18-29-02120, and the NSFC-BRICS STI Framework Program (No. 51861145309). The authors from V.I. Vernadsky Crimean Federal University TVM and ANSh acknowledge support by grant of Russian Science Foundation (project no. 19-72-20154) for modelling of one-dimensional magnetophotonic crystals with bi-layers.

193

194

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

List of Abbreviation and Symbol SQUID OPM BiIG MCG EMF LMPIE TMOKE ES

MF EPFG MO DDS LPE EP EA MOEC DS FMR

Superconducting quantum interferometer magnetometer optically pumped magnetometer bismuth-substituted iron garnet magnetocardiography electromagnetic filter longitudinal magnetophotonic intensity effect transverse MO Kerr effect eddy current magnetic field epitaxial ferrite garnet films magneto-optical dynamic domain structure liquid phase epitaxy easy plane easy axis magneto-optic eddy current domain structure ferromagnetic resonance

References 1 Korepanov, V. and Marusenkov, A. (2012). Flux-gate magnetometers design peculiarities. Surveys in geophysics 33 (5): 1059–1079. 2 Drung, D., Abmann, C., Beyer, J. et al. (2007). Highly sensitive and easy-to-use SQUID sensors. IEEE Transactions on Applied Superconductivity 17 (2): 699–704. 3 Budker, D. and Romalis, M. (2007). Optical magnetometry. Nature physics 3 (4): 227. 4 Aschenbrenner, H.A.N.S. (1936). Eine Anordnung zur Registrierung rascher magnetischer Storungen Hochfrequenztechnik und Elektroakustik. Band 47: 177–181. 5 Kolachevskii, N.N. (1985). Fluctuation Phenomena in Ferromagnetic Materials. Moscow: Nauka. 6 Petta, J.R., Weissman, M.B., and O’brien, K.P. (1996). Multiple magnetization paths in Barkhausen noise. Physical Review E 54 (2): R1029. 7 Deak, J., Miklich, A.H., Slonczewski, J. et al. (1996). A low-noise single-domain fluxgate sensor. Applied physics letters 69 (8): 1157–1159. 8 Vetoshko, P.M., Valeiko, M.V., and Nikitin, P.I. (2003). Epitaxial yttrium iron garnet film as an active medium of an even-harmonic magnetic field transducer. Sensors and Actuators A: Physical 106 (1–3): 270–273. 9 Syvorotka, I.I., Vetoshko, P.M., Skidanov, V.A. et al. (2015). In-plane transverse susceptibility of (111)-oriented iron garnet films. IEEE Transactions on Magnetics 51 (1): 1–3.

References

10 Vetoshko, P., Skidanov, V., and Stempkovskiy, A. (2013). Magnetization distribution near edge of YIG film core in fluxgate magnetometer. Sensor Letters 11 (1): 59–61. 11 Vetoshko, P.M., Gusev, N.A., Chepurnova, D.A. et al. (2016). Flux-gate magnetic field sensor based on yttrium iron garnet films for magnetocardiography investigations. Technical Physics Letters 42 (8): 860–864. 12 Van der Ziel, A. (1954). Noise. Englewood Cliffs: Prentice-Hall. 13 Kwong, J.S.W., Leithäuser, B., Park, J.-W. et al. (2013). Diagnostic value of magnetocardiography in coronary artery disease and cardiac arrhythmias: a review of clinical data. International journal of cardiology 167 (5): 1835–1842. 14 Rogachev, A.E., Vetoshko, P.M., Gusev, N.A. et al. (2016). Vector magneto-optical sensor based on transparent magnetic films with cubic crystallographic symmetry. Appl. Phys. Lett. 109 (16): 162403. ˇ 15 Višnovský, Š., Lišková-Jakubisová, E., Harward, I., and Celinski, Z. (2018). Vector MO magnetometry for mapping microwave currents. AIP Advances 8 (5): 056642. 16 Arakelyan, S., Galstyan, O., Lee, H. et al. (2016). Direct current imaging using a magneto-optical sensor. Sensors and Actuators A: Physical 238: 397–401. 17 Zhang, Z., Guo, T., Zhang, X. et al. (2016). Plasmonic fiber-optic vector magnetometer. Applied Physics Letters 108 (10): 101105. 18 Deng, M., Huang, C., Liu, D. et al. (2015). All fiber magnetic field sensor with Ferrofluid-filled tapered microstructured optical fiber interferometer. Optics express 23 (16): 20668–20674. 19 Vetoshko, P.M., Volkovoy, V.B., Zalogin, V.N. et al. (1991). Measuring low alternating magnetic fields by means of Bi-containing rare-earth ferrite-garnet films with planar anisotropy. Journal of applied physics 70 (10): 6298–6300. 20 Belotelov, V.I., Kreilkamp, L.E., Akimov, I.A. et al. (2013). Plasmon-mediated magneto-optical transparency. Nature Communications 4: 2128. 21 Belotelov, V.I., Bykov, D.A., Doskolovich, L.L. et al. (2009). Giant magnetooptical orientational effect in plasmonic heterostructures. Optics Letters 34: 398–400. 22 Loughran, T.H.J., Keatley, P.S., Hendry, E. et al. (2018). Enhancing the magneto-optical Kerr effect through the use of a plasmonic antenna. Optics express 26 (4): 4738–4750. 23 Knyazev, G.A., Kapralov, P.O., Gusev, N.A. et al. (2018). Magnetoplasmonic crystals for highly sensitive magnetometry. ACS Photonics 5 (12): 4951–4959. 24 Koch, R.H., Deak, J.G., and Grinstein, G. (1999). Fundamental limits to magnetic-field sensitivity of flux-gate magnetic-field sensors. Applied Physics Letters 75 (24): 3862–3864. 25 Homola, J. (2008). Surface plasmon resonance sensors for detection of chemical and biological species. Chem. Rev. 108: 462. 26 Anker, J.N., Hall, W.P., Lyandres, O. et al. (2008). Biosensing with plasmonic nanosensors. Nat. Mater. 7: 442. 27 Slavik, R. and Homola, J. (2007). Ultrahigh resolution long range surface plasmon-based sensor. Sens. Actuators B: Chem. 123: 10. 28 Konopsky, V.N. (2010). Plasmon-polariton waves in nanofilms on one-dimensional photonic crystal surfaces. New J. Phys. 12: 093006.

195

196

5 Bismuth-substituted Iron Garnet Films for Magnetophotonics: Part B – Devices and Applications

29 Maccaferri, N., Gregorczyk, K.E., de Oliveira, T.V.A.G. et al. (2015). Ultrasensitive and label-free molecular-level detection enabled by light phase control in magnetoplasmonic nanoantennas. Nat. Commun. 6: 6150. 30 Belotelov, V.I., Kreilkamp, L.E., and Kalish, A.N. (2014). Magnetophotonic intensity effects in hybrid metal-dielectric structures. Phys. Rev. B 89: 045118. 31 Sukhorukov, A.P., Ignatyeva, D.O., and Kalish, A.N. (2011). Terahertz and infrared surface wave beams and pulses on gyrotropic, nonlinear and metamaterial interfaces. J. Infrared Millim. Terahertz Waves 32: 1223. 32 Kalish, A.N., Ignatyeva, D.O., Belotelov, V.I. et al. (2014). Transformation of mode polarization in gyrotropic plasmonic waveguides. Laser Phys. 24: 094006. 33 Ignatyeva, D.O., Kapralov, P.O., Knyazev, G.A. et al. (2016). High-Q surface modes in photonic crystal/iron garnet film heterostructures for sensor applications. JETP letters 104 (10): 679–684. 34 Ignatyeva, D.O., Knyazev, G.A., Kapralov, P.O. et al. (2016). Magneto-optical plasmonic heterostructure with ultranarrow resonance for sensing applications. Scientific reports 6: 28077. 35 Andrä, W. and Nowak, H. (2007). Magnetism in Medicine, 2e, 655. Weinheim, Germany: Wiley–VCH Verlag GmbH. 36 Nikitin, P.I., Vetoshko, P.M., and Ksenevich, T.I. (2007). Magnetic Immunoassays. Sensor Letters 5: 296–299. 37 Malmivuo, J. and Plonsey, R. (1995). Bioelectromagnetism: Principles and Applications of Bioelectric and Biomagnetic Fields. New York: Oxford University Press. 38 Vetoshko, P.M., Zvezdin, A.K., Skidanov, V.A. et al. (2015). Impact of profile of disk magnetic element on the saturation field and noise magnetomodulation sensor of magnetic field. Technical Physics Letters 41 (9): 103–110. 39 Vishnevskii, V. and Berzhansky, V. (2009). Visualizer of magnetic protection of securities and banknotes. Sensor Letters 7: 282. 40 Joubert, P.-Y. and Pinassaud, J. (2006). Linear magneto-optic imager for nondestructive evaluation. Sens. Actuators, A 129: 126–130. 41 Cheng, I.H., Liu, X.M., Bai, L.B., and Tian, G.I. (2013). Magneto-optic microscope technology for nondestructive testing. Appl. Mech. Mater. 330: 291–298. 42 Pinassaud, J., Joubert, P.-I., and Lemistre, M. (2005). Quantitative magneto-optical imager for nondestructive evaluation. Proc. SPIE 5768: 196–203. 43 Radtke, U., Zielke, R., and Rademacher, H. (2001). Application of magneto-optical method for real-time visualization of eddy currents with high spatial resolution for nondestructive testing. Opt. Laser. Eng. 36: 251–268. 44 Vishnevskii, V., Berzhansky, V., Mikhailov, V. et al. (2013). Magneto-optic eddy current introscopy based on garnet films. Solid State Phenomena 200: 245–249. 45 Vishnevskii, V.G., Berzhansky, V.N., Lugovskoy, N.V. et al. (2014). Features of linear defects eddy current images obtained with use of ferrite garnet films. Solid State Phenomena 230: 273–278. 46 Lugovskoy, N., Vishnevskii, V., Berzhansky, V. et al. (2014). Features of magneto-optic eddy current images obtained with use of garnet films. In: International Conference on Oxide Materials for Electronic Engineering OMEE-2014, 234–235. Lviv Polytechnic National University.

References

47 Grechishkin, R., Chigirinsky, S., Gusev, M. et al. (2008). Magnetic imaging films. In: Magnetic Nanostructures in Modern Technology (eds. B. Azzerboni et al.), 195–224. Springer. 48 Grechishkin, R., Kustov, M., Ilyashenko, S. et al. (2016). Magneto-optical imaging and analysis of magnetic field micro-distributions with the aid of biased indicator films. J. Appl. Phys. 120 (174502). 49 Kandaurova, G.S. (2002). New phenomena in the low-frequency dynamics of magnetic domain ensembles. PHYS-USP 45 (10): 1051–1072. 50 Kandaurova, G. and Svidersky, A. (1990). Self-organization processes in multidomain magnetic media and the formation of stable dynamic structures. JETPh. 97: 1218–1229. 51 Berzhansky, V.N., Karavainikov, A.V., Mikhailova, T.V. et al. (2017). Nano- and micro-scale bi-substituted iron garnets films for photonics and magneto-optic eddy current defectoscopy. Journal of Magnetism and Magnetic Materials: 440, 175–178. 52 Prokopov, A.R., Vetoshko, P.M., Shumilov, A.G. et al. (2016). Epitaxial Bi-Gd-Sc iron-garnet films for magnetophotonic applications. Journal of Alloys and Compounds 671: 403–407. 53 Lugovskoy, N., Berzhansky, V., Glechik, D., and Prokopov, A. (2018). Epitaxial film of garnet ferrite with anisotropy “easy plane” for magneto-optical eddy current flaw detection. J. Phys.: Conf. Ser. 1124: 051063, 4 p. 54 Lugovskoy, N.V., Berzhansky, V.N., E. Yu. Semuk et al. (2019). Magneto-optical properties of easy-plane and easy-axis garnet ferrite films for eddy current testing. IOP Journal of Physics: Conference Series in print. 55 A.G. Gurevich, G.A. Melkov Magnetization Oscillations and Waves Magnetization Oscillations and Waves: 1996, CRC-Press P. 464 56 Berzhansky, V.N., Lugovskoy, N.V., Filippov, D.M., et al. (2017). Investigation of welds by the method of magneto-optical eddy current flaw detection. EPJ Web Conference Moscow International Symposium on Magnetism (MISM 2017) V 185. 57 Filippov, D.M., Kozik, G.P., A.V. Fursenko, and V.N. Fedorovsky (2016). The Secondary Sources Method Analysis and Experimental Modeling of the Permanent Magnet Eddy Currents, International Conference on Industrial Engineering, Applications and Manufacturing, ICIEAM, 2 58 Berzhansky, V.N., Filippov, D.M., and Nazar, V. (2016). Lugovskoy/magneto-optical visualization of Eddy current magnetic fields. Physics Procedia C 82: 27–31.

197

199

6 MEMS, NEMS, AEMS, and Quantum Films for the Next Generation of Computing and Information Technology Haishuai Chai 1 , Junmei Wang 1 , and Yujun Song 1,2 1 University of Science and Technology Beijing, Center for Modern Physics Technology, Applied Physics Department, School of Mathematics and Physics, 30 Xueyuan Road, Beijing 100083, China 2 Zhejiang Key Laboratory for Pulsed Power Technology Translational Medicine, Hangzhou Ruidi Biotechnology Company, Hangzhou 310000, China

6.1 Introduction Microelectromechanical systems (MEMS) adopt silicon micromachining technology to manufacture thousands of microelectronic devices (the metal oxide semiconductor field-effect transistor, MOSFET) on a single silicon wafer, which addresses the bottleneck of mass production of traditional processes, thereby reducing costs and improving efficiency in the integrated microelectronic circuits. MEMS conjugate thermal, optical, magnetic, chemical, biological, and other functional structures and devices onto microchips through microelectronics and other micromachining processes and through integration with circuits and constructing complex microsystems. Due to the high degree of cross-linking and permeability of MEMS, research and development efforts are diverse. As great progress has been achieved in microfabrication and microprocessing technologies, the size of electronic devices has been scaled down to submicrometers. This is especially embodied in the Si metal oxide semiconductor (MOS)-based devices, e.g. chips and memories, obeying the well-known Moore’s law. To minimize the short-channel effect, the thickness of dielectric oxide in a MOS device is approaching a few nanometers [1, 2]. Recently, basic research in MEMS based on nanotechnology has been scaled down to the nanoscale or even atom level in one or more dimensions, leading to nanoelectromechanical systems (NEMS, or AEMS). Usually, thin films are modeled by a rectangular deep potential well. However, in recent years, the profile of the wells, for example, parabolic or other forms, has been considered. Due to the technology of film growth, their surfaces may not be the same. In the case of classical films, the difference in surfaces could be considered, most likely, by the introduction of different surface scattering parameters or by different band bends [3, 4]. In the case of quantum films, a solution of the Poisson equation with given surface charges, which is an asymmetric potential well, could determine the difference in surfaces. This will affect the energy spectrum of the carriers [4]. Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications, First Edition. Edited by Yujun Song. © 2021 WILEY-VCH GmbH. Published 2021 by WILEY-VCH GmbH.

200

6 MEMS, NEMS, AEMS, and Quantum Films

6.2 Typical Fabrication Methods for MEMS, NEMS, and AEMS 6.2.1

Fabrication of Microstructures

Microstructures are usually fabricated by the LIGA process. LIGA process was first developed at the Institute for Microstructure Technology (IMT) (at the Research Center Karlsruhe) in the early eighties under the leadership of Dr. W. Ehrfeld [5]. The acronym LIGA comes from the German name for the process (Lithographie, Galvanoformung, Abformung). LIGA uses lithography, electroplating, and molding processes to produce microstructures. It is capable of creating very finely defined microstructures of up to 1000 μm or higher. Using this process, it is possible to mass-produce at low-cost microstructures with high aspect ratio. By making use of the full potential of the three basic steps in the LIGA process, metal, polymeric, and even ceramic microstructures can be produced. Danny Banks has summarized [5] the detailed process of the LIGA process. As briefly described in Figure 6.1, the X-ray or other radiation sources (UV, excimer lasers, synchrotron-based radiation) are represented as an appropriately designed mask onto a thick photoresist layer (sensitive to the type of radiation source), which is coated on a desired conductive substrate. This resist is then developed to form a “negative” feature. In the process, as originally developed, a special kind of photolithography using X-rays (X-ray lithography) is used to produce patterns in very thick layers of photoresist. The X-rays from a synchrotron source are shone through a special mask onto a thick photoresist layer (sensitive to X-rays) that covers a conductive substrate (Figure 6.1a). This resist is then developed (Figure 6.1b). The pattern formed is then electroplated with metal (Figure 6.1c). The metal structures produced can be the final product; however, it is common to produce a metal mold (Figure 6.1d). This X-Rays

(a)

(c)

(e)

(b)

(d)

(f)

Figure 6.1

The typical steps of the LIGA process.

6.2 Typical Fabrication Methods for MEMS, NEMS, and AEMS

mold can then be filled with a suitable material, such as a plastic (Figure 6.1e), to produce the finished product in that material (Figure 6.1f). As the synchrotron source makes LIGA expensive, alternatives are being developed. These include high-voltage electron beam lithography that can be used to produce structures of the order of 100 high and excimer lasers capable of producing structures of up to several hundred microns high. Electroplating is not limited to use with the LIGA process, but may be combined with other processes and more conventional photolithography to produce microstructures. The related fabrication techniques or the MEMS were further developed, initially, for microelectronics applications to construct sensors and actuators for the above hardware for computers and information technology [5, 6]. Based on the LIGA process and the advances in the field of Si-based semiconductor complementary metal oxide semiconductor (CMOS) or field-effect transistor (FET) thin films, design and fabrication techniques for microstructures and nanostructures have been further developed for miniaturized sensing and reaction systems of biological and chemical processes together with those for computers and information technology, as well as those microfluidic devices [6–8]. In the past decades, the development of micro/nanofabrication technologies and the corresponding design strategies have given an enormous impetus also to the progress in the field of fabrication of varieties of structures and micro/nanoelectronic devices [7–10]. Different fabrication processes were developed for the construction of microstructures based on a variety of materials and applications, including most commonly metals, ceramics, polymers, glass, silicon, stainless steel, and their composites [11]. The fabrication methods can be broadly classified as mechanical machining, LIGA (lithography, electroplating, and molding; using either UV or X-ray radiation) (Figure 6.1), micro molding, embossing, chemical etching, electroplating, laser ablation, E-beam lithography in addition to original classical bulk machining, deep reactive ion etching (DRIE), etc. [5, 6]. Microstructural thin films, reported to date, have been fabricated by “standard” microfabrication techniques using borosilicate glass or polymers as substrates that have been used in the fabrication of multilevel microstructures for advanced microelectronics devices. Figure 6.2 gives one typical process to fabricate multilevel microstructures using SU-8 as construction materials on pre-micromachined poly-ether-ether-ketone (PEEK) substrates [11, 12]. Photolithography and wet etching process are usually used to fabricate Si or glass-based microelectronic devices [5]. The etch rates for micromachining using different etchers for various materials are summarized in reference [13]. Typically, a thin layer of metal, such as chromium, is deposited on the surface of a glass substrate [5]. A layer of positive photoresist is then spin coated on top of the chromium to a depth of 0.5–2.0 μm. The pattern of the required network of interconnecting channels is transferred to the photoresist layer using photolithography. After exposure, the photoresist is developed to open the areas of glass to be etched. The plate is then allowed to dry to ensure complete evaporation of volatiles before performing the chemical etching. The channels are then etched using, for example, a mixture of

201

202

6 MEMS, NEMS, AEMS, and Quantum Films Holesϕ 2 mm

PEEK

SU-8

Thin SU-8 layer

Thick SU-8 layer

(3)

(4)

Threads (1)

(2)

Sealed microfluidics Embedded structure SSPT

Sacrificial PI layer (6)

(5)

Enforced layer

Semisolid SU-8 layer

Embedded substrate (7)

Microfluidics with enforced substrate (8)

Binding the second channel with the first Forming the second microstructure layer by solid transfer process after aligning Forming the third microstructure layer with aligning marks layer with aligning marks (9) Sealed microfluidics with 3 layers of 3D structures by aligning (12)

(10) Thousands of layers of microstructures with connectors (13)

(11) Sealed microfluidics with desired layers of 3D structures (e.g. 3 layer) by SSPT (14)

Figure 6.2 Process for the fabrication of multilevel microstructures using UV–LIGA process and SU-8 (an epoxy photonic resin) as construction material on premachined PEEK substrates. (1) Connection point between micromachining and external macro equipment. (2) Seal joint points with su-8 photosensitive resin and mask to form uncleaned openings. (3) Rotate and apply a thin layer of su-8 and mask to form an uncleaned opening. (4) The su-8 layer is rotationally coated, and a mask is used to form the microreactor body. (5) Flush the unexploded su-8 to form an open microchannel. (6) Preparation of semisolid su-8 sealing membrane. (7) The semisolid su-8 sealed membrane seals the microchannel, forming a semi-open microfluidic reactor. (8) The microfluidic reactor is strengthened with an enhancement layer. (9) Prepare the second microchannel layer. (10) Bind the second microchannel to the first. (11) Preparation of the third microchannel layer. (12) Bind the third microchannel to the second microchannel. (13) Multilayer microchannel system required for replication. (14) Use SSPT technology to seal the entire microchannel system [11, 12]. Source: Modified from Song [11].

1% HF and 5% NH4 F in water at 65 ∘ C, resulting in an etch rate of 0.3–0.5 μm/min. A glass block, with predrilled holes to act as reservoirs and if necessary electrode supports, is aligned with the channel geometry and thermally bonded to the glass base plate, producing a glass-based microfluidic reactor. In addition to the specific microfabrication processes described above, hybrid fabrication techniques use a combination of microfabrication processes, such as low-pressure chemical vapor deposition (LPCVD), buffered oxide etch (BOE), chemical etching, dry etching and sputter, producing multilevel microstructures with heating parts, sensing parts, control parts, and thin catalyst membranes for reaction that can be all fabricated in wafers or other high-temperature stable substrates (e.g. Si) for special applications [14–17]. Ye et al. described the fabrication of a typical palladium membrane microreactor using this approach as one MEMS example for materials synthesis [14]. The unique feature of the hybrid fabrication process, where typically microreactors are fabricated using Si wafer, glass, or other ceramics, is that they can be used at high temperatures and are flexible to incorporate different types of thin membranes for catalytic applications. Such

6.2 Typical Fabrication Methods for MEMS, NEMS, and AEMS

microreactors have been successfully developed to carry out gas or liquid-phase oxidation or partial oxidation of organic compounds by depositing a thin catalyst membrane in the channel walls [17–20].

6.2.2 Fabrication Process of Complementary Metal Oxide Semiconductor (CMOS) Mizuki Ono et al. described a complementary metal oxide semiconductor–fieldeffect-transistor (CMOSFET)-compatible tip fabrication process. The process flow for cantilever fabrication is schematically shown in Figure 6.3 [22]. The starting material is a p-type silicon wafer with a diameter of 100 mm. After the company Austrian Mikro Systeme Int. used a traditional CMOS process to manufacture piezoresistive and on-chip circuit components, the film has been released from the back of the wafer with 27 wt% KOH at 90 ∘ C [23]. Anisotropic corrosion is stopped at the p2n junction between the n-well and the p-type substrate [24] using electrochemical corrosion resistance technology. Therefore, the obtained thin film is composed of CMOS n-holes. At the tip of this area, a deeper n-well CMOS process is used. The underside of the cantilever is uneven; however, extra material is consumed at the tip. Silicon dioxide is then deposited on both sides of the wafer. Plasma-enhanced chemical vapor deposition (PECVD) technology is used to mix silane (SiH4 ) and nitrous oxide (N2 O) at 3001 ∘ C. Then, a photolithography process is performed to determine a mask for needle tip manufacturing. The silica on the front is a reactive ion etching (RIE) pattern with CF4 (carbon tetrafluoride) and CHF3 (trifluoromethane). This step can also be completed by a wet etching process, for example, a mixture of NH4 F, CH3 COOH, and HO(CH2 )2 OH. After removing the photoresist with oxygen plasma, the tip is prepared at a temperature of 301 ∘ C with 25% by weight of tetramethylazanium hydroxide (TMAH) without any stirring. The wafers are immersed in diluted hydrofluoric acid (HF) to remove the natural oxide film prior to tip manufacturing. The etching depth in the silicon [1] direction is 3 mm. The total corrosion time is 2.5 hours. After the preparation of the tip is completed, a mixture of NH4 F, CH3 COOH, and HO(CH2 )2 OH is used to remove silicon dioxide on both sides of the wafer. As mentioned earlier, neither silicon nor aluminum is attacked during this process. Photolithography is performed to determine the shape of the cantilever: Using a patterned photoresist layer as a mask, silicon is etched in an RIE process using a mixture of sulfur hexafluoride (SF6 ) and CHF3 to release the cantilever beam. Finally, the photoresist is removed using an oxygen plasma.

6.2.3

Fabrication Process of Field Emission Transistors (FET)

In order to fabricate 100 μm channels (Figure 6.3, [21], aluminum electrodes were deposited on the TiO2 nanofibers with thermal evaporation of a mask after being calcined. After that, dipped a bit of polyvinylpyrrolidone (PVP) aqueous solution onto the two sides of SiO2 substrates, which aims at connecting the top-gate electrode. At last, the top-gate electrode was made of a Si wafer, which was highly doped,

203

6 MEMS, NEMS, AEMS, and Quantum Films

100 nm

AI

Si SiNx

TiO2 AI

SiO2

Si

AI

AI

SiO2

204

(AIt + A)

Figure 6.3 Schematic diagram of a single TiO2 nanofiber nonvolatile memory device based on a top- gate field effect transistor and SEM images and photos of TiO2 nanofiber. Source: Zhang et al. [21]. © 2019 Elsevier.

coated with a 1.6-μm-thick SiO2 dielectric. The gap between mechanically contacted top-gate electrodes is automatically filled with dry air (5–20% RH), which is used in conjunction with SiO2 as the gate dielectric.

6.2.4 Giant Magnetoresistance (GMR) Sensor and Its Fabrication Method Giant magnetoresistance (GMR) refers to a very large change in resistance in ultra-thin magnetic multilayer films (Figure 6.4). The basic GMR material structure consists of a fixed layer and a free layer; the free layer is affected by an external magnetic field. A suitable external magnetic field has a range larger than the saturated magnetic field of the free layer and smaller than the stable magnetic field of the fixed layer. The free layer is magnetized with the change of the magnetic field when the free layer rotates. In the case of fixed reference layer magnetization and free layer synchronous magnetization, the magnetic resistance is a simple cosine function of the angle of the rotor relative to the fixed sensor. The resistance of a spin value is related to the angle between the free and the pinned layer magnetizations in R∕Rp = 1 + 1∕2GMR(1 − c)

(6.1)

where the Rp is the minimum of the resistance value when the two magnetizations are parallel. At last, the GMR is the biggest percentage magnetoresistance. The finite element modeling method divides the solution domain into smaller regions. The program uses Maxwell’s equations as the basis for electromagnetic field analysis. In magnetostatics, the unknowns (degrees of freedom) are usually magnetic vector potentials that can be approximated by polynomial shape functions. Other magnetic field quantities, such as magnetic flux density, magnetic field strength, current density, energy, force, loss, inductance, and capacitance, all come from degrees of freedom [9]. The size of the element must be small enough to provide sufficient precision [10].

6.3 From MEMS to NEMS and then to Quantum Films and AEMS Spin up

Spin down

Spin up

Spin down

Ta gap Antiterromagnet

FeMn pinning layer Co pinned layer Cu spacer NiFe free layer Ta buffer Si substrate

Ferromagnet

Figure 6.4 The GMR effectively measures the difference in angle between the two magnetizations in the magnetic layers. Small angles (parallel alignment) give a low resistance; large angles (anti-parallel alignment) give a higher resistance.

The manipulator is required to finish many complex works, in order to obtain an accurate angle signal, thus, making the angle sensor important in the sensor system. The GMR sensor is ideal for systems that are highly integrated (Figure 6.5) [25].

6.3 From MEMS to NEMS and then to Quantum Films and AEMS for the Next Generation of Information Technology 6.3.1

The Trend of Microsystem Integration Technology

Microsystem technology that combines microelectronics, microelectromechanical, and light electricity technique, through system architecture and software algorithm, and the micro sensor, micro control, micro actuator, micro energy, and various interfaces form integrated soft hardware such as multifunctional integration, micro/nanomanufacturing and micro-integration technique that was adopted to realize the system structure of micro/nanoscale; it is recognized as one of the revolutionary technology of the twenty-first century. Microsystems have the advantages of high integration, miniaturization, low power consumption, high reliability, and high efficiency. New materials, new methods, new processes, and other technological changes in microsystems will certainly disrupt the development and manufacture of dual-use systems. China’s microsystem technology has made some progress. application-specific integrated circuit (ASIC) devices have been tested in orbit and started to be applied in space, but they need to be improved at the manufacturing process level. At this stage, China will “manufacture its own micro-devices” as the development goal

205

206

6 MEMS, NEMS, AEMS, and Quantum Films

0.5 mm Ø6 mm

0.7 mm

Ø3 mm

Live shaft Bearing Ø4.5 mm GMR sensor chip PCB

Figure 6.5 An example of the GMR sensor in 5-finger dexterous. Source: Lan et al. [25]. © 2009 Elsevier.

of China’s microsystems and will self-innovate while using existing resources and learning from foreign advanced technology experience. In recent years, many new micro/nanoprocessing technologies, new packaging technologies, and new materials have been deeply studied in the domestic technology platform. Throughout the 5th International Symposium on Microsystems and Nano-Engineering, the development trends of domestic microsystems are the following: (i) The reliability and stability of microsystems have been greatly improved. (ii) Research and application prospects in physical and medical integration are broad. (iii) The integration of micro/nanodevices and processing technology development has been put on the agenda. In recent years, China has been gradually increasing its support for microsystem technology through establishing policies and various conditions; and new breakthroughs in microelectronics technology has also been made: ultra-deep submicron integration technology has reached the international advanced level – chip design. For example, integrated circuits have grown significantly in the fields of high-end IC cards for digital TVs, media signal processing for mobile phones above 5G, and information security, and we have achieved major breakthroughs in the development and industrialization of core chips designed for chip products with independent intellectual property (the design level of integrated circuits has reached 0.13 μ). With the strong support of domestic policies, China is gradually relying on technological innovation to replace from low end to high end. Water silicon-based microelectronics technology continues to narrow the gap with foreign countries.

6.3 From MEMS to NEMS and then to Quantum Films and AEMS

6.3.2

The Development Trend of Microsystem Packaging Technology

In 2012, China has a research team for electronic packaging technology. The 25 domestic electronics packaging industry chain-related units have established the “Integrated Circuits and Tests Industry Chain Technology Innovation Alliance” to establish a high-density IC packaging technology engineering laboratory, mainly researching foreign developed countries. The monopoly of packaging technology, the progress of the application of packaging and testing, has played an effective role in the electronic packaging industry chain [26]. In recent years, the international competitiveness of China’s electronic packaging technology has gradually improved, mainly in the following aspects: new packaging technologies such as chip-scale package (CSP), multiple-chip package (MCP), and ball grid array (BGA) have been applied in some electronic packaging production lines; TSV silicon channel, SiP RF, and 50 three-dimensional stacking and packaging technologies for ultrathin chips of μm and below have been widely used; and electronic packaging technologies such as supervised primitive fitting network (SPFN), management information system (MIS), and FBP have also obtained independent intellectual property rights. In addition, domestic innovation in MIS flip-chip packaging technology can effectively reduce the cost of packaging and technically achieve good support for the current mainstream packaging technology [26]. For monolithic microwave-integrated circuit (MMIC), foreign MMIC chips are developing toward terahertz, while packaging in the direction of 3D-MMIC. The spectral range of this product is from microwave to millimeter wave. In semiconductor materials, the emergence of wide bandgap semiconductor provides a solid foundation for the development of microsystems. The second generation of semiconductors represented by GaAs has matured in foreign countries, and the third generation of semiconductor MMIC represented by Gan is booming. After decades of development, the three-dimensional MMIC technology has evolved from a single device to an integrated transceiver chip. The multifunctional multiboard multistage amplifier is integrated on one chip. While improving monolithic performance, it reduces the design and manufacture of space structure, only one chip built-in tertiary amplifier, gaining up to 22 dB.

6.3.3

Challenges in the Development of Microsystems Technology

In the past 10 years, as the characteristic size of transistors has decreased, quantum efficiency should become more and more significant, Moore’s law gradually fails, and the growth rate of transistors tends to be linear. With the introduction of new principle device materials and processing technologies, it is still feasible to further reduce the feature size. For many MEMS devices, controlling the level of strain in the deposited film is critical. A film structure made of polysilicon, a common MEMS mechanical material, requires a tensile film strain of less than −0.001; otherwise, the film will break. Many MEMS sensors have been fabricated that use many different methods to convert physical changes into electrical signals. The most common method of

207

208

6 MEMS, NEMS, AEMS, and Quantum Films

Glass

Top fixed electrode

Figure 6.6 Schematic illustration of a sensor designed to correlate measured changes in capacitance with pressure changes. Source: Kordi´c [29]; Baltes and Popovic [30].

Movable bottom electrode Silicon Flexible membrane

detecting motion involves piezoresistance. Other common transducer methods for mechanical deflection include piezoelectric effects, capacitance changes, and magnetic effects (such as the Hall effect). The piezoelectric effect [27, 28] is a phenomenon via a force to certain crystalline materials that causes an electrical charge to be generated on the surface of the crystal. The amount of the charge is directly related to the applied force. The application of a stress causes deformation, which created electric dipoles. These electric dipoles induce surface charges on the crystal that in turn create an electric field [26, 29]. The electric field induced by surface charge cancels the stress-induced electric field from dipoles. For example, a sensor based on the Hall effect was fabricated (Figure 6.6) [29–31]. These sensors have been used for the detection of electric current, proximity detection, and position of rotating shafts. Piezoresistivity is a property of materials that describes the change in electrical resistance as a function of mechanical stress applied to the material [32]. The discovery of piezoresistive silicon is critical to the development of silicon pressure sensors. The theoretical interpretation of piezoresistance is based on holes as a function of lattice orientation. For MEMS devices made with (100) silicon wafers, the most important parameter is the doping type. A detailed description of the piezoresistive can be found in the scientific file [33, 34]. Stephen A. Campbell shows a bar of single-crystal silicon subjected to a tensile force F, with a current flowing in a direction perpendicular to F. R denotes the resistance of the silicon bar. The force F generates a stress 𝜎 in the bar and a resistance change due to the piezoresistivity effect. The resistance change ∇R is 𝛻R∕R = 𝜋l 𝜎 where 𝜋 f is defined as the transverse piezoresistance coefficient. For situations in which the current flow is parallel to the stress, the longitudinal piezoresistance coefficient 𝜋 l i used. In general case in which both transverse and longitudinal stresses are present, we have 𝛻R∕R = 𝜋t 𝛿t + 𝜋l 𝛿l

(6.2)

To use Eq. (6.2) to calculate a resistance change, the piezoresistance coefficients need to be calculated. As previously mentioned, the values of 𝜋 t and 𝜋 l depend on several properties of the crystal and on the orientation of the stress 𝜎 relative to the crystal lattice orientation of the single-crystal silicon.

6.3 From MEMS to NEMS and then to Quantum Films and AEMS

Bulk micromachining refers to MEMS fabrication processes that involve the removal of significant amounts of the silicon substrate in order to form the desired structure. Etching is the cornerstone of bulk micromachining [33, 35, 36]. Historically, wet etching with both isotropic and anisotropic etchants has dominated MEMS devices, but more recently, other techniques, such as isotropic vapor-phase etching and high-density plasma-based processes, have been used. Silicon isotropic dry etchants such as hydrofluoric acid–nitric acid–acetic acid are well known, showing an uncontrolled etch profile. The undercut of the features defined by the masking material results in the contours of the circular holes and grooves. Stephen A. Campbell shows etch profiles for isotropic etchants with and without agitation of the etchant. With agitation, the profile generally exhibits the shape expected for etch rates that are nearly the same for all crystal orientations. Without agitation, the profile shows u flatter bottom, which is the result of reduced etching in the vertical direction. This reduction in etch rate is caused by a depletion of the etching species near the etch surface (the etch is diffusion rate limited). The agitation assists the diffusion of fresh etchant to the surface, resulting in the expected isotropic etch profile. For bulk-micromachined structures, etch depths often approach the full wafer thickness. When isotropically etching these deep structures, the considerable undercutting of the etch profile mask means the features must be separated by at least the depth of the etch. The MEMS fabrication process utilizes many of the same process steps as conventional silicon LC processes. However, differences have arisen, such as the nature of electronic circuits of MEMS sensors and actuators. MEMS devices typically include structural components that are movable, implying an independent or unsupported element. Material mechanics include the stress/strain relationships and motion of the membrane and cantilever beam. Two traditional methods of forming MEMS devices are bulk micromachining and surface micromachining. Batch micromachining is a relatively simple processing technique for devices such as piezoresistive pressure sensors. Batch micromechanical devices cannot easily be integrated with standard IC processing of on-chip signal processing circuits for MEMS devices. Surface micromachining is more suited to integration with on-chip electronics. But this integration requires considerable care. Surface micromachining using multiple layers of structural material and sacrificial material enables the fabrication of complex structures not attainable with bulk micromachining, such as small motors, resonators, and optical elements out of the plane of the surface of the wafer. MEMS actuators, which convert an electrical signal to motion using many different transduction methods, are finding application in many fields for small controlled motion. High-aspect-ratio microsystem (HARM) processing combines X-ray lithography with electroplating to form structures that can be hundreds of microns high. These high-aspect-ratio plated structures can then be used as molds for precise formation of plastic parts using molding processes, as well as for mechanical structures such as gears and micromotors.

209

210

6 MEMS, NEMS, AEMS, and Quantum Films

6.4 NEMS and AEM 6.4.1

NEMS

Single-crystal diamond has the best properties for NEMS/MEMS among the semiconductors owing to its outstanding mechanical strength, thermal conductivity, electronic properties, and chemical inertness. However, on-chip single-crystal diamond (SCD) NEMS/MEMS with all electrical sensing and actuation integrated into one SCD wafer have the challenges such as arising from the lack of shallow bulk dopants necessary for high electrical conductivity to overcome. A heterodyne frequency downconversion method is utilized to characterize the resonance frequency of the MEMS and NEMS cantilevers. The cantilevers are actuated by the gate electrode, and the motion detected electrically by the S–D electrodes using the “1ω” method [37], as shown in Reference [38]. The on-chip SCD NEMS/MEMS displays its advantages: high self-sensitivity, low-voltage actuation (mV), little energy dissipation, and high-frequency (>MHz) and high-temperature operation (873 K). It shows that the actuation force is enhanced 800 times by a self-sensing readout voltage as low as 1 V compared with those without sensing. The SEA mechanism enables the removal of the direct current (DC) voltage prebiased on the gate, differing from the reported configurations [33, 37, 39, 40]. A radiofrequency (RF) signal with an amplitude of g ac V and a frequency of ω was applied along the bottom gate, while a second RF signal with an amplitude of d ac V and a frequency of 𝜔 + Δ𝜔 was applied to the drain electrode. The readout was performed by a lock-in amplifier at Δ𝜔 = 5 kHz at an impedance of 50 Ω. The finite element method (FEM) revealed that a strong electric field is confined around the edge of the SCD cantilever. Traditional MEMS devices are characterized as either sensors or actuators. Examples of MEMS sensors include acceleration sensors used for automobile airbag deployment control, pressure sensors mounted on the tip of catheters for use in intracardiac (within the heart) monitoring of blood pressure, and chemical sensors that quantitatively detect gaseous compounds. Example of MEMS actuators include video display systems using digital mirror devices consisting of over one million individually controlled micromirrors, ink-dispensing nozzles used in inkjet printers, and valves and pumps used in miniature fluidic systems (fluid volumes in the microliter range). Many processing issues do affect both standard IC and MEMS fabrication, including thin film stress, planarization, and selective wet and dry etching. NEMS switches used for various portable electronic NEMS switches used for various portable electronic NEMS and RF communication systems are the ideal bi-stable switch and the off state can be achieved through the electric switch in a reliable way. In this chapter, the nonideal characteristics of the cantilever NEMS switch, namely, the stability and parameter sensitivity of three representative system models based on Euler–Bernoulli beam theory, the electrostatic interaction between parallel plate capacitor and surface form of Leonard–Jones more physical fields, are discussed. The

6.4 NEMS and AEM

device geometry, material properties, and surface features are compressed in several devices of dimensionless quantities, thus creating the space, the dimensional parameters low enough to provide physical-related balance of all access system model analysis about this kind of bistable system and three steady-state conditions. The multistability region and instability boundary of dimensionless parameter space are determined, and the instability sensitivity of dimensionless parameter space is studied. Under the defined method, cantilever MEMS/NEMS capacitor switches show unity, gender, and triple stability under the condition of physical accessibility, even including the most inherent Lennard–Jones surface interaction model. The results provide a framework for exploring the coupling between multistability and surface interaction, which is easy to expand and is related to the design and performance of such devices.

6.4.2

AEMS

As Si transistors rapidly approach their projected scaling limit of ∼5-nm gate lengths, exploration of new channel materials and device architectures is of utmost interest [41–43]. This scaling limit arises from short-channel effects [44]. Direct source-to-drain tunneling and the loss of gate electrostatic control on the channel severely degrade the off state leakage currents, thus limiting the scaling of Si transistors [45, 46]. Certain semiconductor properties dictate the magnitude of these effects for a given gate length. Heavier carrier effective mass, larger bandgap, and lower in-plane dielectric constant yield lower direct source-to-drain tunneling currents. Uniform and atomically thin semiconductors with low in-plane dielectric constants are desirable for enhanced electrostatic control of the gate. Thus, investigation and introduction of semiconductors that have more ideal properties than Si could lead to further scaling of transistor dimensions with lower off state dissipation power. Transition metal dichalcogenides (TMDs) are layered two-dimensional (2D) semiconductors that have been widely explored as a potential channel material replacement for Si [47–50], and each material exhibits different band structures and properties [51–55]. The layered nature of TMDs allows uniform thickness control with atomic-level precision down to the monolayer limit. This thickness scaling feature of TMDs is highly desirable for well-controlled electrostatics in ultrashort transistors (3). For example, monolayer and few-layer MoS2 have been shown theoretically to be superior to Si at the sub-5-nm scaling limit [56]. As for atom-electromechanical system (AEMS), Sujay B. Desai et al. [57] reported a MoS2 with 1-nm gate lengths. The device is based on 1D-gated, 2D semiconductor field-effect transistors (1D2D-FET). The experimental device structure of the 1D2D-FET consists of a MoS2 channel (number of layers vary), a ZrO2 gate dielectric, and a single-walled carbon nanotube (SWCNT) gate on a 50-nm SiO2 /Si substrate with a physical gate length (LG ∼ d) of ∼1 nm. Long, aligned SWCNTs grown by chemical vapor deposition were transferred onto a n + Si/SiO2 substrate (50-nm-thick SiO2 ) [58], located with a scanning electron microscope (SEM), and contacted with palladium via lithography and metallization. These steps were

211

212

6 MEMS, NEMS, AEMS, and Quantum Films

followed by atomic layer deposition (ALD) of ZrO2 and pick-and-place dry transfer of MoS2 onto the SWCNT covered by ZrO2 [53]. Nickel source and drain contacts were made to MoS2 to complete the device. Qiu C. et al. [59] reported a kind of carbon nanotube field-effect transistors (CNT FET), together with three cross-sectional transmission electron microscopy (TEM) images showing a p-type FET, an n-type FET, and the gate stack structure of these devices. Thanh Luan Phan et al. [10] reported a screening-engineered carbon nanotube (CNT) network/MoS2 /metal heterojunction vertical field-effect transistor (CNT–VFET) is fabricated for an efficient gate modulation independent of the drain voltage. Compared with gr-VFFET, m-SWCNT network realizes a new VFET design based on vdWH with high on/off ratio (>104 ) at room temperature. By theoretical calculation and experimental systems, we find that a high gate modulation independence of the CNT network through the empty space between the gate voltage and the drain modulation is the key to achieving high performance of VFET [60]. Compared with the conventional field-effect transistors, their actual current CNT–VFET is 7.3 times and 112.2 times MoS2 barristor and MoS2 planar FETs.

6.5 Quantum Films for Information Technology When the size of molecular devices is reduced to the nanometer scale, Ohm’s law in the macroscopic system no longer applies, but instead appears as the superposition of multiple discrete orbital transmission coefficients when electrons pass through the molecule, that is, the quantum interference effect of single-molecule electrical transport, similar double slit interference. Recently, Prof. J. Fraser Stoddart’s group at Northwestern University, in collaboration with Prof. Wenjing Hong’s group at Xiamen University and Prof. Guo Hong’s group at McGill University, used single-molecule electricity, which has the accuracy of micro/nanoscale electrical measurement and the sensitivity of subnanometer displacement control. The quantum interference effect based on charged macrocyclic molecular system has been studied by the measurement technology, and a new self-gated quantum interference mechanism has been proposed, which can reach more than 50 times of the single-molecule conductance regulation. In this study, the author relies on the high electrical measurement accuracy and the displacement control crack sensitivity of the scanning tunnel junction technique and firstly proposes the quantum interference based on the gated mechanism to make the two-channel conductance of the molecular system more than that of the single-channel system. More than 50 times, it has broken through the traditional rules of quantum superposition theory, restricted the interdisciplinary research results, confirmed the important role of the strong electrostatic interaction between molecular orbitals on the electrical transmission process, and provided an important theoretical basis for the

6.5 Quantum Films for Information Technology

future design and preparation of high-performance single-molecule electronic devices. Nanofilms of the asymmetric profile can also be modeled by well-known asymmetric potential functions such as the Morse potential, the electrostatic potential of the electric field [61], the semi-parabolic potential [62], etc. These potentials are well-known and have applications in various fields of physics, in particular, solid-state physics, for example, the Morse potential as a potential for interaction between atoms in molecules [63], in a solids [64], in metallurgy [65]. The semi-parabolic quantum well also attracts the attention of researchers. While the interband absorption is under the action of intense laser radiation, a linear optical absorption are considered in [62, 66]. The optical properties in electric and magnetic fields are considered in another paper [67]. It should also be noted [68, 69], where the influence of the asymmetry of the quantum well on the mobility of carriers and the electron–phonon interaction is investigated. In this study, we used an asymmetric distribution of Morse potential to account for the difference in the surface of the quantified membrane [61]. According to the solution of the one-dimensional Schrödinger equation, the energy spectrum of the electrons in the film with Morse potential distribution can be obtained. The results show that the density of the electronic states in the asymmetric membrane is sufficiently strong as the energy increases, and it can be several times the density of the electronic states in symmetric holes such as parabolic and rectangular. In addition, the influence of the asymmetry of the potential well on the electron concentration dependence of Fermi energy and the electron thermoelectric potential of the electron gas was investigated. The results show that in the classical strong magnetic field, the electron thermal potential of the strongly degraded electron gas is an oscillation function of the electron concentration as a function of the period. The period of the electron concentration first increases, then passes through the maximum value, and then decreases to a smaller value without becoming zero. For symmetrical, such as parabolic or rectangular deep wells, the corresponding period is even increased or infinitely extended. In addition, Fermi energy was found to be a piecewise linear function of electron concentration. Comparison with the experiment [70, 71] shows a qualitative agreement. A further decrease in transmission was observed when the complex was coupled or mixed into the polymer film with graphene quantum dots (GQD). The photodegradation threshold recorded by PCS embedded in the polymer film is higher than the photodegradation threshold in solution, resulting in higher stability and better NLO response [72]. The transmittance of the representative z-scan section of the dimethyl formamide (DMF) is decreased near the focus, which may be due to the absorption of the reverse saturated excited state, which is manifested as the positive nonlinear absorption of incident light. The optical-limiting properties improve when the complexes are embedded in thin films, and at the same time, the third-order susceptibilities and hyperpolarizabilities increase, and the limiting-threshold values are further reduced.

213

214

6 MEMS, NEMS, AEMS, and Quantum Films

6.6 Summary and Perspectives The MEMS/NEMS technique is more developed now, and the technology of AEMs has great advantage compared with the earlier devices, including the improvement of production and life, the development of quantum film, its easy preparation and convenient storage, and the superior performance of quantum thin film material. These technologies will certainly have a profound impact on the next generation of computing technology and bring revolutionary improvement to computing technology. Driven by many scientific and technological innovations, the semiconductor industry has shown a significant trend of miniaturization. However, if this trend continues faster, cheaper computers will become available, and then the size of microelectronic circuit components will soon need to reach the atomic or molecular scale, a goal that will conceptually require new device structures. The idea that several or even one molecule could be embedded between electrodes and perform the basic functions of digital electronics (rectification, amplification, and storage) was proposed in the mid-1970s and it has now been implemented for individual components. However, the economic manufacture of complete circuits at the molecular level remains challenging because of the difficulty of connecting molecules to each other, which could be resolved using “monomolecular” electronics, in which individual molecules would integrate the basic functions and interconnections required for computing.

Acknowledgments This chapter is supported by the NSFC-BRICS STI Framework Program (No. 51861145309), the National Natural Science Foundation of China (No. 51971029), the National S&T Major Project (No. 2018ZX10301201), the Postdoctor Research Foundation of Shunde Graduate School of University of Science and Technology Beijing (No. 2020BH005), the Project funded by China Postdoctoral Science Foundation (No. 2020M680336), the “All English teaching demonstration course construction project of University of Science and Technology Beijing” (No. KC2015QYW06, 2016), the “1125” Zhihui Zhengzhou Talent project of Henan Province (Fund No. in USTB: 39080070), the “100 Talent Plan” fund of Fujian Province (Fund No. in USTB: 39080067), and the development of a high sensitive magneto-optical biomolecular sensor experimental prototype (Fund No. in USTB: 2019-0649) by Hangzhou Ruidi Biotechnology Co. Ltd.

List of Abbreviations AEMS ALD BGA

atom-electromechanical systems atomic layer deposition ball grid array

References

CMOS CNT CNT-VFET CSP FET GMR GQD HARM HF LPCVD MCP MEMS MMIC MOS MOSFET NEMS PECVD SCD SEM SWCNT TMDs

complementary metal oxide semiconductor carbon nanotube vertical field-effect transistor chip-scale package field-effect transistor giant magnetoresistance graphene quantum dots high-aspect-ratio microsystems hydrofluoric acid low-pressure chemical vapor deposition multiple-chip package microelectromechanical systems monolithic microwave integrated circuit metal oxide semiconductor metal oxide semiconductor field-effect transistor nanoelectromechanical systems plasma-enhanced chemical vapor deposition single-crystal diamond scanning electron microscopy single-walled carbon nanotube transition metal dichalcogenides

References 1 Stathis, J.H. and Dimaria, D.J. (1999). Oxide scaling limit for future logic and memory technology. Microelectronic Engineering 48 (1/4): 395–401. 2 Goldschmidtböing, F., Engler, M., and Doll, A. (2006). Silicon microfabrication for microfluidics. In: Micro Process Engineering. 3 Kuliev, B.I. and Figarova, S.R. (1977). Galvano- and thermomagnetic phenomena in thin films with nonidentical surfaces. Soviet Physics Journal 20 (10): 1326–1330. 4 Guliyev, B., Kerimli, G., and Bashirov, N. (2019). Fermi energy and electron thermopower in quantum films of an asymmetric profile. Chinese Journal of Physics 59: 110–116. 5 Grum, J. (2008). Book review: microengineering, MEMS, and Interfacing: a practical guide by D. Banks. International Journal of Microstructure & Materials Properties 3 (4-5): 707–708. 6 Wise, K.D. (1998). Special issue on integrated sensors, microactuators, and microsystems (MEMS). Proceedings of the IEEE 86: 1531–1533. 7 deMello, A. (2006). Control and detection of chemical reactions in microfluidic systems. Nature 442 (7101): 394–402. 8 Korzeniewska, E., Tomczyk, M., and Walczak, M. (2020). The influence of laser modification on a composite substrate and the resistance of thin layers created using the PVD process. Sensors 20 (7): 1920.

215

216

6 MEMS, NEMS, AEMS, and Quantum Films

9 Zhang, W., Wang, Y., Wang, Q. et al. (2018). The surface and interface engineering multi-layered nanopore films for enhanced Fabry–Pérot interferences. The Journal of Physical Chemistry C 122 (51): 29457–29463. 10 Thanh Luan, P., Quoc An, V., Kim, Y.R. et al. (2019). Efficient gate modulation in a screening-engineered MoS2/single-walled carbon nanotube network heterojunction vertical field-effect transistor. Acs Applied Materials & Interfaces 11 (28): 25516–25523. 11 Song, Y., Kumar, C.S.S.R., and Hormes, J. (2004). Fabrication of an SU-8 based microfluidic reactor on a PEEK substrate sealed by a flexible semi-solid transfer (FST) process. Journal of Micromechanics and Microengineering 14 (7): 932–940. 12 Song, Y. (2010). Fabrication of multi-level 3-dimension microstructures by phase inversion process. Nano Micro Letters 2 (2): 95–100. 13 Williams, K.R., Gupta, K., and Wasilik, M. (2003). Etch rates for micromachining processing-Part II. Journal of Microelectromechanical Systems 12 (6): 761–778. 14 Ye, S.Y., Tanaka, S., Esashi, M. et al. (2006). MEMS-based thin palladium membrane microreactors. Proceedings of SPIE The International Society for Optical Engineering 6032: 603207–603210. 15 Salloum, K.S., Hayes, J.R., Friesen, C.A., and Posner, J.D. (2008). Sequential flow membraneless microfluidic fuel cell with porous electrodes. Journal of Power Sources 180 (1): 243–252. 16 Yamada, Y., Watanabe, T., Uozumi, Y. et al. (2010). Suzuki–Miyaura couplings with palladium membrane installed microreactors. Synfacts 2010 (12): 1439–1439. 17 Wagner, L.T., Yang, J., Ghobadian, S. et al. (2012). A microfluidic reactor for energy applications. Open Journal of Applied Biosensor 1 (3): 21–25. 18 Kusakabe, K., Morooka, S., and Maeda, H. (2001). Development of a microchannel catalytic reactor system. Korean Journal of Chemical Engineering 18 (3): 271–276. 19 Tiggelaar, R.M., Berenschot, J. W.; Oosterbroek, R. E.; et al. (2003). A new technique for accurately defined deposition of catalyst thin films in deep flow channels of high-temperature gas microreactors. TRANSDUCERS ’03. 12th International Conference on Solid-State Sensors, Actuators and Microsystems. Digest of Technical Papers (Cat. No.03TH8664), 8-12 June 2003. pp. 746-749. 20 Kulah, H., Chae, J., and Najafi, K. (2015). TRANSDUCERS 2003 - 12th International Conference on Solid-State Sensors, Actuators and Microsystems, Digest of Technical Papers. 21 Zhang, H., Ye, J., and Qin, X. (2019). Facile fabrication and transistor properties of mixed crystalline TiO2 nanofibers FET devices. Materials Letters 246: 99–102. 22 Mizuki, O., Lange, D., Brand, O. et al. (2002). A complementary-metal-oxide-semiconductor-field-effect-transistor-compatible atomic force microscopy tip fabrication process and integrated atomic force microscopy cantilevers fabricated with this process. Ultramicroscopy 91 (1): 9–20. 23 Müller, T., Brandl, M., Brand, O., and Baltes, H. (2000). An industrial CMOS process family adapted for the fabrication of smart silicon sensors. Sensors and Actuators A: Physical 84 (1–2): 126–133.

References

24 Kloeck, B., Collins, S.D., de Rooij, N.F., and Smith, R.L. (1989). Study of electrochemical etch-stop for high-precision thickness control of silicon membranes. IEEE Transactions on Electron Devices 36 (4): 663–669. 25 Lan, T., Liu, Y.W., Jin, M.H. et al. (2009). Study of ultra-miniature giant magneto resistance sensor system based on 3D static magnetic analysis technique. Measurement 42 (7): 1011–1016. 26 Ortiz-Ibarra, H. and Medina, J.A. (2004). Uniform thin film electrodeposition onto large circular wafer substrates. Journal of Applied Electrochemistry 34 (7): 751–756. 27 Guckel, H., Randazzo, T., and Burns, D.W. (1985). A simple technique for the determination of mechanical strain in thin films with applications to polysilicon. Journal of Applied Physics 57 (5): 1671–1675. 28 Auld, B.A. and Green, R.E. (1974). Acoustic fields and waves in solids: two volumes. Physics Today 27 (10): 63–64. ´ S. (1986). Integrated silicon magnetic-field sensors. Sensors and Actuators 29 Kordic, 10 (3–4): 347–378. 30 Baltes, H.P. and Popovic, R.S. (1986). Integrated semiconductor magnetic field sensors. Proceedings of the IEEE 74 (8): 1107–1132. ´ R. (1989). Hall-effect devices. Sensors and Actuators 17 (1–2): 39–53. 31 Popovic, 32 Pfann, W.G. and Thurston, R.N. (1961). Semiconducting stress transducers utilizing the transverse and shear piezoresistance effects. Journal of Applied Physics 32 (10): 2008–2019. 33 Kauzlarick, D. (2003). Fundamentals of microfabrication, the science of miniaturization, 2nd edition [Book Review]. IEEE Engineering in Medicine and Biology Magazine 22 (2): 109–111. 34 Flinn, P.A. (1988). Principles and applications of wafer curvature techniques for stress measurements in thin films. MRS Online Proceedings Library Archive 130: 41. 35 Eccardt, P.C., Niederer, K., and Fischer, B.. (1997) Micromachined transducers for ultrasound applications. 1997 IEEE Ultrasonics Symposium Proceedings. An International Symposium (Cat. No.97CH36118), 5-8 October 1997, pp 1609-1618. 36 Kendall, D.L., C.B. Fleddermann, and K.J. Malloy, Chapter 7 Critical technologies for the micromachining of silicon. Semiconductors and Semimetals, 1992. 37: 293–337. 37 Barton, R.A., Parpia, J., and Craighead, H.G. (2011). Fabrication and performance of graphene nanoelectromechanical systems. Journal of Vacuum Science & Technology, B: Nanotechnology & Microelectronics: Materials, Processing, Measurement, & Phenomena 29 (5): 050801–050801-10. 38 Liao, M., Sang, L., Teraji, T. et al. (2019). Ultrahigh performance on-chip single crystal diamond NEMS/MEMS with electrically tailored self-sensing enhancing actuation. Advanced Materials Technologies 4 (2): 1800325. 39 Bunch, J.S., van der Zande, A.M., Verbridge, S.S. et al. (2007). Electromechanical resonators from graphene sheets. Science 315 (5811): 490–493.

217

218

6 MEMS, NEMS, AEMS, and Quantum Films

40 Verd, J., Uranga, A., Teva, J. et al. (2006). Integrated CMOS–MEMS with on-chip readout electronics for high-frequency applications. IEEE Electron Device Letters 27 (6): 495–497. 41 Theis, T.N. and Solomon, P.M. It’s time to reinvent the transistor. Science 327 (5973): 1600–1601. 42 Chau, R., Doyle, B., Datta, S. et al. Integrated nanoelectronics for the future. Nature Materials 6 (11): 810–812. 43 Franklin, A.D. Nanomaterials in transistors: from high-performance to thin-film applications. Science 349, 6249: aab2750. 44 Lundstrom, M. (2003). Moore’s law forever? Science 299 (5604): 210–211. 45 Luisier, M., Lundstrom, M., Antoniadis, D. A., Bokor, J. (2011). Ultimate device scaling: intrinsic performance comparisons of carbon-based, InGaAs, and Si field-effect transistors for 5 nm gate length. 2011 International Electron Devices Meeting, 5-7 December 2011; pp. 11.2.1–11.2.4. 46 Kawaura, H., Sakamoto, T., and Baba, T. Observation of source-to-drain direct tunneling current in 8 nm gate electrically variable shallow junction metal–oxide–semiconductor field-effect transistors. Applied Physics Letters 76 (25): 3810. 47 Radisavljevic, B., Radenovic, A., Brivio, J. et al. Single-layer MoS2 transistors. Nature Nanotechnology 6 (3): 147–150. 48 Sarkar, D., Xie, X., Liu, W. et al. A subthermionic tunnel field-effect transistor with an atomically thin channel. Nature 526 (7571): 91–95. 49 Liu, H., Neal, A.T., and Ye, P.D. (2012). Channel length scaling of MoS2 MOSFETs. ACS Nano 6 (10): 8563–8569. 50 Wang, H., Yu, L., Lee, Y.-H. et al. Integrated circuits based on bilayer MoS2 transistors. Nano Letters 12 (9): 4674–4680. 51 Mak, K.F., McGill, K.L., Park, J., and McEuen, P.L. (2014). The valley hall effect in MoS2 transistors. Science 344 (6191): 1489–1492. 52 Jariwala, D., Sangwan, V.K., Lauhon, L.J. et al. Emerging device applications for semiconducting two-dimensional transition metal dichalcogenides. ACS Nano 8 (2): 1102–1120. 53 Fang, H., Battaglia, C., Carraro, C. et al. Strong interlayer coupling in van der Waals heterostructures built from single-layer chalcogenides. Proceedings of the National Academy of Sciences 111 (17): 6198–6202. 54 Novoselov, K.S., Jiang, D., Schedin, F. et al. (2005). Two dimensional atomic crystals. Proceedings of the National Academy of Sciences of the United States of America 102 (30): 10451–10453. 55 Lee, C.-H., Lee, G.-H., van der Zande, A.M. et al. Atomically thin p–n junctions with van der Waals heterointerfaces. Nature Nanotechnology 9 (9): 676–681. 56 Yoon, Y., Ganapathi, K., and Salahuddin, S. (2011). How good can monolayer MoS2 transistors be? Nano Letters 11 (9): 3768–3773. 57 Desai, S.B., Madhvapathy, S.R., Sachid, A.B. et al. MoS2 transistors with 1-nanometer gate lengths. Science 354: 97–102.

References

58 He, X., Gao, W., Xie, L. et al. (2016). Wafer-scale monodomain films of spontaneously aligned single-walled carbon nanotubes. Nature Nanotechnology 11 (7): 633–638. 59 Qiu, C., Zhang, Z., Xiao, M. et al. (2017). Scaling carbon nanotube complementary transistors to 5-nm gate lengths. Science 355 (6322): 271–276. 60 Yu, W.J., Li, Z., Zhou, H. et al. (2012). Vertically stacked multi-heterostructures of layered materials for logic transistors and complementary inverters. Nature Materials 12 (3): 246–252. 61 Elton, B.L.R. (1958). Quantum Mechanics, Non-Relativistic Theory: Vol. 3 of Course of Theoretical Physics. Physics Bulletin 9 (10): 270–271. 62 Niculescu, E.C. and Eseanu, N. Interband absorption in square and semiparabolic near-surface quantum wells under intense laser field. European Physical Journal B 79 (3): 313–319. 63 Kaplan, I.G. and Rodimova, O.B. (1978). Intermolecular interactions. Uspekhi Fizicheskih Nauk 126 (11): 403–449. 64 Slutsker, A.I. Characteristics of elementary acts in the kinetics of metal fracture. Physics of the Solid State 46 (9): 1658–1666. 65 Girifalco, L. and Weizer, V. (1959). Application of the morse potential function to cubic metals. Physical Review 114 (3): 687–690. 66 Tien, N.T., Hung, N.N.T., Nguyen, T.T., and Thao, P.T.B. (2017). Linear intersubband optical absorption in the semiparabolic quantum wells based on AlN/AlGaN/AlN under a uniform electric field. Physica B: Physics of Condensed Matter 519: 63–68. 67 Yan, R.-Y., Tang, J., Zhang, Z.-H., and Yuan, J.-H. (2018). Optical properties in GaAs/AlGaAs semiparabolic quantum wells by the finite difference method: combined effects of electric field and magnetic field. International Journal of Modern Physics B 32 (13): 1850159. 68 Lima, F.M.S., Fonseca, A.L.A., Nunes, O.A.C., and Fanyao, Q. (2002). Electric field effects on electron mobility in n-AlGaAs/GaAs/AlGaAs single asymmetric quantum wells. Journal of Applied Physics 92 (9): 5296. 69 Stavrou, V.N., Babiker, M., and Bennett, C.R. Influences of asymmetric quantum wells on electron-phonon interactions. Journal of Physics: Condensed Matter 13 (30): –6489, 6498. 70 Martin, J. (2009). Enhanced Seebeck coefficient through energy-barrier scattering in PbTe nanocomposites. Physical Review B: Condensed Matter 79 (11): 5311. 71 Miller, N., Haller, E.E., Koblmüller, G. et al. (2011). Effect of charged dislocation scattering on electrical and electrothermal transport in n -type InN. Physical Review B 84 (7): 2989–2996. 72 Nwaji, N., Mack, J., Britton, J., and Nyokong, T. (2017). Synthesis, photophysical and nonlinear optical properties of a series of ball-type phthalocyanines in solution and thin films. New Journal of Chemistry 41 (5): 2020–2028.

219

221

7 Metamaterial or Metastructural Thin Films for EM Wave Control Menglin L.N. Chen 1 , Luzhou Chen 2 , Xunwang Dang 3 , Maokun Li 4 , Li Jun Jiang 1 , and Wei E.I. Sha 5 1 The University of Hong Kong, Department of Electrical and Electronic Engineering, Pokfulam 999077, Hong Kong 2 École Polytechnique de Montréal, Department of Electrical Engineering, 2500 Chemin de Polytechnique, Montréal, QC H3T 1J4, Canada 3 Science and Technology on Electromagnetic Scattering Laboratory, Beijing 100854, China 4 Tsinghua University, Beijing National Research Center for Information Science and Technology, Department of Electronic Engineering, State Key Laboratory on Microwave and Digital Communications, ShuangQing Road No. 30, Beijing 100084, China 5 Zhejiang University, College of Information Science and Electronic Engineering, Key Laboratory of Micro-Nano Electronic Devices and Smart Systems of Zhejiang Province, Department of Electronic Engineering, 38 Zheda Road, Hangzhou 310027, China

7.1

Introduction

Electromagnetic (EM) response from bulk materials governed by Maxwell’s equations depends on the constitutive relations with bulk permittivities, permeabilities, conductivities, etc. The homogenization of many-body behaviors of microscopic electrons, ions, and protons results in elegant constitutive relations with the bulk EM parameters. By exploring the same principle, the spatially inhomogeneous EM parameters caused by periodic, quasiperiodic, and random scatterers can also be homogenized in the macroscopic level if these scatterers are significantly smaller than the EM wavelength. Consequently, the scatterers can be reshaped, orientated, translated, transformed, and dynamically tuned to flexibly modify the properties of EM waves locally and form a specific EM structure with peculiar and fascinating homogenized EM parameters. Initially, the arrangements of the scatterers spread to three-dimensional (3D) space, and the corresponding structures are called metamaterials in the literature. Metamaterials control EM wave properties in an extraordinary way in comparison with traditional homogeneous bulk materials. Metamaterials with negative, zero, and gradient refraction indices were invented and studied extensively in both microwave and optical frequencies. For example, at microwave frequencies, a metamaterial sample formed by periodically arranged split-ring resonators and

Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications, First Edition. Edited by Yujun Song. © 2021 WILEY-VCH GmbH. Published 2021 by WILEY-VCH GmbH.

222

7 Metamaterial or Metastructural Thin Films for EM Wave Control

wires was experimentally demonstrated to exhibit negative refraction index (NRI) [1, 2]. With the NRI, the metamaterial can amplify evanescent waves so that it can be used as superlens to overcome the diffraction limit for imaging [3, 4]. Besides, zero-index-metamaterials (ZIM) with zero-phase delay nearly independent of their geometries were proposed as a coupler [5] and cloak [6]. Furthermore, by adopting transformation optics, the impedance of metamaterial-based cloaks can be matched to free space for an arbitrary cloaked object [7]. Metamaterials can also be applied to enhance the directivity of antennas, such as gradient-index metalens [8, 9] and metamaterial-based antennas [10, 11]. Additionally, metamaterials with anisotropic EM response can be designed as polarizers [12, 13]. However, the 3D structural thin films on desired substrates are not easy to be fabricated, integrated, and particularly not compatible to modern planar designs. Afterward, the scatterers are spatially arranged in a flat or curved plane like a metastructural film, which is called metasurface in the literature. Unlike metamaterials that change the properties of EM wave gradually along the wave path in the media, metasurfaces alter the EM wave property by the abrupt phase shift at the scatterers [14]. Compared with metamaterials, metasurfaces not only manipulate EM waves to achieve marvelous functionalities but also remain ultrathin configuration. Therefore, at microwave frequencies, metasurfaces can be conveniently fabricated using printed circuit board (PCB) technique, on different substrates based on the design targets. By designing scatterers with different geometries or orientations, phase shifts from 0 to 2𝜋 can be covered. Therefore, a metastructural film formed by those scatterers can realize arbitrary beam forming, such as beam shaping, steering, and focusing [15–17]. In this chapter, we will review the theoretical foundations, design routes, numerical methods, and engineering applications of the metastructural film (metasurface) in a comprehensive manner. Because the response of a metasurface differs locally, the simulation of both the single scatterer and the whole metasurface needs to be conducted. In Sections 7.2 and 7.3, we will introduce several modeling approaches for metasurfaces to facilitate fast design process and accurate simulation. In Section 7.4, metasurfaces designed for orbital angular momentum (OAM) generation will be reviewed. The metasurfaces introduce additional degrees of freedom in the microwave and optical communications and could be applied to the OAM-based multiplexing and demultiplexing. In Section 7.5, we will introduce the potential application of metasurface in modifying spontaneous emission (SE) of quantum emitters.

7.2

Modeling and Synthesis Methods of Metasurfaces

To transform arbitrary incident field into targeted reflected and transmitted fields as shown in Figure 7.1, we need a synthesis method that can link an effective macroscopic parameter, which represents the homogenized property of the metasurface, to the EM fields at the two sides.

7.2 Modeling and Synthesis Methods of Metasurfaces

Figure 7.1 Schematic presentation of a metasurface that transforms an arbitrary incident field into desired reflected and transmitted fields.

Ei, Hi Er, Hr

Et, Ht

7.2.1

Jones Vector and Jones Matrix

A simple approach to model scatterers on metasurfaces is the Jones matrix [18]. We consider a monochromatic plane wave propagating along the z direction. The incident electric field can be decomposed into x and y components and described by Jones vector: ( ) i (7.1) Ei (r, t) = x e−ikz iy where the time harmonic factor is omitted, k is the wave number, and the complex amplitudes ix and iy represent the polarization states of the incident waves. After the incident wave impinging on the metasurface, the transmitted wave is described in the same manner: ( ) t (7.2) Et (r, t) = x e−ikz ty where tx and ty represent the polarization states of the transmitted wave. Then, the local behavior of a scatterer on the metasurface can be modeled by Jones matrix, J. It connects the transmitted field components, tx and ty to the incident ones: )( ) ( ) ( ) ( J J ix i tx = xx xy =J x (7.3) ty iy iy Jyx Jyy where the first and second subscripts of J denote the polarization states of the transmitted wave and incident wave, respectively. The reflected field can be modeled in the same way. When the scatterer is rotated by an angle 𝛼, as shown in Figure 7.2, the new Jones matrix can be easily obtained by using the rotation matrix R(𝛼): ( ) cos(𝛼) sin(𝛼) J(𝛼) = R(−𝛼)JR(𝛼), R(𝛼) = (7.4) − sin(𝛼) cos(𝛼) Additionally, the Jones matrix, J, can be transformed into circular basis by coordinate transformation so that circularly polarized fields can be directly manipulated: ( ) J++ J+− c J = J−+ J−− ( ) 1 (Jxx + Jyy ) + i(Jxy − Jyx ) (Jxx − Jyy ) − i(Jxy + Jyx ) (7.5) = 2 (Jxx − Jyy ) + i(Jxy + Jyx ) (Jxx + Jyy ) − i(Jxy − Jyx )

223

224

7 Metamaterial or Metastructural Thin Films for EM Wave Control

Inc.

Inc.

Figure 7.2 Schematic representation of a scatterer that rotates an angle of 𝛼.

α x

y

x

y

where Jc connects the incident circularly polarized Jones vectors to the transmitted or reflected circularly polarized ones. + and − represent the left circularly polarized (LCP) and right polarized components, respectively. We should see that only the normal incidence and reflection/transmission are considered by using the Jones matrix. However, the coupling between scatterers is not taken into consideration rigorously.

7.2.2

Polarizability Model

Incident wave induces polarization electric current and magnetic current, leading to the discontinuities of field components when crossing the metasurface plane. For this situation, the conventional boundary conditions fail to describe the system but should be replaced by the generalized sheet transition conditions (GSTCs) [19] n × (H1∕∕ − H2∕∕ ) = Jtoe = i𝜔P∕∕ − n × ∇∕∕ Mn P n × (E1∕∕ − E2∕∕ ) = −Jtom = −i𝜔𝜇0 M∕∕ − n × ∇∕∕ n 𝜀0 P ∕∕ n ⋅ (E1∕∕ − E2∕∕ ) = −∇ ⋅ 𝜀0 n ⋅ (H1∕∕ − H2∕∕ ) = −∇ ⋅ M∕∕

(7.6) (7.7) (7.8) (7.9)

where P and M are the surface electric and magnetic polarization densities, correspondingly. Subscripts “//” and “n” denote the tangential and normal components, respectively, while superscripts “1” and “2” refer to the fields at the two sides of the metasurface. Jtoe and Jtom denote the effective total electric and magnetic currents. Equations (7.6)–(7.9) provide us the information of the induced polarization density that transforms the impinging field in a desired manner, but still it cannot give an intuitive insight of the metasurface design. To investigate this problem, models based on different homogenized parameters such as the polarizability [20], susceptibility [21], and equivalent impedance [22] have been proposed and demonstrated. Polarization density in Eqs. (7.6)–(7.9) can be expressed as a form of the polarizability and the incident field [20]: P=

μM

𝛼 ̂ee 𝛼 ̂ E + em Hi S i S

=

𝛼 ̂me 𝛼 ̂ E + mm Hi S i S

(7.10)

(7.11)

7.2 Modeling and Synthesis Methods of Metasurfaces

where Ei and Hi are the known incident fields and S is the area of the unit cell of the metasurface. 𝛼 ̂ is the effective polarizability dyadic, which represents the collective effect of a single scatterer (inclusion) itself, together with the coupling and interaction from the whole metasurface array [22]. To synthesize a metasurface, one should first determine the polarization by substituting the known incident fields (Ei , Hi ), desired transmitted fields (Et , Ht ), and reflected fields (Er , Hr ), into ̂ components through the obtained (P, M) and Eqs. (7.6)–(7.9) and then find the 𝛼 Eqs. (7.10) and (7.11). For normal incident plane wave and the metasurface with uniaxial symmetry, a simple form of the polarizability and fields can be derived [20] [( ) i𝜔 1 co co cr cr 𝜂0 𝛼 ̂ee + 𝛼 ̂em + 𝛼 ̂me − 𝛼 ̂ I Er = − 2S 𝜂0 mm t ) ] ( 1 cr cr co co + 𝜂0 𝛼 ̂ee −𝛼 ̂em −𝛼 ̂me − 𝛼 ̂ J ⋅ Ei (7.12) 𝜂0 mm t ( )] [[ i𝜔 1 co co cr cr 𝜂0 𝛼 ̂ee + 𝛼 ̂em − 𝛼 ̂me + 𝛼 ̂ It Et = 1 − 2S 𝜂0 mm ) ] ( i𝜔 1 cr cr co co 𝜂0 𝛼 − ̂ee −𝛼 ̂em +𝛼 ̂me + 𝛼 ̂ J ⋅ Ei (7.13) 2S 𝜂0 mm t where I t = I − z ⋅ z is the tangential unit dyadic, while J t = z × I t is the transverse rotation dyadic. Superscripts “co” and “cr” mean the symmetric and antisymmetric components of 𝛼 ̂, respectively:

7.2.3

co cr 𝛼 ̂ee = 𝛼 ̂ee It + 𝛼 ̂ee Jt

(7.14)

co cr 𝛼 ̂em = 𝛼 ̂em It + 𝛼 ̂em Jt

(7.15)

co cr 𝛼 ̂me = 𝛼 ̂me It + 𝛼 ̂me Jt

(7.16)

co cr 𝛼 ̂mm = 𝛼 ̂mm It + 𝛼 ̂mm Jt

(7.17)

Susceptibility Model

Besides the polarizability, alternatively, metasurface can be homogenized by surface susceptibility. Polarization density is described by [21]. √ P = 𝜀𝜒 ee Eav + 𝜇𝜀 𝜒 em Hav (7.18) √ 𝜀 M = 𝜒 mm Hav + 𝜒 E (7.19) 𝜇 me av u = [Eu + (Eu + Eu )]∕2 and H u = [H u + (H u + H u )]∕2 (u = x, y, z) are the where Eav r av r t t i i average fields at the two sides of the metasurface. Closed-form relation of the fields and the susceptibility tensor can be obtained for simplified case, assuming only tangential components of the polarizations are induced, so that Pz = M z = 0. Substitute Eqs. (7.18) and (7.19) into Eqs. (7.6)–(7.9), it leads to

225

226

7 Metamaterial or Metastructural Thin Films for EM Wave Control

( (

y)

−HΔ HΔx y

EΔ −EΔx

(

xy ) ( x ) Eav y Eav

𝜒eexx 𝜒ee = i𝜔𝜀 yx yy 𝜒ee 𝜒ee

)

( = i𝜔𝜇

xy

xx 𝜒mm 𝜒mm yx yy 𝜒mm 𝜒mm

)(

x Hav y Hav

( xx xy ) ( x ) √ Hav 𝜒em 𝜒em + i𝜔 𝜀𝜇 yx yy y 𝜒em 𝜒em Hav )

( xx xy ) ( x ) √ Eav 𝜒me 𝜒me + i𝜔 𝜀𝜇 yx yy y 𝜒me 𝜒me Eav

(7.20)

(7.21)

with EΔu = Etu − (Eiu + Eru ) and HΔu = Htu − (Hiu + Hru ) denote the differences of the fields at the two sides of the metasurface, correspondingly. To this point, susceptibility tensor matrix still cannot be completely determined for a specified field transformation. The number of the unknown matrix components should be reduced by enforcing some extra conditions. For example, to a mono-anisotropic and uniaxxy yx xy yx ial medium, so that 𝜒 em = 𝜒 me = 0 and 𝜒ee = 𝜒ee = 𝜒mm = 𝜒mm = 0, Eqs. (7.20) and (7.21) degrades to a simple relation: 𝜒eexx = 𝜒eeyy = xx 𝜒mm yy 𝜒mm

−HΔy

(7.22)

x i𝜔𝜀Eav x HΔ

(7.23)

y

i𝜔𝜀Eav EΔy = x i𝜔𝜇Hav −EΔx = y i𝜔𝜇Hav

(7.24) (7.25)

Therefore, the metasurface can be synthesized according to the desired fields at the two sides.

7.2.4

Equivalent Impedance Model

Equivalent impedance model based on the transmission line theory is also a powerful method for the metasurface design [22, 23]. Impinging plane wave to the metasurface is analogy to a propagating signal in a transmission line with proper equivalent parameters. Metasurface described by Eqs. (7.6)–(7.9) that can be modeled by a T-circuit as shown in Figure 7.3, with the equivalent impedance matrix that connects the voltages and currents by ) ( 1) ( 1) ( Z11 Z12 i v = (7.26) v2 Z21 Z22 i2 where Z 11 = Z 1 + Z 3 , Z 22 = Z 2 + Z 3 , Z 12 = Z 21 = Z 3 . Linking the tangential fields at the two sides of the metasurface to the voltages and currents of the transmission line, i1

i2 Z1 v1

Z2 Z3

v2

Figure 7.3 Equivalent transmission line model (T-circuit) of a metasurface. Source: Asadchy et al. [23]. © 2016 American Physical Society.

7.3 Simulation Algorithms of Quasi-periodic Electromagnetic Surfaces

(

E1∕∕ E2∕∕

)

) )( ( n × H1∕∕ Z11 Z12 = Z21 Z22 −n × H2∕∕

(7.27)

one can obtain the corresponding impedance matrix of the metasurface.

7.3 Simulation Algorithms of Quasi-periodic Electromagnetic Surfaces 7.3.1

Introduction to EM Surfaces

With the development of EM theory, EM surfaces have attracted more and more attentions due to their flexible design features and rich functions. An EM surface is a planar structure with certain EM properties. In general, its thickness is much smaller than the wavelength, and the aperture area is larger than the wavelength, and it has scattering, transmission, or absorption effects on EM waves. EM surfaces could achieve specific functions through a variety of structures. Most EM surfaces are planar arrays that consist of scatterers arranged in small electrical sizes. Some scatterers in the array are identical, forming a periodic EM surface; some scatterers in the array are not exactly the same, forming a quasi-periodic EM surface. The different scatterers of these EM surfaces respond differently to space EM waves, increasing the degree of freedom for designing the EM surface and enabling many interface characteristics that are not found in nature, such as artificial magnetic conductor (AMC) [24], frequency selective surface (FSS) [25], etc. Figure 7.4 shows a schematic pattern of a general quasi-periodic EM surface. Researchers have proposed many types of quasi-periodic EM surfaces, such as reflect-array [26], transmit-array antenna [27], nano-optical arrays [28], and so on, which can modulate the amplitude and phase of EM waves. The scatterers of quasi-periodic EM surface are generally not the same. Usually, there are some changes such as scaling, rotation, etc., which can effectively adjust the amplitude and phase of the EM waves, and provide flexibility for EM surface design to achieve more functions. For example, a scatterer whose phase changes linearly with the size will achieve a low-profile reflection array with a relative 3 dB bandwidth of more than 20% [29], which is suitable for a wideband mobile communication system. A dual band scatterer with an independent phase adjustment is proposed to realize a circularly polarized low-profile reflectarray [30]. An FSS can combine with a quasi-periodic reflectarray to achieve a reflective EM surface insensitive to incident angles [31]. Quasi-periodic EM surfaces are also flexible in radiation pattern design. A multi-beam reflectarray is designed for wireless communications

1

Figure 7.4

2

3

4

Illustration of a quasi-periodic EM surface.

N−1

N

227

228

7 Metamaterial or Metastructural Thin Films for EM Wave Control

and electronic countermeasure systems [32]. A satellite antenna array with the beam shape covering a European region is designed using beam-forming techniques [33]. A metasurface, which could detect OAM mode for multiplexing communication, is proposed using quasi-periodic EM surface at a microwave band [34]. A nano-metasurface is designed for wideband and wide-angle holographic imaging [35]. In addition, quasi-periodic EM surfaces can also introduce reconfigurable designs to dynamically control their performances. A mechanically controlled reconfigurable quasi-periodic EM surface uses rotating scatterers to control the phase responding to circularly polarized incident waves, thus changing the reflected beam direction accordingly [36]. A reconfigurable quasi-periodic EM surface with PIN diodes is proposed in [37]. The scatterers’ performance in phase response is achieved by the direct current (DC) bias control of a PIN diode, which has a good integration property. A reflectarray with power amplifiers is designed to compensate the incident power [38]. A reconfigurable mid-wavelength optical quasi-periodic EM surface with graphene and gold is designed for beam-forming [39]. The scatterers are controlled by different DC voltages to achieve different phase shifts. These studies have extended the applications of quasi-periodic EM surfaces and proved their advantages over traditional designs in controlling EM waves.

7.3.2

Design of Quasi-periodic EM Surfaces

The design of quasi-periodic EM surfaces mainly contains three steps: scatterer simulation, array computation, and full wave validation. (1) Scatterer simulation. Use the periodic boundary condition (PBC) [40, 41] to simulate and model the scatterer with different parameter values. The PBC is based on Floquet or Bloch theorem. It expands the current distribution on the scatterer into different modes and can model the response of scatterers with respect to different incident waves. Many commercial simulation software integrates such functions. (2) Array computation. The configurations of the scatterers are determined according to the design requirements, and then array computation is performed on the entire quasi-periodic EM surface through the array method or the aperture field method. The array method does not consider the polarization direction of the EM field, and the aperture field method considers it. Consequently, the far field pattern of the co-polarization and the cross-polarization can be calculated. (3) Full-wave validation. After the design is completed, it is often necessary to simulate the entire quasi-periodic EM surface using full-wave methods to verify whether the design achieves the desired goal. Here we take the design of a reflectarray antenna [26] as an example to specifically describe the actual design process of a quasi-periodic EM surface. Reflectarray antennas use feed antennas to generate incident waves. The scatterers are located in the far field of the feed antenna, so the directional pattern of the feed antenna can be used as the amplitude information of the EM wave as an excitation. The phase of incident EM waves are determined by path lengths. The response of scatterers can

7.3 Simulation Algorithms of Quasi-periodic Electromagnetic Surfaces

be calculated using commercial simulation software. As an example, in HFSS, the amplitude and phase responses of the scatterers to plane waves can be described by the S-parameters of different modes of the Floquet ports. The far-field radiation characteristics of the array are computed using either the array method or the cross-polarized aperture field method. In the design of a quasi-periodic EM surface, the PBC is an approximation to the actual situation, and the array calculation process is also not accurate enough. It is pointed out that since the size of each scatterer and the surrounding scatterers are different, the phase of the actual contribution of one scatterer will be different from the phase required [42], and increase the phase error of this scatterer. In order to design a high-performance quasi-periodic EM surface, accurate full-wave simulation is needed and then the array design is adjusted according to the phase error to improve the gain of the array and reduce the side lobes. However, full-wave simulation cannot be widely used in the array optimization process because the time of full-wave simulation of a single array is too long, which usually takes tens of hours. Therefore, the new quasi-periodic EM surface simulation algorithm is urgently needed to reduce the full-wave simulation time to the level of the scatterer simulation time so that a large number of full-wave simulations can be applied during the design process to improve the design accuracy and help to discover more physical phenomena on quasi-periodic EM surfaces.

7.3.3

Simulation Algorithms of Quasi-periodic EM Surfaces

Basic EM simulation methods mainly include method of moments (MoM) [43], finite-element method (FEM) [44], finite-difference time-domain (FDTD) [45], etc. They all discretize the original Maxwell equation into the matrix equation. These full-wave simulation methods can be divided into two categories based on integral equations and partial differential equations. The MoM is based on the integral equation method, and the FEM and FDTD are methods based on partial differential equations. As electrical size of the EM surface is usually large, its simulation is usually based on the surface integral equation using MoM. This is because the MoM can define the unknown on the surface of the geometry to be solved without discretizing the whole 3D space and can reduce the number of unknowns compared to other methods. At the same time, the equations of MoM include the radiation boundary condition of EM waves and do not need to deal with extra radiation boundary conditions, which leads to higher accuracy. Currently, the MoM has been widely applied to the simulation modeling of large scatterers [46], periodic arrays [47], infinite multilayer media [48], and random rough surfaces [49]. In practical applications, the matrix equations corresponding to the MoM contain high-dimensional dense matrices. It is difficult to solve large-scale problems and it needs to be accelerated. To this end, researchers have developed various general acceleration algorithms, such as multilevel fast multipole algorithm (MLFMA) [50], adaptive cross-approximation [51], and domain decomposition [52]. With the development of computers, software developers compile the codes of these algorithms as executable programs and pack

229

230

7 Metamaterial or Metastructural Thin Films for EM Wave Control

them into commercial software such as CST (3DS–CST, www.3ds.com), HFSS (ANSYS–HFSS, www.ansys.com), FEKO (Altair–Feko, www.altairhyperworks .com), COMSOL (COMSOL, www.comsol.com), and so on. At the same time, it can be accelerated according to the characteristics of the programming language, hardware platform, and other tools, such as the use of CUDA graphics processing units (GPU) [53], parallel programming [54], and so on. The full-wave simulation of quasi-periodic EM surfaces is a large- and multi-scale problem. The matrix equation with a large condition number has millions of unknowns. It is generally difficult to effectively solve the problem by using existing full-wave simulation algorithms. This limits the flexibility and accuracy of the quasi-periodic EM surface design and is a bottleneck restricting the development of quasi-periodic EM surfaces. If we design specific fast simulation algorithms for quasi-periodic EM surfaces at the algorithm level, we can effectively reduce the time for full-wave simulation, then improve the precisions of the design processes, and help to design quasi-periodic EM surfaces efficiently.

7.3.4

Review of Simulation Algorithms of Quasi-periodic EM Surfaces

When solving MoM equations on quasi-periodic EM surfaces, common acceleration methods can be divided into four types: approximation algorithms, dimension reduction, matrix compression, and fast Fourier transform (FFT), as shown in Figure 7.5. The main idea of the approximate simulation method is to physically approximate the solution area such as PBC, impedance homogenization, and so on. The PBC is based on the Floquet theory and the current distribution of the scatterers is expanded into linear combinations of different modes [55]. This expansion method is only valid for PBC, which corresponds to the infinite periodic surface formed by the same scatterer, and is usually used for the design phase of the quasi-periodic surface as an approximate model for scatterer simulation. Impedance homogeneity means that when the size of the array unit is small, the EM surface is uniformly processed, and the characteristics of the scatterer are described by electrical parameters. For example, by modeling quasi-periodic surfaces with impedance boundary conditions, the MoM matrix equations can be established [56]. The GSTC can also be used to simplify the EM field equation [57]. The equivalent surface impedance is generally expressed in terms of the polarizability of the electric and magnetic fields. It is related to thermoelectric (TE) or transition-metal (TM) polarization and can be obtained by the Fresnel formula [58]. The GSTC can also be combined with the integral equation Quasi-periodic EM surface simulation algorithms

Approximate simulation method

Figure 7.5

Dimension reduction method

Matrix compression method

FFT-based method

Classification of quasi-periodic EM surface simulation algorithms.

7.3 Simulation Algorithms of Quasi-periodic Electromagnetic Surfaces

method. For example, the researchers used the surface integral equation in the spectral domain to build the MoM equation on the EM surface [59]. When the EM surface mathematically satisfies a certain continuous function, its dielectric constant can be approximated as a continuous function, and its surface characteristics can be directly derived from the EM field propagation equation. An approximate analysis of a quasi-periodic surface with the cosine function shaped scatterers is proposed [60]. The solution to this equation is the Mathieu function. From this, we can get the electric field distribution of the scatterer at different frequencies. For the array of rotating scatterers, a quasi-periodic array consisting of rotating scatterers based on Green’s function modeling is proposed [61], where each scatterer of this array is rotated by a certain angle different from the previous one. The Green’s function of its electric field can be expressed by the Fourier expansion, similar to the Green’s function of the periodic array. Researchers use hierarchical dipole approximation (HDA) to align periodic EM surfaces for modeling [62]. It reduces the number of unknowns and does not require a triangular mesh. It is suitable for small-sized scatterers. In general, the approximate methods have clear physical pictures and can effectively reduce the number of unknowns. The disadvantage is that they are only suitable for electrically small scatterers, or a relatively smooth varying EM surface, and the simulation accuracy is insufficient for large and complex scatterers. The dimension reduction method [63] basically adopts new basis functions to reduce the total number of unknowns in the matrix equation. The first method is to employ global basis functions. For example, for a disc-type quasi-periodic surface, a Gaussian ring basis function defined on the entire EM surface is used to effectively reduce the number of unknowns and accelerate the solution [64]. The reference [65] is based on Fourier–Bessel global base function for elliptic quasi-periodic EM surfaces. In addition, the similarities among scatterers in simulating quasi-periodic EM surfaces and linearly combine the original basis functions can be used to obtain a new basis function. The reference [66] used the characteristic basis function (CBF) method to simulate the quasi-periodic EM surface, which is composed of rotating scatterers and generates the EM waves with angular momentum (AM) mode. The CBF is proposed in [67]. It extracts the information of the current distributed on the array scatterers with EM waves at different incident angles and thus obtains a set of CBFs that can be defined on the whole scatterer. The basic idea of this method is that EM waves can be expanded into superpositions of plane waves of different angles. As long as the number of incident angles is enough, it can be used to represent the responses from different excitations on the array’s scatterers. After the set of CBFs is obtained, the equations of the original MoM can be projected onto the set of basis functions, thereby reducing the number of unknowns. Synthetic basis function (SBF) method has been adopted to apply a certain acceleration method to the actual quasi-periodic EM surface simulation and the corresponding software have been developed [68]. The method of SBF uses the scattered current generated by the point sources around the scatterers in the array as the basis function for the current expansion [69]. Compared with CBF, it extracts basis functions from the near-field EM response, which can represent the role of near-field coupling more effectively. Generally, it can use fewer unknowns than the feature basis function [70]. Simulating periodic EM surfaces with the macro basis functions (MBF), which

231

232

7 Metamaterial or Metastructural Thin Films for EM Wave Control

extract the main vector components from the current distribution of multiple scattering, has been developed as a new basis function [71]. In general, the dimension reduction method can effectively reduce the number of unknowns and has strong versatility. The disadvantage is that it does not use the characteristics of the scatterer arrangement information. Thereby, the performance of the algorithm still has room for further improvement. The method of matrix compression mainly relies on finding the low-rank representation of the impedance matrix because the interaction of basis functions that are far apart can be represented by fewer multiple items and have mathematically low-rank characteristics. Adaptive Cross Approximation (ACA) is explored to calculate the near-field interaction in quasi-periodic EM surface simulation, which can reduce the computational complexity [72]. The ACA method applies the greedy algorithm to decompose the far-field part of the impedance matrix into the product of two low-rank matrices [51]. The literature [73] uses the H-matrix algorithm to compress the near-field interaction of quasi-periodic EM surfaces. The H-matrix algorithm was proposed as a general framework to hierarchically represent low-rank matrices and to reduce the storage and computational complexity [74]. The matrix compression method is versatile and can effectively reduce the complexity of the representation of the impedance matrix. The disadvantage is lack of the scatterer alignment information, and thus the performance of such algorithms can be further improved based on the periodic alignment of scatterers. FFT-based method generally relays on the translation invariance or periodicity of the basis functions. The matrix vector multiplication should be converted into cyclic convolution operation and then employ FFT to accelerate. The matrix generated by the translation invariant basis functions has the Toeplitz property, that is, the entries on each subdiagonal line are equal. In some simple problems, the basis functions themselves can provide periodicity. However, for the general three-dimensional MoM equations, a surface triangular mesh is generally used, on whichthe basis functions are also defined. They do not exhibit translation invariance. This requires that the basis functions defined on the triangular mesh are represented by those defined on uniform grids to obtain translation invariant basis functions and the Toeplitz property of the corresponding impedance matrix. For example [73], uses integral equation fast Fourier transform (IE–FFT )) [75] to accelerate the simulation of quasi-periodic EM surface, and adopts the H-matrix algorithm [74] to further compress the matrix. The IE–FFT method was proposed by Seo and Lee [75]. It uses the Lagrangian interpolation on uniform grids to represent the Green’s function interaction between basis functions. The adaptive integral method (AIM) is combined with an SBF to simulate a quasi-periodic EM surface by De Vita et al. [68] and improves the sparse matrix multiplication in the AIM [76]. The number of unknowns in the original AIM can be reduced, which approximates the basis function with the superposition of impulse functions on uniform grids [77]. It should be noted that the acceleration based on the FFT method is suitable for matrix vector multiplication operation, which needs to be combined with a matrix iterative solver, such as the conjugate gradient method and the fast Fourier transform (CG–FFT) method [78–80]. In addition, other iterative algorithms, such as BCG (bi-conjugate gradient)–FFT [81, 82], TFQMR (transpose free quasi minimal residual)–FFT [83],

7.4 Orbital Angular Momentum of Electromagnetic Waves: Generation and Detection

and stable BCG and so on [84, 85], can also be used as an extension to the basic iterative algorithm to achieve the O(Nlog N) computational complexity, where N is the number of total unknowns. The FFT based methods can effectively speed up the matrix vector multiplication. The disadvantage is that the number of unknowns will increase when the triangular grid is mapped to the uniform grids. Additionally, this kind of algorithm has not been used to explore the scatterer similarity of the quasi-periodic EM surface. From the above literature, the following is a preliminary summary of the research on periodic EM surface fast simulation algorithms: (1) The core idea of the acceleration algorithms is to compress redundant information in quasi-periodic EM surfaces. (2) Some methods required certain assumptions such as small scatterers, smooth change of the surface, etc., with a narrower scope of application. (3) It is necessary to make full use of the characteristics of the EM surface itself. (4) The advantages of existing fast algorithms can be merged by using hybrid methods.

7.4 Orbital Angular Momentum of Electromagnetic Waves: Generation and Detection 7.4.1

Introduction

EM waves carry two types of AM, namely, the spin angular momentum (SAM) and the OAM [86]. In view of a single photon, its value of SAM is ±ℏ for circularly polarized EM waves (ℏ is the reduced Planck constant). Here, +/− sign is taken for left/right-handed circular polarization (L/RHCP). Different from SAM, OAM manifests the orbital rotation of photons and each photon can have an OAM of value lℏ, where l is known as the OAM index and can be any integer. In 1992, Allen et al. first found that Laguerre–Gaussian (LG) beams carry well defined OAM [87]. For EM waves carrying OAM, different values of l correspond to mutually orthogonal wave states, and therefore, these states can be utilized for multiplexing in communication systems [88–91]. Also, there have been other emerging applications of OAM, such as in super resolution imaging [92], optical tweezers [93], and detectors [94]. With the great potential of OAM in various applications, research on its generation has been undertaken extensively. At optical frequencies, to generate OAM beams, optical devices, such as spiral phase plates (SPPs) [95], computer-generated holograms (CGHs) [96], and q plates [97] are commonly used. Since 2010, antenna arrays that generate OAM waves at radio frequencies have been analyzed [98], followed by the traveling-wave antennas [99] and circularly polarized antennas [100] demonstrated to produce OAM waves in radio. In addution, with the proposed concept of metasurface, various prototypes of metasurfaces are designed to radiate EM waves carrying OAM, at both optical and radio regimes [14]. The OAM detection is also an area worthy of study. Although OAM detection is a reciprocal process of OAM generation, it is much more challenging due to the divergence and spatial-dependence nature of an OAM wave. The approaches for

233

234

7 Metamaterial or Metastructural Thin Films for EM Wave Control

OAM detection can be classified into three categories, mode analysis based on field data [101, 102], observation of OAM induced effects such as the rotational Doppler shift [103], and the beam reforming using holographic technology [104].

7.4.2

Generation of Orbital Angular Momentum

Compared with plane waves, the OAM-carrying waves have helical wave fronts, resulting from an additional phase term, eil𝜑 , where 𝜑 is the azimuthal angle. Therefore, by introducing the azimuthal-angle dependent phase delay to the path of plane wave, the plane wave will be modulated to carry an OAM. There are several ways to introduce the phase delay, but generally, they can be categorized as independent of and dependent on the wave polarization. The first scheme employs isotropic materials, such as the SPPs and CGHs. The thickness of an SPP gradually changes along the azimuthal coordinate, proportional to the value of required phase delay, l𝜑, while that for CGHs, they are diffractive optical elements. Different orders of OAM can be observed at the locations according to different diffraction orders [105]. Alternatively, the excitation of an antenna or antenna array can be directly modulated to fulfill the required phase condition so that an OAM wave can be radiated out [106]. Moreover, by utilizing the abrupt phase shift at scatterers or resonators, the desired phase delay can be produced with a reduced device size. For example, by designing the scatterers with different sizes and geometries, a full coverage of 2𝜋 phase shift can be realized [14]. Therefore, by patterning the scatterers according to the required phase distribution, arbitrary beam forming can be achieved, including generating OAM waves [107]. The second scheme is based on the coupling and conversion between SAM and OAM [108] and requires inhomogeneous and anisotropic media, such as q plates. A q plate has a spatially varying optical axis and at the same time, which changes the helicity of circular polarization. Its behavior can be explained by the AM conservation law. The helicity change of the circular polarization indicates a change of ±2ℏ of SAM. When the q plate is cylindrically symmetric, the outcoming wave must carry an OAM of ∓2ℏ to conserve the total AM. When the q plate is not cylindrically symmetric, it introduces extra AM to the system so that different orders of OAM can be generated. q plates are usually implemented using liquid crystals. Recently, metasurfaces have been used to implement the feature of q plates, where anisotropic scatterers are designed to change the helicity of circular polarization and they are placed with different orientations, correlating with the spatially varying optical axis of q plates [109, 110]. Apparently, the phase shift of scatterers depends on their orientations. This phase shift originates from the change in the polarization state along different paths on the Poincare sphere known as geometric phase [111]. Therefore, this type of metasurfaces is also known as geometric-phase metasurface, which is fundamentally different from the metasurfaces adopting the first scheme. 7.4.2.1

Geometric-phase Metasurfaces

In the following, we will introduce several geometric-phase metasurfaces for OAM generation.

7.4 Orbital Angular Momentum of Electromagnetic Waves: Generation and Detection

al

z x

g

as

p

p

y

h (b)

(a) Amplitude

(c)

(d)

Phase

0

1

–π

π

0

1

–π

π

Figure 7.6 Schematic and the response of the ultrathin complementary metasurface. (a) Geometry of the unit cell; (b) Geometry of the metasurface for l = 4; (c, d) Amplitude and phase distributions of the cross-circularly polarized component of the electric field on a transverse plane. The results in (c) are calculated from the equivalent dipole model with the distributed dipole moments shown in (b). The results in (d) is from the full-wave simulation. Source: Chen et al. [112]. © 2017 IEEE.

Figure 7.6 shows an ultrathin complementary metasurface to radiate an OAM wave with high efficiency [112]. The unit cell consists of four complementary split ring resonators (CSRRs) in different sizes and orientations. The fundamental resonance of the larger two CSRRs is excited by y polarization, while that of the smaller two CSRRs is excited by x polarization. By tuning their sizes, the transmitted x and y components can have same amplitude and a 𝜋 phase difference. Therefore, the unit cell transmits the incident circularly polarized wave with a reversed helicity. The double-layer complementary structure guarantees the high transmittance. Then, a whole metasurface is built by arranging the unit cells with varying orientations, as shown in Figure 7.6b. The axial rotation angle for each unit cell, 𝛼 = 2𝜑. So, the phase shift introduced by the unit cell is ±2𝛼 (i.e. ±4𝜑). The OAM order it generates

235

236

7 Metamaterial or Metastructural Thin Films for EM Wave Control

should be ±4. The sign depends on the helicity of incident circular polarization. The EM responses of the metasurface are obtained by two approaches, namely, the equivalent dipole approximation and the full-wave simulation using CST MWS. In the equivalent dipole model, each unit cell is considered as two orthogonal magnetic dipoles as shown in Figure 7.6b (blue and yellow arrows). Then, the EM response of the whole metasurface is calculated by adding up the responses of all the equivalent magnetic dipoles. This approach is useful for analyzing metasurfaces consisting of a large number of scatterers, since these scatterers can be decomposed into point sources with different strengths, locations, and orientations. The calculated field distributions are shown in Figure 7.6c. Clearly, we can observe the phase singularity at the center. The phase changes by 8𝜋 along a circular path around the center, which verifies the successful generation of an OAM of order 4. Full-wave simulation results are shown in Figure 7.6d. The divergence between the results from the dipole model and the full-wave simulation comes from the coupling among unit cells. Other than the transmission-type metasurface, a reflection-type metasurface is also shown below [113]. The metasurface contains an anisotropic perfect electric conductor (PEC) layer on top of an isotropic perfect magnetic conductor (PMC) layer. The top views of two metasurfaces are shown in Figure 7.7. The PEC layers are composed of the red strip arrays, which are metals. For simplicity, the isotropic PMC layers are indicated by the blue surfaces beneath the red strips. Due to a 𝜋 reflection phase difference from a PEC and a PMC layers, the composite structure reflects the incident LCP/right circularly polarized (RCP) plane wave to a RCP/LCP y Phase

Amplitude

𝛼 x 0

1

–π

π

0

1

–π

π

(a)

(b)

Figure 7.7 The composite PEC–PMC metasurfaces for OAM generation. (a) l = 1 and (b) l = 2. The left panels show the top views of the PEC layers. The middle panels and the right panels show the amplitude and phase distributions of the reflected waves on a transverse plane away from the metasurfaces.

7.4 Orbital Angular Momentum of Electromagnetic Waves: Generation and Detection

wave. And the locally different inclination angle between the metal-strip tangent and x axis, 𝛼 introduces a locally modulated phase shift, ±2𝛼. Therefore, to generate an OAM of order l, at each azimuthal location, 𝛼 should be equal to l𝜑/2 so that a phase factor e±il𝜑 can be introduced to the reflected wave. In Figure 7.7a, we make 𝛼 = 𝜑/2 and in Figure 7.7b, 𝛼 = 𝜑. Then, the metasurfaces are illuminated by an LCP Gaussian wave and the amplitude and phase distributions of the reflected RCP field components at a transverse plane away from the metasurfaces are plotted. As expected, in both Figure 7.7a,b, clear phase singularities are observed. We see the field distribution of an OAM wave with l = 1 in Figure 7.7a, because the phase encounters a total 2𝜋 change along a closed path enclosing the center. The total phase change around the center is 4𝜋 in Figure 7.7b, so the reflected wave carries an OAM of order 2. The patterning of the metasurface is accomplished with the assistance of grating vectors [114]. It is different from existing design protocols for geometric-phase-based metasurfaces, where complicated optimization process of single scatterer is needed. Since the PEC layer presents a quasi-continuous pattern, the distortions from local-response discontinuity of discrete scatterers are avoided. Moreover, by keeping the local period small enough, no high-order diffraction exists. 7.4.2.2

Photonic Crystals

In the following, a different scheme based on the superposition of two vibrational modes with proper weights and phases is explored for OAM generation. Particularly, in Figure 7.8, the EM energy is transferred from guided wave in a line defect in a photonic crystal (PC) to localized resonant mode in a point defect and then to unbounded OAM state in free space [115]. For a PC, there exists a band gap where no EM state is allowed to propagate. However, a localized state will be supported at the photonic band gap (PBG) if a defect is introduced [116]. In Figure 7.9a, we show the localized quadrupole modes (quadrupole-xy and quadrupole-diag modes) inside the PBG by increasing the

Figure 7.8 A schematic representation of OAM generation in PCs. Source: Chen et al. [115]. © 2018 American Physical Society.

237

238

7 Metamaterial or Metastructural Thin Films for EM Wave Control

θ = π/2 BE2zeiθ

AE z1

0

+ 0

(a)

1

= 0

1

–π

π

–π

π

θ = – π/2

1

(b)

0

1

Figure 7.9 The quadrupole states (quadrupole-xy and quadrupole-diag states) and the superposed states (amplitude and phase patterns). Ez1 and Ez2 are the electric fields of the two quadrupole modes. A and B are the weights of two modes and 𝜃 is the relative phase of the mode 2 with respect to the mode 1.

radius of a dielectric rod in the PC. It is well known that combining two vibrational modes with a proper phase delay will generate a rotational mode [117]. Therefore, in Figure 7.9b, we superpose the two quadrupole modes and find that pure OAM states with the spatial phase dependence of e±i2𝜑 can be generated when the optimal weights (A = B = 1) and phase (𝜃 = ±𝜋/2) are chosen. To excite the two quadrupole modes simultaneously with the same weights and a phase difference of 𝜋/2, we did optimization in CST MWS. The simulation model is same as in Figure 7.8. To be specific, there are dielectric rods sandwiched between two metallic plates. A row of dielectric rods is removed to form a waveguiding channel. Right above the defect rod, there is a circular opening on the top metallic plate to allow the modes to radiate. The eigenfrequencies of the quadrupole modes are fixed when the size of the defect rod is fixed. Thus, at any frequency between the eigenfrequencies, the mode can be considered as the superposition of the two quadrupole modes. By sweeping the frequency, the weights and relative phase will change dramatically. Therefore, the operating frequency should be carefully chosen. The other parameter that needs to be optimized is the height of the rods, because it directly determines the modes at the air–dielectric interface. By feeding the structure from the other end of the channel, the helicity of the radiated OAM wave will be changed. OAM states of order ±1 can be generated by exciting two dipole modes simultaneously. Experimental results are shown in Figure 7.10. Higher-order OAM waves may be produced by continuously increasing the radius or permittivity of the defect so that higher-order localized modes will appear and can be manipulated.

7.4.3

Detection of Orbital Angular Momentum

As discussed in the introduction part, detection of OAM is another issue that needs to lay emphasis on. In the following, a holographic metasurface is proposed for the detection of multiple OAM beams [118]. The detection process is summarized in Figure 7.11. The metasurface converts the incident wave to a Gaussian wave and the radiation direction of the Gaussian wave is distinguishable according to the order of

7.4 Orbital Angular Momentum of Electromagnetic Waves: Generation and Detection

Defect rod

Probe

Absorber

Horn antenna

y

y

110 mm

–π

(b)

160 mm x

110 mm

Waveguide channel

(a)

–π

π (c)

π

160 mm x

Figure 7.10 The experimental results of the proposed PC. (a) Experimental setup. The phase distributions of E z at a transverse plane 30 mm above the structure with (b) quadrupole defect, measured at the frequency f = 8.75 GHz and (c) dipole defect, measured at the frequency f = 9.55 GHz. Source: (a) Reproduced with permission from Chen et al. [115]. © 2018 American Physical Society, (b, c) Chen et al. [115]. © 2018 American Physical Society.

Figure 7.11 Schematic representation of multiple OAM-beam detection by using a single metasurface.

239

240

7 Metamaterial or Metastructural Thin Films for EM Wave Control

incident OAM. Consequently, by locating the Gaussian wave, the incident OAM can be conveniently determined. To start with, we model the metasurface by a transmittance function: ∑ Am ej(lm 𝜙+kxm x+kym y) (7.28) t(r, 𝜙) = m

where r is the radial position, 𝜑 is the azimuthal position, Am is the weight of the mth beam, lm is the corresponding OAM index, and kxm , kym are the transverse wave numbers of the mth beam. Then, the far-field response of the metasurface illuminated by an incident wave carrying an OAM of order l0 is calculated by doing the Fourier transform: ∑ E = F{Ein ⋅ t} = Am F{EOAM(lm +l0 ) (kxm , kym )} (7.29) m

Therefore, multiple beams are generated and at the designed k-space position (kxm , kym ), the OAM order is lm + l0 . When lm + l0 = 0, the beam is Gaussian and its beam axis is at (kxm , kym ). It is known that OAM wave has a singularity at its beam axis. Therefore, by examining the field intensity at the positions of (kxm , kym ), we can identify the Gaussian beam, i.e. the one that is not null. Then, the incident OAM order l0 can be determined. As a proof of concept, a five-beam case with lm = 2, 1, 0, −1, −2 at the directions of 𝜃 = 40∘ and 𝜑 = 90∘ , 18∘ , 306∘ , 234∘ , 162∘ is demonstrated. The transmittance t(r, 𝜑) is calculated first. Then, its phase information is extracted and implemented using the unit cell shown in Figure 7.6a. Full-wave simulated radiation patterns are shown in Figure 7.12. It can be seen that the maximum radiation directions (axis of the Gaussian beam) when l0 = −2, −1, 0, 1, 2 are at 𝜃 = 40∘ and 𝜑 = 90∘ , 18∘ , 306∘ , 234∘ , 162∘ , respectively, which is as expected. In summary, in this section, we have demonstrated several types of geometric-phase based metasurfaces for OAM generation based on the local phase manipulation. The local phase variation on the metasurfaces can be in discrete and continuous formats. For discrete scatterers, the patterning of the metasurface is straightforward and more flexible. While when the scatterers are quasi-continuous, the patterning of the metasurface is achieved through grating vectors. Additionally, a PC has been designed for OAM generation by introducing line and point defects. Based on the transmission function, a holographic OAM detection method has been proposed and implemented also by using geometric-phase-based metasurfaces.

ϕ−162° ϕ−234°

ϕ−306° ϕ−90°

l = −2

ϕ−18°

l = −1

l=0

l=1

l=2

Figure 7.12 Full-wave simulated far-field power patterns when the incident wave carries OAM of order −2, −1, 0, 1, and 2. Source: Chen et al. [118]. © 2018 IEEE.

7.5 Application in Spontaneous Emission Modification

7.5

Application in Spontaneous Emission Modification

7.5.1 Spontaneous Emission in Inhomogeneous Electromagnetic Environment SE as a fundamental phenomenon of light-matter interaction, is responsible for most of the photon creation process. From daily lighting and displaying, to optical communication, toward quantum information transportation, manipulation of SE is a vital requirement to build up efficient and controllable photonic devices. Since the pioneering work of Purcell in 1946, people open the gate of using macroscale EM environment to control the SE of an atom or molecule. To describe it quantitatively, the concept of spontaneous emission rate (SER) can be introduced by Fermi’s golden rule that how “fast” an quantum emitter transits from an initial state |i⟩ to a final state |f ⟩, and depends on both the atomic structure and density of EM mode of its environment [119] 𝜋𝜔0 2 𝛾= |p| 𝜌(r, 𝜔0 ) (7.30) 3ℏ𝜀0 where p is the transition dipole moment, ℏ is the reduced Planck’s constant, 𝜀0 is vacuum permittivity, and 𝜌(r, 𝜔0 ) is the local density of state (LDOS) of the EM field of frequency 𝜔0 for a quantum emitter located at r, which can be modified by many nanophotonic structures and therefore tune the SER flexibly. Typically, modification factor of the SER is denoted by the Purcell factor 𝛾 Fp = (7.31) 𝛾0 which compares the SER in the studied structure with free space. Currently, one of the most widely studied structures is the PC as shown in Figure 7.13a. Composing by two alternatively arranged materials with different refractive index (namely, n1 and n2 ), PCs introduce the artificial periodic EM environment that analogous to electron crystals. Due to the overlapped destructive interference, optical mode at certain frequency 𝜔0 doesn’t allow to exist along certain direction inside the PC. This frequency region is the so-called PBG where LDOS equals to zero, and therefore, SE is completely inhibited. If the 𝜔0 locates at where optical modes are dense, for example, near the PBG edge, SE will be highly increased. Besides, PC cavity with a point defect can also realize such enhancement effect when the emission mode is resonant with the PC cavity mode. Using single mode approximation, SER can be enhanced by a factor of Q/V (Q is the quality factor of the PC cavity, V is the mode volume). Another kind of the popularly investigated structures for SER modification is the plasmonic nanoparticle (Figure 7.13b). Emission of an atom or molecule close to a metallic nanoparticle is different from the situation in free space. Forster resonant energy transfer theory explains such phenomenon where the energy transfer between the quantum emitter and the closely located nanoparticle can tailor the emission rate, which is a bit similar to the case of two coupled dipoles. The former is the energy exchange process between photon and excited plasmon of the nanoparticle, while the latter is a nonradiative process.

241

242

7 Metamaterial or Metastructural Thin Films for EM Wave Control

z

(b)

(a)

Tip

y gnp

x

z y x

E Excitation 0

1.0

Spectrometer, CCD,APD

Figure 7.13 Typical nanostructure for spontaneous emission modification. (a) Photonic crystal cavity. Source: Englund et al [120]. © 2005 American Physical Society. (b) Plasmonic nano-antenna. Source: Kühn et al. [121]. © 2006 American Physical Society.

7.5.2

Calculation of Spontaneous Emission Rate

For nanostructures of regular geometrical shape such as some cavities and PCs, LDOS, and therefore, the SER might be obtained analytically according to its definition. A more universal method is widely applied based on the dyadic Green’s function by which LDOS can be calculated from the imaginary part of the Green’s tensor at the location of the quantum emitter [119] 6𝜔0 ⟨p ⋅ Im[G(r, r, 𝜔0 )] ⋅ p⟩ (7.32) 𝜋c2 For quantum emitter has a specified dipole orientation, for example, along the y-axis, SER reduces to 𝜌(r, 𝜔0 ) =

𝛾=

2p2 𝜔20 ℏ𝜀0 c2

Im[Gyy (r, r, 𝜔0 )]

(7.33)

While if it does not have a fixed dipole axis, SER should be considered as the average value of all the directions 𝛾=

2p2 𝜔20 3ℏ𝜀0 c2

Im{Tr[G(r, r, 𝜔0 )]}

(7.34)

It is worth to note that this technology is only available for weak coupling situation when atom-field coupling constant 𝜅 is much smaller than 𝛾. Under such assumption, decay rate of a classical dipole emitter and a quantum emitter come to agreement.

7.5.3

Metamaterials Enhance Spontaneous Emission

Possibility of using metamaterials to change SER draws wide attentions in recent years. One of the most famous structure is the hyperbolic metamaterial (HMM),

7.5 Application in Spontaneous Emission Modification

𝜀



>

𝜀

0, 𝜀 < 0 ⃦



> 0, 𝜀⊥< 0

h

MA

R6G/PM

Si

kz

W

Ag

ky

H

kx d

x

(a)

(b)

te

a ubstr

z

ss Glas y

(c)

Figure 7.14 Iso-frequency surfaces for of two kinds of hyperbolic metamaterials: (a) 𝜀zz = 𝜀∥ < 0, 𝜀xx = 𝜀yy = 𝜀⟂ > 0 and (b) 𝜀∥ > 0, 𝜀⟂ < 0. Source: Poddubny et al. [122]. © 2013 Springer Nature. (c) Patterned hyperbolic medium for enhancing spontaneous emission and far field radiation. Source: Lu et al. [123]. © 2014 Springer Nature.

which consists of periodically stacked metal-dielectric layers of subwavelength thickness (1D HMM), or metal nanowires array embedded in the dielectric medium host (2D HMM). Along one direction, HMM behaves like metal, while along the other directions, it is more similar to a dielectric medium. This anisotropic property results in opposite signs of the effective permittivity along different directions 𝜀⟂ • 𝜀// < 0, therefore making this dispersion relation an open hyperbolic curve (Figure 7.14a,b) kx2 + ky2 𝜀⟂

+

kz2 𝜀∕∕

=

𝜔2 c2

(7.35)

which is very different from the isotropic medium for which the above formula is a closed circle. LDOS, which illustrates how “dense” the optical modes accumulate at frequency 𝜔0, can be intuitively represented by the volume between the iso-frequency surfaces at 𝜔0 and 𝜔0 + d𝜔. Theoretically, open hyperbolic shape shell in the k-diagram has infinite large LDOS and is hence significantly higher than ordinary medium with closed spherical or ellipsoid shell. However, this infinite is limited, in practice, by the finite period of the HMM, which prevents the emission to access large k region. Different from cavities where resonating condition is required for obvious SER modification, leading to the drawback of narrow bandwidth, HMM can enhance SER in a wide frequency range due to its hybrid plasmonic mode supported by the multiple-metal-dielectric interfaces and therefore has great potential for emission enhancement (Figure 7.14c).

7.5.4

Metasurfaces Enhance Spontaneous Emission

Compared with 3D bulk metamaterials, practical implementation of metasurface is less demanding and less challenging especially in the optical region. Besides, for some photonic devices such as the light-emitting diodes and solar cells, using a planar metasurface cover instead of changing the whole epitaxy structure can avoid perturbing the inner light emitting/absorption region and thus ease fabrication difficulty.

243

7 Metamaterial or Metastructural Thin Films for EM Wave Control

n0 n

z

Et1

y

X=0 Er1

X1

X2

X3

z=0

Ei1

MS1

d1 Emitter d2

Er2

MS2

Ei2

(b)

backreflector

(a)

3000 2500 2000 1500 1000 500 0 –500 –1000 –1500 –2000 –2500 –3000

3000 2500 2000 1500 1000 500 0 –500 –1000 –1500 –2000 –2500 –3000

x (nm)

x

x (nm)

244

–4000

–2000

0 z(nm)

2000

4000

(c)

–4000

–2000

0 z(nm)

2000

4000

Figure 7.15 Metasurface-cavity enhanced single-photon source. (a) Design of the metasurface cavity, where MS1 is a partially reflective metasurface while MS2 is a completely reflective metasurface. Radiation power of (b) a traditional planar singe-photon source and (c) metasurface-cavity enhanced single-photon source. Source: Chen et al. [124]. © 2017 American Physical Society.

Figure 7.15 shows a double metasurface cavity [124], which has great potential to improve single-photon source by enhancing the SE rate, light extraction efficiency and far field directivity, simultaneously. Schematic view of the single-photon source is shown in Figure 7.15a with a quantum emitter embedded in the semiconductor medium. A partially-reflective metasurface (MS1 in Figure 7.15a) is placed at the upper semiconductor-air interface, which turns the upward incident dipole field Ei1 of all the incident angles into two parts: vertically reflected field Er1 and vertically transmitted field Et1 . Another metasurface (MS2 in Figure 7.15a) is responsible for turning all the downward incident field Ei2 of all the incident angles into vertically reflected field Er2 . Both of the metasurfaces can be synthesized using the susceptibility tensor technology as mentioned above by Eqs. (7.22)–(7.25). For MS1, yy 𝜒ee1 =

x x x 2 Ht1 − (Hi1 + Hr1 ) y y y i𝜔𝜀0 Et1 + (Ei1 + Er1 ) y

xx = 𝜒mm1

y

(7.36)

y

2 Et1 − (Ei1 + Er1 ) x x i𝜔𝜇0 Ht1 + (Hi1x + Hr1 )

(7.37)

and for MS2 yy

x x x 2 (Hi2 + Hr2 ) − Ht2 y y y i𝜔𝜀0 (Ei2 + Er2 ) + Et2

(7.38)

y y y 2 (Ei2 + Er2 ) − Et2 = x x i𝜔𝜇0 (Hi2x + Hr2 ) + Ht2

(7.39)

𝜒ee2 = xx 𝜒mm2

yy

yy

y

xx xx x = 𝜒mm1 = 𝜒ee2 = 𝜒mm2 = 0 and Et2 = Ht2 = 0. Since now all the emitted with 𝜒ee1 wave become vertically propagating inside the semiconductor region, leading to much stronger resonance and field confinement at the dipole location compared with no metasurface situation. SER is, therefore, obviously enhanced by a factor of 1.9 times. Besides, light trapping issue due to the sudden change of refractive index between the semiconductor/air interface is almost resolved due to the vertical emission (Figure 7.15b,c). Far field directivity is also naturally improved by the increased radiation aperture. Given these advantages, metasurface cavity has great potential to develop highly efficient single-photon source with high-single photon purity.

Acknowledgments

7.5.5

Other Potential Application in Quantum Optics

Beyond the application in SE modification of the quantum emitter, metasurface can realize anisotropic response, which could induce quantum interference between two spontaneous decay channels [125]. This is accessible by synthesizing an anisotropic metasurface cavity, which has different response to the dipole emitters polarized along different orientations. For example, enhancing the SER of a y-polarized dipole while suppressing the SER of an x-polarized dipole. The degree of quantum interference typically defined as QI = (𝛾 x − 𝛾 y )/(𝛾 x + 𝛾 y ) can be flexibly controlled by the metasurface. This offers more degrees of freedom to improve and control many quantum-interference-based phenomena such as EM induced transparency and lasing without inversion [126].

7.6

Conclusion and Perspectives

Metastructural film or metasurface, which exploits scatterers to realize the abruptly local change of EM response, plays more and more important roles in engineering control of propagation, scattering, and radiation of EM waves. The capabilities of flexible beam shaping make metastructural films useful in various applications, such as antenna engineering, flat optics, etc. Using metasurfaces improves the performance of antennas in many ways. For example, the Huygens’ metasurface antennas have high directivity and efficiency [127]. The bandwidth of circular polarized antennas can be largely increased using metasurfaces [128], and the size can be miniaturized [129, 130]. Metasurfaces are also employed for reconfigurable design of antennas [131]. The utilization of metasurfaces for antenna designs has achieved specified and advanced functions and becomes one of the most important applications of metasurface. At optical frequencies, metasurfaces can lead to a new class of flat optical components, such as planar lens [132, 133], planar wave plates [134–136], beam transformers, and splitters [137]. By exploring the spin degree of freedom, metasurfaces are also applied in spin-controlled photonics [138]. Furthermore, metasurfaces can replace the complex cascades of optical devices by providing different optical functions simultaneously, which cannot be achieved by conventional optical components [139]. Research on metasurfaces will be extended to nonlinear and quantum regimes, where multiscale and multiphysics modeling and designs should be developed. EM field-matter interaction should be controlled at both microscopic and macroscopic levels by using the metastructural films or metasurfaces.

Acknowledgments This work was supported in part by the Research Grants Council of Hong Kong (GRF 17209918) and National Natural Science Foundation of China (Nos. 61975177, 61701424, and 61571264)

245

246

7 Metamaterial or Metastructural Thin Films for EM Wave Control

List of Abbreviations 3D ACA AIM AM AMC BCG CBF CG–FFT CGHs CSRRs EM FDTD FEM FFT FSS GPU GSTC HDA HMM IE–FFT LCP LDOS LG MBF MLFMA MoM NRI OAM PBC PBG PC PCB PEC PMC RCP SBF SE SER SPPs TFQMR ZIM

three-dimensional Adaptive Cross Approximation adaptive integral method angular momentum artificial magnetic conductor bi-conjugate gradient characteristic basis function conjugate gradient method and the fast Fourier transform computer generated holograms complementary split ring resonators electromagnetic finite-difference time-domain finite-element method Fast Fourier Transform frequency selective surface graphics processing units generalized sheet transition condition hierarchical dipole approximation hyperbolic metamaterial integral equation FFT left circularly polarized local density of state Laguerre–Gaussian macro basis functions multi-level fast multipole algorithm method of moments negative refraction index orbital angular momentum periodic boundary condition photonic band gap photonic crystal printed circuit board perfect electric conductor perfect magnetic conductor right circularly polarized synthetic basis function spontaneous emission spontaneous emission rate spiral phase plates transpose free quasi-minimal residual zero-index metamaterials

References

References 1 Smith, D.R., Pendry, J.B., and Wiltshire, M.C.K. (2004). Metamaterials and negative refractive index. Science 305 (5685): 788–792. 2 Shelby, R.A., Smith, D.R., and Schultz, S. (2001). Experimental verification of a negative index of refraction. Science 292 (5514): 77–79. 3 Zhang, X. and Liu, Z.W. (2008). Superlenses to overcome the diffraction limit. Nature Materials 7 (6): 435–441. 4 Grbic, A. and Eleftheriades, G.V. (2004). Overcoming the diffraction limit with a planar left-handed transmission-line lens. Physical Review Letters 92 (11): 117403. 5 Edwards, B. et al. (2008). Experimental verification of epsilon-near-zero metamaterial coupling and energy squeezing using a microwave waveguide. Physical Review Letters 100 (3): 033903. 6 Hao, J.M., Yan, W., and Qiu, M. (2010). Super-reflection and cloaking based on zero index metamaterial. Applied Physics Letters 96 (10): 101109. 7 Schurig, D., Schurig, D., Mock, J.J. et al. (2006). Metamaterial electromagnetic cloak at microwave frequencies. Science 314 (5801): 977–980. 8 Chen, X. et al. (2011). Three-dimensional broadband and high-directivity lens antenna made of metamaterials. Journal of Applied Physics 110 (4): 044904. 9 Kundtz, N. and Smith, D.R. (2010). Extreme-angle broadband metamaterial lens. Nature Materials 9 (2): 129–132. 10 Ziolkowski, R.W. and Erentok, A. (2006). Metamaterial-based efficient electrically small antennas. IEEE Transactions on Antennas and Propagation 54 (7): 2113–2130. 11 Dong, Y.D. and Itoh, T. (2012). Metamaterial-based antennas. Proceedings of the IEEE 100 (7): 2271–2285. 12 Chin, J.Y., Lu, M.Z., and Cui, T.J. (2008). Metamaterial polarizers by electric-field-coupled resonators. Applied Physics Letters 93 (25): 251903. 13 Chen, M.L.N. et al. (2016). Polarization control by using anisotropic 3-D chiral structures. IEEE Transactions on Antennas and Propagation 64 (11): 4687–4694. 14 Yu, N.F. et al. (2011). Light propagation with phase discontinuities: generalized laws of reflection and refraction. Science 334 (6054): 333–337. 15 Sun, S.L. et al. (2012). Gradient-index meta-surfaces as a bridge linking propagating waves and surface waves. Nature Materials 11 (5): 426–431. 16 Bosiljevac, M. et al. (2012). Non-uniform metasurface luneburg lens antenna design. IEEE Transactions on Antennas and Propagation 60 (9): 4065–4073. 17 Pfeiffer, C. and Grbic, A. (2013). Metamaterial Huygens’ surfaces: tailoring wave fronts with reflectionless sheets. Physical Review Letters 110 (19): 197401. 18 Menzel, C., Rockstuhl, C., and Lederer, F. (2010). Advanced Jones calculus for the classification of periodic metamaterials. Physical Review A 82 (5): 053811. 19 Idemen, M.M. (2011). Discontinuities in the Electromagnetic Field. Hoboken, New Jersey, USA: Wiley-IEEE Press.

247

248

7 Metamaterial or Metastructural Thin Films for EM Wave Control

20 Niemi, T., Karilainen, A.O., and Tretyakov, S.A. (2013). Synthesis of polarization transformers. IEEE Transactions on Antennas and Propagation 61 (6): 3102–3111. 21 Achouri, K., Salem, M.A., and Caloz, C. (2015). General metasurface synthesis based on susceptibility tensors. IEEE Transactions on Antennas and Propagation 63 (7): 2977–2991. 22 Tretyakov, S. (2003). Analytical Modeling in Applied Electromagnetics. Boston, MA: Artech House. 23 Asadchy, V.S. et al. (2016). Perfect control of reflection and refraction using spatially dispersive metasurfaces. Physical Review B 94 (7): 075142. 24 Feresidis, A.P. et al. (2005). Artificial magnetic conductor surfaces and their application to low-profile high-gain planar antennas. IEEE Transactions on Antennas and Propagation 53 (1): 209–215. 25 Sarabandi, K. and Behdad, N. (2007). A frequency selective surface with miniaturized elements. IEEE Transactions on Antennas and Propagation 55 (5): 1239–1245. 26 Huang, J. (2007). Reflectarray Antenna. Wiley Online Library. 27 Abdelrahman, A.H., Elsherbeni, A.Z., and Yang, F. (2014). Transmitarray antenna design using cross-slot elements with no dielectric substrate. IEEE Antennas and Wireless Propagation Letters 13: 177–180. 28 Ni, X. et al. (2012). Broadband light bending with plasmonic nanoantennas. Science 335 (6067): 427–427. 29 Chen, Q.-Y. et al. (2012). Low-profile wideband reflectarray by novel elements with linear phase response. IEEE Antennas and Wireless Propagation Letters 11: 1545–1547. 30 Deng, R. et al. (2017). An FSS-backed 20/30-GHz dual-band circularly polarized reflectarray with suppressed mutual coupling and enhanced performance. IEEE Transactions on Antennas and Propagation 65 (2): 926–931. 31 Edalati, A. and Sarabandi, K. (2012). Wideband reflectarray antenna based on miniaturized element frequency selective surfaces. 2012 6th European Conference on Antennas and Propagation (EUCAP), IEEE. 32 Nayeri, P., Yang, F., and Elsherbeni, A.Z. (2012). Design and experiment of a single-feed quad-beam reflectarray antenna. IEEE Transactions on Antennas and Propagation 60 (2): 1166–1171. 33 Zhou, M. et al. (2013). Direct optimization of printed reflectarrays for contoured beam satellite antenna applications. IEEE Transactions on Antennas and Propagation 61 (4): 1995–2004. 34 Chen, M., Jiang, L.J., and Wei, E. (2017). Detection of orbital angular momentum with metasurface at microwave band. IEEE Antennas and Wireless Propagation Letters 17 (1): 110–113. 35 Yifat, Y. et al. (2014). Highly efficient and broadband wide-angle holography using patch-dipole nanoantenna reflectarrays. Nano Letters 14 (5): 2485–2490. 36 Yang, X. et al. (2017). A broadband high-efficiency reconfigurable reflectarray antenna using mechanically rotational elements. IEEE Transactions on Antennas and Propagation 65 (8): 3959–3966.

References

37 Yang, H. et al. (2017). A 1600-element dual-frequency electronically reconfigurable reflectarray at X/Ku-band. IEEE Transactions on Antennas and Propagation 65 (6): 3024–3032. 38 Yang, X. et al. (2016). Design of an amplifying reflectarray antenna with improved isolation performance. 2016 IEEE International Symposium on Antennas and Propagation (APSURSI), IEEE. 39 Sherrott, M.C. et al. (2017). Experimental demonstration of >230∘ phase modulation in gate-tunable graphene–gold reconfigurable mid-infrared metasurfaces. arXiv preprint arXiv:1701.08221. 40 Pozar, D. and Metzler, T. (1993). Analysis of a reflectarray antenna using microstrip patches of variable size. Electronics Letters 29 (8): 657–658. 41 Encinar, J.A. (2001). Design of two-layer printed reflectarrays using patches of variable size. IEEE Transactions on Antennas and Propagation 49 (10): 1403–1410. 42 Liu, T. et al. (2017). Phase error analysis for reflectarray antennas based on study of quasi-periodic effect. 2017 11th European Conference on Antennas and Propagation (EUCAP), IEEE. 43 Gibson, W.C. (2014). The Method of Moments in Electromagnetics. CRC Press. 44 Volakis, J.L., Chatterjee, A., and Kempel, L.C. (1998). Finite Element Method Electromagnetics: Antennas, Microwave Circuits, and Scattering Applications, vol. 6. Wiley. 45 Elsherbeni, A.Z. and Demir, V. (2016). The Finite-Difference Time-Domain Method for Electromagnetics with MATLAB Simulations. The Institution of Engineering and Technology. 46 Killian, T.N., Rao, S.M., and Baginski, M.E. (2011). Electromagnetic scattering from electrically large arbitrarily-shaped conductors using the method of moments and a new null-field generation technique. IEEE Transactions on Antennas and Propagation 59 (2): 537–545. 47 Xiong, X.Y. et al. (2015). Efficient calculation of large finite periodic structures based on surface wave analysis. IEEE Transactions on Antennas and Propagation 63 (1): 69–80. 48 Karami, H., Sheshyekani, K., and Rachidi, F. (2017). Mixed-potential integral equation for full-wave modeling of grounding systems buried in a lossy multilayer stratified ground. IEEE Transactions on Electromagnetic Compatibility 59 (5): 1505–1513. 49 Warnick, K.F. and Chew, W.C. (2001). Numerical simulation methods for rough surface scattering. Waves in Random Media 11 (1): R1–R30. 50 Song, J., Lu, C.-C., and Chew, W.C. (1997). Multilevel fast multipole algorithm for electromagnetic scattering by large complex objects. IEEE Transactions on Antennas and Propagation 45 (10): 1488–1493. 51 Kurz, S., Rain, O., and Rjasanow, S. (2002). The adaptive cross-approximation technique for the 3D boundary-element method. IEEE Transactions on Magnetics 38 (2): 421–424. 52 Peng, Z., Wang, X.-C., and Lee, J.-F. (2011). Integral equation based domain decomposition method for solving electromagnetic wave scattering from

249

250

7 Metamaterial or Metastructural Thin Films for EM Wave Control

53

54

55

56

57

58

59

60

61

62

63 64

65

66

non-penetrable objects. IEEE Transactions on Antennas and Propagation 59 (9): 3328–3338. De Donno, D. et al. (2010). Introduction to GPU computing and CUDA programming: a case study on FDTD [EM programmer’s notebook]. IEEE Antennas and Propagation Magazine 52 (3): 116–122. Velamparambil, S. et al. (1998). ScaleME: a portable scaleable multipole engine for electromagnetic and acoustic integral equation solvers. Antennas and Propagation Society International Symposium, IEEE. Jin, J.-M. and Volakis, J.L. (1990). Electromagnetic scattering by a perfectly conducting patch array on a dielectric slab. IEEE Transactions on Antennas and Propagation 38 (4): 556–563. Francavilla, M.A. et al. (2015). On the numerical simulation of metasurfaces with impedance boundary condition integral equations. IEEE Transactions on Antennas and Propagation 63 (5): 2153–2161. Holloway, C.L. et al. (2012). Use of generalized sheet transition conditions to model guided waves on metasurfaces/metafilms. IEEE Transactions on Antennas and Propagation 60 (11): 5173–5186. Holloway, C.L., Kuester, E.F., and Dienstfrey, A. (2011). Characterizing metasurfaces/metafilms: the connection between surface susceptibilities and effective material properties. IEEE Antennas and Wireless Propagation Letters 10: 1507–1511. Chamanara, N., Achouri, K., and Caloz, C. (2017). Efficient analysis of metasurfaces in terms of spectral-domain GSTC integral equations. arXiv preprint arXiv:1704.06801. Matekovits, L. (2010). Analytically expressed dispersion diagram of unit cells for a novel type of holographic surface. IEEE Antennas and Wireless Propagation Letters 9: 1251–1254. Van Orden, D. and Lomakin, V. (2011). Fundamental electromagnetic properties of twisted periodic arrays. IEEE Transactions on Antennas and Propagation 59 (8): 2824–2833. Salary, M.M., Forouzmand, A., and Mosallaei, H. (2016). Model order reduction of large-scale metasurfaces using a hierarchical dipole approximation. ACS Photonics 4 (1): 63–75. Quarteroni, A. and Rozza, G. (2014). Reduced Order Methods for Modeling and Computational Reduction, vol. 9. Springer. González-Ovejero, D. and Maci, S. (2015). Gaussian ring basis functions for the analysis of modulated metasurface antennas. IEEE Transactions on Antennas and Propagation 63 (9): 3982–3993. Bodehou, M. et al. (2018). Fourier–Bessel basis functions for the analysis of elliptical domain metasurface antennas. IEEE Antennas and Wireless Propagation Letters 17 (4): 675–678. Dong, T., Ma, X., and Mittra, R. (2013). Modeling large nonuniform optical antenna arrays for metasurface application. Journal of Applied Physics 114 (4): 043103.

References

67 Prakash, V.V.S. and Mittra, R. (2003). Characteristic basis function method: a new technique for efficient solution of method of moments matrix equations. Microwave and Optical Technology Letters 36 (2): 95–100. 68 De Vita, P. et al. (2013). Metasurface-by-design: an emerging concept for antennas and beam forming networks. 2013 IEEE–APS Topical Conference on Antennas and Propagation in Wireless Communications (APWC), IEEE. 69 Zhang, B. et al. (2010). Analyzing large-scale non-periodic arrays with synthetic basis functions. IEEE Transactions on Antennas and Propagation 58 (11): 3576–3584. 70 Ansari Oghol Beig, D. and Mosallaei, H. (2012). A macrobasis function model for characterizing finite large-area metamaterials. Progress in Electromagnetics Research 43: 211–238. ´ I. and Mosig, J.R. (2004). Efficient evaluation of 71 Stevanovic, macro-basis-function reaction terms in the subdomain multilevel approach. Microwave and Optical Technology Letters 42 (2): 138–143. 72 Freni, A., De Vita, P., and Bandinelli, M. (2014). A hybrid ACA–MLayAIM technique for the analysis of metasurfaces. 2014 8th European Conference on Antennas and Propagation (EuCAP), IEEE. 73 Ansari-Oghol-Beig, D. and Mosallaei, H. (2015). Array integral equation–fast Fourier transform solver for simulation of supercells and aperiodic penetrable metamaterials. Journal of Computational and Theoretical Nanoscience 12 (10): 3864–3878. 74 Börm, S., Grasedyck, L., and Hackbusch, W. (2003). Introduction to hierarchical matrices with applications. Engineering Analysis with Boundary Elements 27 (5): 405–422. 75 Seo, S.M. and Lee, J.-F. (2005). A fast IE–FFT algorithm for solving PEC scattering problems. IEEE Transactions on Magnetics 41 (5): 1476–1479. 76 De Vita, P. et al. (2006). Fast analysis of large finite arrays with a combined multiresolution – SM/AIM approach. IEEE Transactions on Antennas and Propagation 54 (12): 3827–3832. 77 Bleszynski, E., Bleszynski, M., and Jaroszewicz, T. (1996). AIM: adaptive integral method for solving large-scale electromagnetic scattering and radiation problems. Radio Science 31 (5): 1225–1251. 78 Sarkar, T., Arvas, E., and Rao, S. (1986). Application of FFT and the conjugate gradient method for the solution of electromagnetic radiation from electrically large and small conducting bodies. IEEE Transactions on Antennas and Propagation 34 (5): 635–640. 79 Catedra, M.F., Cuevas, J.G., and Nuno, L. (1988). A scheme to analyze conducting plates of resonant size using the conjugate-gradient method and the fast Fourier transform. IEEE Transactions on Antennas and Propagation 36 (12): 1744–1752. 80 Catedra, M.F., Gago, E., and Nuno, L. (1989). A numerical scheme to obtain the RCS of three-dimensional bodies of resonant size using the conjugate gradient method and the fast Fourier transform. IEEE Transactions on Antennas and Propagation 37 (5): 528–537.

251

252

7 Metamaterial or Metastructural Thin Films for EM Wave Control

81 Gan, H. and Chew, W.C. (1995). A discrete BCG–FFT algorithm for solving 3D inhomogeneous scatterer problems. Journal of Electromagnetic Waves and Applications 9 (10): 1339–1357. 82 Zhang, Z. and Liu, Q. (2001). Three-dimensional weak-form conjugate-and biconjugate-gradient FFT methods for volume integral equations. Microwave and Optical Technology Letters 29 (5): 350–356. 83 Wang, C. and Jin, J. (1998). Simple and efficient computation of electromagnetic fields in arbitrarily shaped inhomogeneous dielectric bodies using transpose-free QMR and FFT. IEEE Transactions on Microwave Theory and Techniques 46 (5): 553–558. 84 Zhang, Z.Q., Liu, Q.H., and Xu, X.M. (2003). RCS computation of large inhomogeneous objects using a fast integral equation solver. IEEE Transactions on Antennas and Propagation 51 (3): 613–618. 85 Chen, R., Fan, Z., and Yung, E.K. (2003). Analysis of electromagnetic scattering of three-dimensional dielectric bodies using Krylov subspace FFT iterative methods. Microwave and Optical Technology Letters 39 (4): 261–267. 86 Bliokh, K.Y., Bekshaev, A.Y., and Nori, F. (2013). Dual electromagnetism: helicity, spin, momentum and angular momentum. New Journal of Physics 15: 033026. 87 Allen, L. et al. (1992). Orbital angular-momentum of light and the transformation of Laguerre–Gaussian laser modes. Physical Review A 45 (11): 8185–8189. 88 Yan, Y. et al. (2014). High-capacity millimetre-wave communications with orbital angular momentum multiplexing. Nature Communications 5: 4876. 89 Wang, J. et al. (2012). Terabit free-space data transmission employing orbital angular momentum multiplexing. Nature Photonics 6 (7): 488–496. 90 Thide, B. et al. (2007). Utilization of photon orbital angular momentum in the low-frequency radio domain. Physical Review Letters 99 (8): 087701. 91 Bozinovic, N. et al. (2013). Terabit-scale orbital angular momentum mode division multiplexing in fibers. Science 340 (6140): 1545–1548. 92 Tamburini, F. et al. (2006). Overcoming the Rayleigh criterion limit with optical vortices. Physical Review Letters 97 (16): 163903. 93 Grier, D.G. (2003). A revolution in optical manipulation. Nature 424 (6950): 810–816. 94 Lavery, M.P.J. (2013). Detection of a spinning object using light’s orbital angular momentum (vol 341, pg 537, 2013). Science 341 (6151): 1175–1175. 95 Beijersbergen, M. et al. (1994). Helical-wavefront laser beams produced with a spiral phaseplate. Optics Communications 112 (5–6): 321–327. 96 Bazhenov, V.Y., Vasnetsov, M.V., and Soskin, M.S. (1990). Laser-beams with screw dislocations in their wave-fronts. JETP Letters 52 (8): 429–431. 97 Marrucci, L., Manzo, C., and Paparo, D. (2006). Optical spin-to-orbital angular momentum conversion in inhomogeneous anisotropic media. Physical Review Letters 96 (16): 163905. 98 Mohammadi, S.M. et al. (2010). Orbital angular momentum in radio-a system study. IEEE Transactions on Antennas and Propagation 58 (2): 565–572.

References

99 Zheng, S.L. et al. (2015). Transmission characteristics of a twisted radio wave based on circular traveling-wave antenna. IEEE Transactions on Antennas and Propagation 63 (4): 1530–1536. 100 Barbuto, M. et al. (2014). Circular polarized patch antenna generating orbital angular momentum. Progress in Electromagnetics Research-Pier 148: 23–30. 101 Mohammadi, S.M. et al. (2010). Orbital angular momentum in radio: measurement methods. Radio Science 45: RS4007. 102 Schulze, C. et al. (2013). Measurement of the orbital angular momentum density of light by modal decomposition. New Journal of Physics 15: 073025. 103 Courtial, J. et al. (1998). Measurement of the rotational frequency shift imparted to a rotating light beam possessing orbital angular momentum. Physical Review Letters 80 (15): 3217–3219. 104 Genevet, P. et al. (2012). Holographic detection of the orbital angular momentum of light with plasmonic photodiodes. Nature Communications 3: 1278. 105 Moreno, I. et al. (2009). Vortex sensing diffraction gratings. Optics Letters 34 (19): 2927–2929. 106 Liu, K. et al. (2016). Generation of OAM beams using phased Array in the microwave band. IEEE Transactions on Antennas and Propagation 64 (9): 3850–3857. 107 Yu, S.X. et al. (2016). Generating multiple orbital angular momentum vortex beams using a metasurface in radio frequency domain. Applied Physics Letters 108 (24): 241901. 108 Bliokh, K.Y. et al. (2015). Spin-orbit interactions of light. Nature Photonics 9 (12): 796–808. 109 Kang, M. et al. (2012). Wave front engineering from an array of thin aperture antennas. Optics Express 20 (14): 15882–15890. 110 Karimi, E. et al. (2014). Generating optical orbital angular momentum at visible wavelengths using a plasmonic metasurface. Light-Science & Applications 3: e167. 111 Berry, M.V. (1987). The adiabatic phase and pancharatnam phase for polarized-light. Journal of Modern Optics 34 (11): 1401–1407. 112 Chen, M.L.L.N., Jiang, L.J., and Sha, W.E.I. (2017). Ultrathin complementary metasurface for orbital angular momentum generation at microwave frequencies. IEEE Transactions on Antennas and Propagation 65 (1): 396–400. 113 Chen, M.L.N., Jiang, L.J., and Sha, W.E.I. (2016). Artificial perfect electric conductor-perfect magnetic conductor anisotropic metasurface for generating orbital angular momentum of microwave with nearly perfect conversion efficiency. Journal of Applied Physics 119 (6): 064506. 114 Chen, M.L.L.N., Jiang, L.J., and Sha, W.E.I. (2019). Quasi-continuous metasurfaces for orbital angular momentum generation. IEEE Antennas and Wireless Propagation Letters 18 (3): 477–481. 115 Chen, M.L.L.N., Jiang, L.J., and Sha, W.E.I. (2018). Generation of orbital angular momentum by a point defect in photonic crystals. Physical Review Applied 10 (1): 014034.

253

254

7 Metamaterial or Metastructural Thin Films for EM Wave Control

116 Yablonovitch, E. et al. (1991). Donor and acceptor modes in photonic band-structure. Physical Review Letters 67 (24): 3380–3383. 117 Landau, L.D. and Lifshitz, E.M. (1958). Quantum Mechanics: Non-relativistic Theory. Pergamon Press. 118 Chen, M.L.L.N., Jiang, L.J., and Sha, W.E.I. (2018). Detection of orbital angular momentum with metasurface at microwave band. IEEE Antennas and Wireless Propagation Letters 17: 110–113. 119 Novotny, L. and Hecht, B. (2012). Principles of Nano-Optics. Cambridge: Cambridge University Press. 120 Englund, D. et al. (2005). Controlling the spontaneous emission rate of single quantum dots in a two-dimensional photonic crystal. Physical Review Letters 95 (1): 013904. 121 Kühn, S. et al. (2006). Enhancement of single-molecule fluorescence using a gold nanoparticle as an optical nanoantenna. Physical Review Letters 97 (1): 017402. 122 Poddubny, A. et al. (2013). Hyperbolic metamaterials. Nature Photonics 7: 948. 123 Lu, D. et al. (2014). Enhancing spontaneous emission rates of molecules using nanopatterned multilayer hyperbolic metamaterials. Nature Nanotechnology 9: 48. 124 Chen, L. et al. (2017). Simultaneous enhancement of light extraction and spontaneous emission using a partially reflecting metasurface cavity. Physical Review A 95 (5): 053808. 125 Jha, P.K. et al. (2015). Metasurface-enabled remote quantum interference. Physical Review Letters 115 (2): 025501. 126 Scully, M.O. and Zubairy, M.S. (1997). Quantum Optics. Cambridge: Cambridge University Press. 127 Epstein, A., Wong, J.P.S., and Eleftheriades, G.V. (2016). Cavity-excited Huygens’ metasurface antennas for near-unity aperture illumination efficiency from arbitrarily large apertures. Nature Communications 7: 10360. 128 Nasimuddin, N., Chen, Z.N., and Qing, X.M. (2016). Bandwidth enhancement of a single-feed circularly polarized antenna using a metasurface metamaterial-based wideband CP rectangular microstrip antenna. IEEE Antennas and Propagation Magazine 58 (2): 39–46. 129 Dong, Y.D., Toyao, H., and Itoh, T. (2012). Design and characterization of miniaturized patch antennas loaded with complementary split-ring resonators. IEEE Transactions on Antennas and Propagation 60 (2): 772–785. 130 Dong, Y.D., Toyao, H., and Itoh, T. (2011). Compact circularly-polarized patch antenna loaded with metamaterial structures. IEEE Transactions on Antennas and Propagation 59 (11): 4329–4333. 131 Kandasamy, K. et al. (2015). Low-RCS and polarization-reconfigurable antenna using cross-slot-based metasurface. IEEE Antennas and Wireless Propagation Letters 14: 1638–1641. 132 Aieta, F. et al. (2012). Aberration-free ultrathin flat lenses and axicons at telecom wavelengths based on plasmonic metasurfaces. Nano Letters 12 (9): 4932–4936.

References

133 Ni, X.J. et al. (2013). Ultra-thin, planar, babinet-inverted plasmonic metalenses. Light-Science & Applications 2: e72. 134 Drezet, A., Genet, C., and Ebbesen, T.W. (2008). Miniature plasmonic wave plates. Physical Review Letters 101 (4): 043902. 135 Khoo, E.H., Li, E.P., and Crozier, K.B. (2011). Plasmonic wave plate based on subwavelength nanoslits. Optics Letters 36 (13): 2498–2500. 136 Yu, N.F. et al. (2012). A broadband, background-free quarter-wave plate based on plasmonic metasurfaces. Nano Letters 12 (12): 6328–6333. 137 Farmahini-Farahani, M. and Mosallaei, H. (2013). Birefringent reflectarray metasurface for beam engineering in infrared. Optics Letters 38 (4): 462–464. 138 Yin, X.B. et al. (2013). Photonic spin hall effect at metasurfaces. Science 339 (6126): 1405–1407. 139 Yu, N.F. and Capasso, F. (2014). Flat optics with designer metasurfaces. Nature Materials 13 (2): 139–150.

255

257

8 Semiconductor Thin Films for Information Technology Na Chen Tsinghua University, School of Materials Science and Engineering, Key Laboratory for Advanced Materials Processing Technology (MOE), Haidian district, Beijing 100084, China

8.1 Introduction Information technology (IT) has changed the world by revolutionizing the manner in which people live [1–3]. The core of IT is the use of computers to perceive, transmit, process, and store information. Semiconductors enable the control of charge carriers for information detection, logical operation, and data storage, which lay foundation of the current IT. In particular, Si-based semiconductors are the key materials for fabricating microprocessor and random-access memory in computers. Except for the single-crystalline Si chips applied in integrated circuits, semiconductors in the form of thin film have played an important role in optics and electronics for IT. For example, thin film transistors (TFTs) based on the hydrogenated amorphous Si (a-Si:H) enable the development of large-area flat display screens for laptop computers and mobile phones [4], which are now indispensable in our daily life. Besides a-Si:H, new amorphous semiconductor thin films from an In–Ga–Zn–O system have been developed and exhibited high carrier mobility that can be used for transparent electronics [5]. Although Si-based semiconductors can be used in random-access memory for data store, the memory is volatile. Once power is turned off, the stored information is lost. New device concepts have been proposed to develop nonvolatile memory. Among them, there are two leading candidates. One is the nonvolatile phase-change random-access memory based on chalcogenide semiconductors [6, 7]. These chalcogenide phase-change semiconductor thin films show significant resistance difference between their amorphous and crystalline states, which can be utilized to encode digital information. The other is the magnetic semiconductor (MS)-based spintronic device with nonvolatility, low power dissipation, and rapid response [8–10]. These MS thin films are capable of manipulating both the charge and spin of electrons, thereby realizing two functionalities of data processing and data storage at the same time. Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications, First Edition. Edited by Yujun Song. © 2021 WILEY-VCH GmbH. Published 2021 by WILEY-VCH GmbH.

258

8 Semiconductor Thin Films for Information Technology

With ever-growing necessity for miniaturization of electronic devices for low power consumption and high speed, the search for new materials with a combination of desirable functionalities has spurred the development of advanced semiconductor thin films. These newly developed semiconductor thin films are revolutionizing the current IT, which may partly replace traditional Si-based semiconductors by holding promises for applications in next-generation devices of IT. In this chapter, we aim to focus on the currently used semiconductor thin films for IT by covering the preparation techniques, structure and property characterization, and the electronic devices based on them. In addition, we will introduce the progress made on the newly developed semiconductors including the chalcogenide phase-change materials and MS thin films.

8.2 Fabrication of Semiconductor Thin Films Semiconductor thin films including Si and III–V semiconductors have been widely used in optics and electronics. To prepare semiconductor thin films with reproducible, controlled, and well-defined properties, a variety of deposition processes have been developed as listed in Figure 8.1. Among them, the most common and important ones comprise physical vapor deposition (PVD) including vacuum evaporation and sputtering and chemical vapor deposition (CVD) including metal–organic chemical vapor deposition (MOCVD) and plasma-enhanced chemical vapor deposition (PECVD) [11]. This section focuses on several representative

Semicenductor thin film deposition

Physical process

Chemical process

Vacuum evaporation

Sputter deposition

Resistive heating

Direct current

MOCVD

Electron beam

Radio frequency

PECVD

Laser ablation

Magnetron

MBE

Ion beam

Figure 8.1

Thermal growth

Sol–gel

CVD

Classification of different processes for semiconductor thin film deposition.

8.2 Fabrication of Semiconductor Thin Films

UHV chamber

In situ analysis systems including REED, AES, etc. Sensors

Substrate preparation

Substrate stage Shutter

Single-crystal GaAs substrate

Sn

n M

Al Ga

Figure 8.2

As

Schematic of the MBE system.

deposition techniques in use today. Topics discussed cover molecular beam epitaxy (MBE), magnetron sputtering, and MOCVD.

8.2.1

Molecular Beam Epitaxy (MBE)

MBE is a special and sophisticated epitaxial thin film growth technique, which requires an ultrahigh vacuum (UHV) environment usually greater than 6 × 10−9 Pa (5 × 10−11 Torr) [11]. Figure 8.2 shows schematic of the MBE system, consisting of UHV chambers, substrate preparation, thin film growth and analysis, and sample exchange load-lock chamber. MBE is basically a thermal evaporation process performed under UHV. The evaporated beams of atoms or molecules of the source materials react with the heated crystalline substrate to form the epitaxial layer. Since the early work on MBE-grown thin films of GaAs and related III–V compounds [12, 13], MBE is now a well-established epitaxial process of major importance in the development of optoelectronic and microelectronic devices [14]. The real-time and in situ monitoring of the MBE system allows the deposition of high-quality single-crystal semiconductor thin films with precise control of composition, growth condition, and doping. Owing to its technological advantages, MBE was used to develop new functional materials, leading to a breakthrough in creating III–V-based diluted magnetic semiconductors (DMSs) such as (In,Mn)As and (Ga,Mn)As [15, 16]. The details regarding the DMSs are discussed in a later section.

8.2.2

Magnetron Sputtering

Magnetron sputtering technology is a kind of PVD. Since its development, this technique has made significant progress for deposition of various semiconductor thin

259

260

8 Semiconductor Thin Films for Information Technology

t

Si substrate t t

t t

t

t

t

E Plasma e-

Ar

+

t

t

Ar

+

e-

Target

N

Figure 8.3

S

N

Schematic of magnetron sputtering process.

films and the related microelectronic devices. A typical magnetron sputter deposition system consists of a vacuum chamber, a sputter target, a substrate holder, and a pumping system. Figure 8.3 shows schematic of the sputtering process. In the vacuum chamber, an inert gas such as Ar is filled in the chamber to create a neutral gas condition at low pressure. The target is mounted on a cathode, while the substrate is connected to an anode. A glow discharge is initiated by applying a high voltage on the order of 103 V between them. The positive Ar ions are accelerated by the electric field before they bombard on the surface of the target. Owing to the bombardment of these high velocity ions, the surface atoms of the target are ejected and condense on the substrate by forming a thin film. The magnetron effect means the use of a closed drift path of crossed magnetic and electric fields to trap electrons in. In the regions of the electron trapping, the efficiency of the ionization process is enhanced, which allows the plasma to be generated at lower pressures. As a result, the energy loss of the ejected atoms due to the gas collisions is reduced, thereby enabling the deposition of high-quality thin films at high deposition rates.

8.2.3

Metal–Organic Chemical Vapor Deposition (MOCVD)

Since the first demonstration of single-crystal films of GaAs grown by MOCVD, MOCVD has been widely used for preparing high-quality semiconductor epitaxial thin films applied in optoelectronics and microwave devices [14, 17]. In particular, MOCVD is the main technique for growing light-emitting diode (LED) epitaxial wafers, which become the cornerstone of the current semiconductor lighting technologies. Figure 8.4 shows a schematic of the deposition process. MOCVD provides an efficient way for preparing single-crystal semiconductor thin films with precise control of stoichiometric composition, doping level, and purity.

8.3 Nonmagnetic Semiconductor Thin Films and Typical Applications

Deposition chamber Gas sources

Ga(CH3)3 + AsH3

GaAs + 3CH4

Ga(CH3)3 + NH3

GaN + 3CH4

GaAs Heating stage

Figure 8.4

To pump

Schematic of chemical vapor deposition process.

8.3 Nonmagnetic Semiconductor Thin Films and Typical Applications Semiconductor thin films including Si in amorphous and crystalline states and compound semiconductors play essential roles in optical and electronic devices currently used for IT. In particular, further advances of light-emitting semiconductor devices, displays, and memories have evolved in important research fields with a focus on creating new semiconductor materials and developing innovation technologies. These traditional semiconductors are usually nonmagnetic. The unique optical and electrical properties of these traditional semiconductor thin films are utilized in applications of optic-electrical devices. This section discusses nonmagnetic semiconductor thin films such as GaAs, amorphous Si:H, IGZO, and chalcogenide semiconductors toward the uses in light-emitting devices, TFTs, and nonvolatile memories.

8.3.1

Semiconductor Thin Films for Light-emitting Devices

Semiconductors lay foundation of the current IT, mainly attributed to their capability of being doped into p-type or n-type semiconductors with tunable electrical and optical properties. The semiconductors of p-type and n-type are integrated to form a p–n junction, which is the basic functional component in light-emitting devices including LEDs and laser diodes (LDs). These devices are the key for information display and information storage. A LED consists of multiple layers of semiconductors. When current flows through the LED in the forward bias, the electrons and holes combine to release energy in the active layer, accompanied by the emission of photons. The color of the emitted light can be red, orange, yellow, green, or blue, depending on the bandgap of the semiconductors used in the LEDs (see Figure 8.5). Such a semiconductor electroluminescence effect was firstly discovered by an Englishman, Captain Henry Joseph Round, in 1907. Later in 1927, a Russian scientist, Oleg Vladimirovich Losev, also observed the light emission from the silicon carbide-based diode. This experiment

261

8 Semiconductor Thin Films for Information Technology

Electron energy

262

e–

Conduction band e– e– e–

e– hv

Eg h+

h+

h+

h+

Valence band

h+

Color of light

Wavelenght (nm)

Energy hv(eV)

Blue

480

2.6

Green

530

2.3

Yellow

580

2.1

Orange

610

2.0

Red

680

1.8

Figure 8.5 The emitted light with colors depending on the bandgap energy of the used semiconductors.

can be recognized as the creation of the first LED. Till the 1970s, LEDs have become commercially successful products. Now LEDs are widely used in our daily life, which is largely attributed to the discovery of high-brightness blue diodes in the 1990s. The invention of blue diodes has enabled the development of white LEDs, which are rapidly replacing conventional lighting sources in the 2000s. Compared with the incandescent bulb-based lights, these white LEDs have improved visibility, lower electric power consumption, and longer lifespan. Meanwhile, the discovery of the blue diodes leads to the development of blue LDs, which enabled the development of the large-capacity optical storage Blu-ray disks. The breakthrough made in the development of blue diodes is based on the preparation of high-quality p-type GaN-based semiconductor thin films by using MOCVD technique [18, 19]. Although the electroluminescence effect was firstly observed in silicon carbide, silicon carbide has an indirect transition band structure. The energy efficiency of the silicon carbide-based LEDs is thus very low, posing significant hurdle for the development of blue diodes. Another potential candidate is zinc selenide, which has a suitable bandgap for the development of blue LEDs. Unfortunately, zinc selenide-based LEDs and LDs could not be commercialized due to their short lifespan. Research on GaN started in the 1970s. Despite this, there existed two main challenges that required to be overcome before a possible blue diode using GaN could be fabricated. One was the preparation of high-quality GaN thin films with good crystal quality and uniform thickness, and the other was the preparation of p-type GaN thin films. Three Japanese scientists Isamu Akasaki, Hiroshi Amano, and Shuji Nakamura have made the major contribution to the development and commercialization of blue LEDs and blue LDs by overcoming the challenges through technological innovations. Figure 8.6 shows a schematic of the first blue LED based on a p-type Mg-doped GaN thin film [18], which exhibited the typical I–V characteristics for p–n junctions. The p-type GaN thin film doped with Mg was produced by low-energy electron-beam irradiation treatment. The hole concentration and the resistivity of this p-type GaN:Mg were not sufficient for further developing blue LDs and high-brightness blue LEDs. Later, low resistivity p-type GaN films were obtained by N2 -ambient thermal annealing at temperatures above 700 ∘ C. The invention of the blue diodes enabled the completion of the light spectrum for semiconductor devices, resulting in the development of large-area, high-brightness,

8.3 Nonmagnetic Semiconductor Thin Films and Typical Applications

Al electrode GaN:Mg

P-type GaN:Mg prepared by low-energy electron-beam irradiation treatment Al electrode Undoped n-type GaN AIN buffer layer Sapphire substrate

Figure 8.6 et al. [18].

Schematic of the first GaN-based blue LED. Source: Modified from Amao

and full-color displays, the white LEDs and the LDs. These light-emitting devices enable many applications including information display and information storage in our daily life, which are important for people to live more conveniently and happily.

8.3.2

Thin Film Transistors for Displays

TFTs are fundamental components in displays of smartphones, large flat-panel TVs, computers, and flexible electronics. Despite the dominant role of single-crystal Si in the semiconductor technology, the commercial application of TFTs was driven by the use of hydrogenated amorphous silicon (a-Si:H) [20]. In addition to a-Si:H, amorphous oxide semiconductors including indium–gallium–zinc oxide (a-IGZO) have gained much attention due to their potential application in transparent TFTs for electronic paper displays. Certainly polycrystalline silicon and organic semiconductors can also be used for TFT materials. Amorphous semiconductors have several advantages over polycrystalline semiconductors. First, large-area amorphous semiconductor thin films can be easily prepared by conventional deposition techniques, which enables the development of large-area flat-panel displays with high performance and low cost. Second, the thin film properties are uniform down to sub-nanometer scale because of the homogeneous amorphous structure without crystalline defects such as crystal boundaries and dislocations. Third, the amorphous structure can accommodate local strains by atom rearrangements in terms of structural relaxation. This character makes amorphous materials be suitable for flexible electronics. Fourth, amorphous materials, unlike crystalline materials, have no solubility limitation for foreign elements to dissolve in. Therefore, a large amount of foreign elements such as hydrogen can be included in them without altering the amorphous nature. It is known that amorphous Si shows high resistivity and high dangling bond density. These dangling bonds behave as the electron trapping centers to degrade the electrical transport properties, thereby hindering the amorphous Si from the use in TTFs. As shown in Figure 8.7, the hydrogen incorporation in amorphous Si resulted in the reduction of the dangling bond density by several orders of magnitude [4]. The improvement in the electrical properties of the a-Si:H semiconductor led to the development of useful TFTs, which dominate the flat-panel display

263

264

8 Semiconductor Thin Films for Information Technology

Figure 8.7 Schematic of the fourfold-coordinated amorphous silicon random network with dangling bonds indicated by dashed line and a dangling bond passivated by a hydrogen atom. Source: Modified from Street [4].

Si

H

industry, particularly for active matrix LCDs and for active matrix electrophoretic displays. The most common architecture of the a-Si:H-based TFTs is an inverted staggered structure primarily consisting of four fabrication steps [20]. The first step is to prepare a bottom metal gate by sputtering deposition and photolithographic definition. On the top of the metal gate electrode, a gate dielectric layer a-SiNx :H, an intrinsic a-Si:H layer, and n-type a-Si:H thin films are sequentially deposited by a plasma-enhanced CVD technique. Then the metal source and drain contacts are deposited, followed by the etching of the exposed region of the n-type a-Si:H. The a-Si:H-based TFTs are characteristics of uniformity, reproducibility, and reliability, which fulfill the requirements for the commercial application of the displays. Despite the dominant role of a-Si:H in the TFTs used for the displays, amorphous oxide semiconductor thin films such as a-IGZO are becoming competitive TFT materials for driving high-resolution active matrix displays. The post-transition-metal-oxide semiconductors are characterized by ionic bonding, whereas Si is characterized by covalent bonding (Figure 8.8). Direct overlap between the neighboring metal s orbitals is rather large and is not significantly affected even in an amorphous structure (Figure 8.8b). Compared with a-Si:H, the a-IGZO thin films showed the Hall effect mobility exceeding 10 cm2 /(V−1 s−1 ), about one or two orders of magnitude higher than a-Si:H thin films [21]. In particular, the a-IGZO thin films are wide bandgap semiconductors with high optical transparency. This is advantageous for the development of TFTs that are transparent and flexible for future transparent electronics (Figure 8.9). The a-IGZO thin films are n-type conduction. To form the basic building blocks of p–n junctions in electronics, a p-type amorphous oxide semiconductor ZnO⋅Rh2 O3 has been discovered [22]. The successful fabrication of amorphous oxide p–n heterojunction diodes would enable the transparent oxide electronics to grow rapidly (Figure 8.10).

8.3.3

Phase-change Semiconductor Thin Films

Phase-change materials can be reversibly switched between the crystalline and amorphous states. The most thoroughly investigated phase-change materials are

8.3 Nonmagnetic Semiconductor Thin Films and Typical Applications

Covalent semiconductors, for example,Si Crystalline Si

Post-transition-metal oxide semiconductors Crystalline Oxygen 2p-orbital

sp3-orbital

Metal ns-orbital

Amorphous Si

Amorphous

(a)

(b)

Figure 8.8 Schematic orbital drawings for the carrier transport paths (that is, conduction band bottoms) in crystalline and amorphous semiconductors. (a) Covalent semiconductors have carrier transport paths composed of strongly directive sp 3 orbitals, so structural randomness greatly degrades the magnitude of bond overlap, that is, carrier mobility. Note that the orbitals shown are illustrative and do not show exact wave functions. (b) Amorphous oxide semiconductors composed of post-transition-metal cations. Spheres denote metal s orbitals. The contribution of oxygen 2p orbitals is small. Source: Reprinted from Nomura et al. [21]. © 2004 Springer Nature.

ITO

ITO

Y2O3

a-IG ZO

ITO

PET

Film thicknesses a-IGZO active layer : 30 nm Y2O3 gate : 140 nm ITO electrode : 40 nm

(a)

(b)

(c)

Figure 8.9 Flexible TTFTs. (a) Structure of TTFT fabricated on a plastic sheet. (b) A photograph of the flexible TTFT sheet bent at R = 30 mm. The TTFT sheet is fully transparent in the visible light region. (c) A photograph of the flexible TTFT sheet. The transparent TFT devices are made visible by adjusting the angle of the illumination. Source: Reprinted from Nomura et al. [21]. © 2004 Springer Nature [21].

265

266

8 Semiconductor Thin Films for Information Technology

Figure 8.10 (a) Schematic of the p–n heterojunction structure. Source: Modified from Narushima et al. [22].

Au P-type a-ZnO • Rh2O4 N-type a-InGaZnO4 ITO Glass or plastic substrates

Temperature Melting temperature Crystallization temperature

High/short current or laser pulse Crystal to amorphous Lower/longer current or laser pulse Amorphous to crystal

Time

Figure 8.11 Schematic of the phase transformation processes in chalcogenide-based phase-change materials.

chalcogenide-based semiconductor thin films because of their extremely rapid crystallization process. These chalcogenides are fast transformers, switching between the amorphous and crystalline states in responses to an external stimulus of a laser pulse or a current pulse. Figure 8.11 shows schematic of the phase transformation processes. The crystal thin film transfers to an amorphous phase when heated to the temperatures above its melting temperature by a high-power current or laser pulse within a very short time period. The amorphous thin films recrystallize when heated to the temperature above its crystallization temperature by a lower-power current or laser pulse within a longer time period. The different atomic and electronic structures of these two states result in their significantly different optical and electrical properties. Therefore, the chalcogenide-based phase-change materials offer an opportunity to use the pronounced property changes in their amorphous and crystalline states for optical data storage and phase-change memory technologies, respectively [23–25]. This has revolutionized the exiting data storage industries by developing low-cost, high-speed, portable, and nonvolatile devices to store high-density data. Utilizing the optical or electrical property difference in the amorphous and crystalline states leads to the development of two main data storage technologies. One is the mature rewritable optical storage technology that has been commercialized in products of rewritable compact disks (CDs), digital versatile disks (DVDs), and

8.3 Nonmagnetic Semiconductor Thin Films and Typical Applications

Figure 8.12 The fast switching alloys on the pseudobinary line between GeTe and Sb2 Te3 . Sources: Yamada et al. [29], Raoux et al. [30].

1 5/6 2/3

Sb 0 1/6 Sb2Te 1/3

1/2 1/3

Ge1Sb4Te7 Ge2Sb2Te5

1/2 Sb2Te3 2/3

1/6 5/6 0 1 Ge 1 5/6 2/3 1/2 1/3 1/6 0 Te GeTe

Blu-ray disks. The other is the phase-change random-access memory (PCRAM) technology that is thought to be a leading candidate for next-generation digital memory hierarchy [26, 27]. The first optical switching of a Te81 Ge15 Sb2 S2 thin film was demonstrated on the microsecond timescale in 1971 [28]. The crystallization of these early phase-change materials in the range of microsecond to millisecond became the main barrier for them to realize the commercialization. It was found that the alloys on the pseudobinary line between GeTe and Sb2 Te3 showed fast phase transformations (Figure 8.12), reducing the crystallization time down to tens of nanoseconds [29, 30]. Based on these fast transformers, the phase-change optical storage has evolved into the most prevalent technology. Rewritable optical storage media use the significant difference in the optical reflectivity between the amorphous and crystalline states to store information. In addition to the optical properties, the electrical properties change dramatically after the crystallization of the chalcogenide semiconductor phase-change materials. The rapid progress in the phase-change optical storage technology triggered the development of PCRAM technology, which uses the pronounced resistance difference between the amorphous and crystalline states. The commercialized PCRAM products based on Ge2 Sb2 Te5 thin films presented the writing speed of the tens of nanoseconds determined by the stochastic crystal nucleation and growth velocity. To achieve sub-nanosecond high-speed cache-type PCRAM memory, new phase-change materials with faster SET (writing) speed than Ge2 Sb2 Te5 thin films are needed. Rao and his collaborators proposed a new design principle for finding phase-change materials with enhanced thermodynamic driving force to stabilize crystalline precursors [27]. Meanwhile, the geometric conformability between the crystalline precursor and the crystalline phase should be high enough to decrease the energy barrier for crystal nucleation. It was reported that alloying transition metal Ti into Sb2 Te3 caused superior crystallization speed as compared with GeSbTe [31]. Taken into account the above considerations, two essential criteria were used to select the transition metals for further accelerating the crystallization process [27]. First, the crystal-like motifs in the amorphous state should be in a highly geometrically match to the rock-salt structure of Sb2 Te3 .

267

268

8 Semiconductor Thin Films for Information Technology

Second, the introduction of the transition metal should bring in chemical bonds of high strength, which stabilize the crystal-like motifs in the amorphous state. Based on this strategy, a new phase-change material of Sc0.2 Sb2 Te3 was developed, which allowed a writing speed of only 700 ps [27]. Compared with Ge2 Sb2 Te5 thin films, the newly developed Sc0.2 Sb2 Te3 thin films enabled the SET speed of the PCRAM devices based on them to be one order of magnitude faster at all voltages. In addition, the ScSbTe-based device showed a cyclability of ∼105 under the sub-nanosecond switching condition [27]. Chalcogenide-based semiconductor thin films are the most important phase-change materials, which are explored for the rewritable optical storage technology and the emerging nonvolatile PCRAM technology. To further improve the performance and realize the miniaturization of commercial products, the material design and optimization are of great importance for the existing and new applications. The currently used phase-change materials are characterized by high-speed phase transition, high thermal stability of the amorphous phase, pronounced optical and electrical difference between the amorphous and crystalline states, large cycle number of reversible phase transition, and high chemical stability. In addition to these crucial properties, additional functionality of ferromagnetism is useful for achieving ultrahigh-density polymorphic data storage. In addition to the data storage, phase-change semiconductor thin films can be used for an optoelectronic framework that has many applications, such as ultrafast, solid-state displays with nanometer-scale pixels, semitransparent glasses, and artificial retina devices [32].

8.3.4

Semiconductor Thin Films for Sensors

A sensor is a detector that perceives a physical or chemical stimulus and converts it into a readable or processable signal. The detected variables consist of force, heat, light, gas pressure, sound, and so on (Figure 8.13). Accordingly, sensors can be categorized as mechanical, thermal, optical, chemical, and acoustic sensors. Semiconductor thin film sensors are devices that use semiconductor thin films as sensor operation. With the development of IT, the demand for intelligent, miniaturized, and programmable sensors is rapidly increasing. To meet this trend, developing semiconductor thin film sensors integrated with semiconductor-based chips is of significant importance. The integrated semiconductor thin film sensors are usually small in size, highly sensitive, and reliable and at low cost, resulting in high performance-to-cost ratio.

Force Heat Light Input Pressure Sound etc.

Figure 8.13

Semiconductor sensor device Signal detedtion

Signal conversion

Output

Signal conditioning

The main functions of a semiconductor device system.

Electrical signal

8.4 Magnetic Semiconductor Thin Films

Table 8.1

Different types of sensors fabricated with the most widely used semiconductors.

Sensor type

Representative semiconductor used

Representative fabrication technique

Acoustic sensor

ZnO, AlN, PZT

Sputtering [34–36]

Mechanical sensor

Polycrystalline Si

Sputtering, surface micromachining [37]

Magnetic Sensor

Si, III–V compound semiconductors

Sputtering, MBE [38]

Radiation sensor

HgCdTe, CdTe, GaAs, InP, InSb

Liquid-phase epitaxy (LPE), MBE, MOCVD [39]

Thermal sensor

Polysilicon

Sputtering, surface micromachining [40]

Chemical sensor

SnO2 , ZnO, Fe2 O3 , TiO2

PVD, CVD, particulate film assembly by aerosol synthesis and deposition of nanoparticles [41]

Surface micromachining is the main processing technique for mass production of thin film sensors, which permits the fabrication of structurally complex sensors by stacking and patterning layers [33]. Table 8.1 lists different type of sensors with the currently used semiconductors including Si, GaAs, ZnO, SnO2 , and CdS. The selected semiconductors are characterized by several essential properties to meet the requirements for industrial sensor applications. First, the deposited thin films are strongly adhesive to the substrate. Second, the fabrication techniques including the deposition and etching of the thin films are compatible with the very-large-scale integration (VLSI) process. Third, the semiconductor thin films exhibit high sensitivity toward the specific sensor applications. Fourth, the cost is as low as possible. Fifth, the semiconductor thin film sensors have good durability/long service lifespan.

8.4 Magnetic Semiconductor Thin Films IT is the use of computers, networking, and other physical devices to create, process, store, secure, and exchange all forms of electronic data. These operations are mainly realized by two fundamental components of microprocessors and hard disk in computers. Si-based semiconductors are the key materials in microprocessors to enable the control of charges for transmitting and processing data, while ferromagnets provide the spin of electrons to be utilized for storing data in hard disk. Generally, semiconductivity and ferromagnetism do not coexist in a material (Figure 8.14a). However, it was found that europium chalcogenides showed anomalous optical, magnetic, and transport phenomena [42, 43]. The physics behind this anomaly result from the strong d–f exchange interaction between the magnetic-exciton electrons of 5d band and the conduction 4f electrons [43]. As a result, there existed a relationship between the Curie temperature and the conduction electron concentration of

269

270

8 Semiconductor Thin Films for Information Technology Nonmagnetic elements

(a)

(b)

Magnetic elements

(c)

Nonmetallic elements

(d)

Figure 8.14 Four types of semiconductors: (a) a nonmagnetic semiconductor without magnetic ions, (b) an europium chalcogenide magnetic semiconductor with magnetic elements as the main constituents, (c) a diluted magnetic semiconductor with few magnetic elements dissolving in the host nonmagnetic semiconductor, and (d) an amorphous magnetic semiconductor with nonmagnetic elements dissolving in the host ferromagnetic alloy.

these semiconducting europium chalcogenides. Based on these phenomena, Methfessel first proposed the concept of ferromagnetic semiconductor, whose magnetic properties could be modified by carrier injection, electrostatic fields, or other means that changed the free carrier concentration in semiconductors [44]; that is, both the charge and spin of electrons can be manipulated simultaneously in these ferromagnetic semiconductors. The europium chalcogenides are therefore regarded as the first-generation MSs (Figure 8.14b). Despite their scientific importance, these magnetic rare earth compounds usually have complex crystalline structure, compared with the widely used semiconductor materials such as Si and GaAs. It is difficult to obtain high-quality interface structure when integrating them with Si or GaAs. In addition, most of them show Curie temperatures far below room temperature. Therefore, the utilization of these europium chalcogenides for practical devices is a challenging task. Nevertheless, MSs hold promise for uses in next-generation spintronic devices, in which they function as both ferromagnets and semiconductors to realize information processing, communications, and storage. The search for new materials with both ferromagnetism and semiconductivity has spurred the development of the MS field. Basically, there are two approaches that can be used for preparing MS thin films. One is to make nonmagnetic semiconductors ferromagnetic (Figure 8.14c), and the other is to make ferromagnetic metals/alloys semiconducting (Figure 8.14d).

8.4.1

Diluted Magnetic Semiconductors

To maintain the most attractive semiconducting properties used in electronic devices, an approach was proposed to introduce magnetic elements into nonmagnetic semiconductors for creating new type of MSs named as diluted magnetic semiconductors [8, 45, 46]. With ever-growing necessity of miniaturization of future electronic devices, DMSs have stimulated great interest due to their potential for realizing new functionalities and revolutionizing device concepts. III–V compound semiconductors including GaAs and InAs are one type of the most important semiconductors used in electronic devices. Thanks to the development

8.4 Magnetic Semiconductor Thin Films VG > 0

VG < 0

VG = 0

Metal gate insulator (In,Mn)As InAs (AI,Ga)Sb AlSb GaAs substrate 50

50 25 0

25

–25

22.5 K

RHall (Ω)

(a)

RHall(Ω)

–50 –8

–4

0

4

B (mT)

8

0 0V +125 V –125 V 0V

–25

–50 –1.0 (b)

–0.5

0

0.5

1.0

B(mT)

Figure 8.15 (a) Field-effect control of the hole-induced ferromagnetism in magnetic semiconductor (In,Mn)As field-effect transistors. Shown are the cross sections of a metal–insulator–semiconductor structure under gate biases V G . This controls the hole concentration in the magnetic semiconductor channel (filled circles). Negative V G increases hole concentration, resulting in enhancement of the ferromagnetic interaction among magnetic Mn ions, whereas positive V G has an opposite effect. The arrow schematically shows the magnitude of the Mn magnetization. The InAs/(Al,Ga)Sb/AlSb structure under the (In,Mn)As layer serves as a buffer relaxing the lattice mismatch between the structure and the GaAs substrate to produce a smooth surface on which the magnetic layer is grown. (b) RHall versus field curves under three different gate biases. Application of V G = 0, +125 and −125 V results in qualitatively different field dependence of RHall measured at 22.5 K. When holes are partially depleted from the channel (V G = +125 V), a paramagnetic response is observed (blue dash-dotted line), whereas a clear hysteresis at low fields ( 0

VG < 0 EDL

M(emu/cm3)

40 EDL

20 0 +2 V, 30 min As–prepared

–20

–2 V, 30 min

–40

–2 V, 60 min

–60 –30 5 nm

(a)

(b)

–20

–10

10 0 H(kOe)

20

30

(c)

Figure 8.22 (a) HRTEM image of the newly developed CFTBO AMS. (b) Schematic diagrams for the electric-field control of carrier concentrations at different V G . (c) Ms increasing with both positive and negative V G . Source: (a) Reprinted from Chen et al. [61], (b, c) Reprinted from Chen et al. [61]. © 2019 Chinese Institute of Electronics.

sample could be an intrinsic MS. Based on this intrinsic MS, we may use V G to control its electrical conduction type of the different parts within the material, which would form a p–n junction at the interface between the different parts. To our best of our knowledge, this should be the first demonstration of an intrinsic MS. Meanwhile, the conduction type of this intrinsic MS can be tuned to be n-type or p-type by using an external electric field.

8.4.5

Prospective for Magnetic Semiconductors

The reverse thinking enabled the development of a new family of MSs by oxidizing originally ferromagnetic AAs. These amorphous MSs showed much higher Curie temperatures than room temperature. Based on them, prototype magnetic semiconductor-based spintronic devices such as spin-light emitting diodes and spin-field effect transistors could be fabricated. One of the most concern issues is the carrier mobility of these amorphous MSs. Generally, the structural disorder would enhance the scattering strength of the charger carriers and hinder them from moving fast in amorphous materials. However, it is possible to discover unprecedented interface effects that could be used to overcome the existing difficulties in the applications of this new family of MSs. For instance, the electron gas may emerge at the interfaces between two MSs. To test the universality of the design concept, various ferromagnetic amorphous alloy systems can be used as hosts for the development of new amorphous MSs. Among them, both p-type and n-type MSs may be prepared and integrated together to form the basic component p–n heterojunctions. These magnetic p–n heterojunctions would provide a platform for exploring the new physics at the interfaces of two MSs and furthering our understanding of the interplay between the electricity and ferromagnetism. In addition, the crystallization induced semiconductor-metal transition is also interesting and worth to be investigated systematically. In these ferromagnetic phase change materials, the coexistence of phase change induced drastic difference in both the resistance and ferromagnetism may open a new avenue for designing high density multi-state memory.

279

280

8 Semiconductor Thin Films for Information Technology

8.5 Conclusion and Outlook This chapter outlines the currently used techniques for fabrication of semiconductor thin films, an overview of the representative semiconductor thin films developed toward the specific applications, and the materials being developed for next-generation electronic/spintronic devices. With the matured manufacturing process of integrated circuits, the traditional semiconductor thin films including Si, Ge, GaAs, and wide bandgap ones have been playing important roles as the key materials in micro/nanoelectronics. Their unique properties and combined functionalities render them to work in a variety of digital applicants including computers and cell phones, which are now essential items of our modern information society. In addition to these conventional semiconductor thin films, amorphous materials such as phase-change chalcogenides, transparent oxides, and MSs have received great attention due to their unique properties. They are being developed or will partly replace the currently used semiconductors for realizing new devices with highly desirable functionalities. The continuous development of IT poses new challenges for further extending Moore’s law that requests miniaturized, high-integration, low power consumption, and rapid response ICs. To meet this demand, the two-dimensional (2D) crystals including graphene, black phosphorus, MoS2 , MoSe2 , and WSe2 exhibit high performance when integrated in nanoelectronic devices. Moreover, new physical phenomena emerge at their surfaces or the interfaces between these 2D thin films. Utilizing the intriguing surface or interface effects may lead to the breakthrough discoveries and innovations that help these materials to work as “post-silicon” materials for further developing IT.

List of Abbreviations ICs IT MSs PCMs TFTs

integrated circuits information technology magnetic semiconductors phase-change materials thin film transistors

References 1 Riordan, M. and Hoddeson, L. (1998). Crystal Fire: The Invention of the Transistor and the Birth of the Information Age. New York: Norton and Company. 2 Orton, J. (2009). Semiconductors and the Information Revolution: Magic Crystals that Made IT Happen. The Netherlands: Academic Press/Elsevier.

References

3 Benda, V., Gowar, J., and Grant, D.A. (1999). Power Semiconductor Devices: Theory and Applications. Chichester: Wiley. 4 Street, R.A. (1991). Hydrogenated Amorphous Silicon. Cambridge: Cambridge University Press. 5 Nomura, K., Ohta, H., Takagi, A. et al. (2004). Room-temperature fabrication of transparent flexible thin-film transistors using amorphous oxide semiconductors. Nature 432: 488–492. 6 Wuttig, M. (2005). Phase-change materials – towards a universal memory? Nat. Mater. 4: 265–266. 7 Xiong, F., Liao, A.D., Estrada, D., and Pop, E. (2011). Low-power switching of phase-change materials with carbon nanotube electrodes. Science 332: 568–570. 8 Ohno, H. (1998). Making nonmagnetic semiconductors ferromagnetic. Science 281: 951–956. 9 Matsumoto, Y., Murakami, M., Shono, T. et al. (2001). Room-temperature ferromagnetism in transparent transition metal-doped titanium dioxide. Science 291: 854–856. 10 Liu, W.J., Zhang, H.X., Shi, J.A. et al. (2016). A room-temperature magnetic semiconductor from a ferromagnetic metallic glass. Nat. Commun. 7: 13497. 11 Barlow, F.D., Elshabini-Riad, A., and Brown, R. (1998). Film deposition techniques and processes. In: Thin Film Technology Handbook (eds. A. Elshabini-Riad and F.D. Barlow), 18–21. New York: McGraw-Hill. 12 Chang, L.L., Esaki, L., Howard, W.E. et al. (1973). Structures grown by molecular-beam epitaxy. J. Vac. Sci. Technol. 10: 655–662. 13 Cho, A.Y. (1975). Impurity profiles of GaAs epitaxial layers doped with Sn, Si, and Ge grown with molecular beam epitaxy. J. Appl. Phys. 46: 1733–1735. 14 George, J. (1992). Epitaxial film deposition techniques, Chapter 7. In: Preparation of Thin Films (ed. J. George), 303–331. New York: Marcel Dekker, Inc. 15 Munekata, H., Ohno, H., von Molnar, S. et al. (1989). Diluted magnetic III–V semiconductors. Phys. Rev. Lett. 63: 1849–1852. 16 Ohno, H., Shen, A., Matsukura, F. et al. (1996). (Ga,Mn)As: a new diluted magnetic semiconductor based on GaAs. Appl. Phys. Lett. 69: 363–365. 17 Manasevit, H.M. (1968). Single-crystal gallium arsenide on insulating substrates. Appl. Phys. Lett. 12: 156–159. 18 Amao, H., Kito, M., Hiramatsu, K., and Akasaki, I. (1989). P-type conduction in Mg-doped GaN treated with low-energy electron beam irradiation (LEEBI). Jpn. J. Appl. Phys. 28: L2112–L2114. 19 Nakamura, S., Mukai, T., Senoh, M., and Iwasa, N. (1992). Thermal annealing effects on p-type Mg-doped GaN films. Jpan. J. Appl. Phys. 31: L139–L142. 20 Brotherton, S.D. (2013). Introduction to Thin Film Transistors – Physics and Technology of TFTs. Cham: Springer International Publishing. 21 Nomura, K., Ohta, H., Takagi, A. et al. (2004). Room-temperature fabrication of transparent flexible thin-film transistors using amorphous oxide semiconductors. Nature 432: 488–492.

281

282

8 Semiconductor Thin Films for Information Technology

22 Narushima, S., Mizoguchi, H., Shimizu, K.I. et al. (2003). A p-type amorphous oxide semiconductor and room-temperature fabrication of amorphous oxide p–n heterojunction diodes. Adv. Mater. 15: 1409–1413. 23 Ohno, E., Yamada, N., Kurumizawa, T. et al. (1989). TeGeSnAu alloys for phase-change type optical disk memories. Jpn. J. Appl. Phys. 28: 1235–1240. 24 Lankhorst, M.H.R., Ketelaars, B.W.S.M.M., and Wolters, R.A.M. (2005). Low-cost and nanoscale non-volatile memory concept for future silicon chips. Nat. Mater. 4: 347–352. 25 Ovshinsky, S.R. (1968). Reversible electrical switching phenomena in disordered structures. Phys. Rev. Lett. 21: 1450–1453. 26 Nam, S.W., Chung, H.S., Lo, Y.C. et al. (2012). Electrical wind force-driven and dislocation-templated amorphization in phase-change nanowires. Science 336: 1561–1566. 27 Rao, F., Ding, K.Y., Zhou, Y.X. et al. (2017). Reducing the stochasticity of crystal nucleation to enable subnanosecond memory writing. Science 358: 1423–1427. 28 Feinleib, J., deNeufville, J., Moss, S.C., and Ovshinsky, S.R. (1971). Rapid reversible light-induced crystallization of amorphous semiconductor. Appl. Phys. Lett. 18: 254–256. 29 Yamada, N., Ohno, E., Akahira, N. et al. (1987). High speed overwritable phase change optical disk material. Jpn. J. Appl. Phys. 26 (Suppl. 26-4): 61–66. 30 Raoux, S., Welnic, W., and Lelmini, D. (2010). Phase change materials and their application to nonvolatile memories. Chem. Rev. 110: 240–267. 31 Zhu, M., Xia, M.J., Rao, F. et al. (2014). One order of magnitude faster phase change at reduced power in Ti-Sb-Te. Nat. Commun. 5: 4086. 32 Hosseini, P., Wright, C.D., and Bhaskaran, H. (2014). An optoelectronic framework enabled by low-dimensional phase-change films. Nature 511: 206–211. 33 Mastrangelo, C.H. and Tang, W.C. (1994). Semiconductor sensor technologies. In: Semiconductor sensors (ed. S.M. Sze), 55–74. New York: Wiley. 34 Yamamoto, T., Shiosaki, T., and Kawabata, A. (1980). Characterization of ZnO pieroelectric films prepared by rf planar-magnetron sputtering. J. App. Phys. 51: 3113–3120. 35 Shiosaki, T., Yamamoto, T., Oda, T., and Kawabata, A. (1980). Low temperature growth of piezoelectric AlN thin film by RF reactive planar magnetron sputtering. Appl. Phys. Lett. 36: 643–645. 36 Sreenivas, M., Sayer, M., Baar, D.J., and Nishioka, M. (1988). Surface acoustic wave propagation on PZT thin film. Appl. Phys. Lett. 52: 709–711. 37 Kloeck, B. and Rooij, N.F. (1994). Mechanical sensors. In: Semiconductor Sensors (ed. S.M. Sze), 153–204. New York: Wiley. 38 Popovic, R.S. (1991). Hall Effect Devices. Bristol: Adam Hilger. 39 Audet, S. and Steigerwald, J. (1994). Mechanical sensors. In: Semiconductor Sensors (ed. S.M. Sze), 271–329. New York: Wiley. 40 Van Herwaarden, A.W. and Meijer, G.C.M. (1994). Thermal sensors. In: Semiconductor Sensors (ed. S.M. Sze), 331–379. New York: Wiley. 41 Morrison, S.R. (1994). Chemical sensors. In: Semiconductor sensors (ed. S.M. Sze), 383–409. New York: Wiley.

References

42 Holtzberg, F., McGuire, T.R., Methfessel, S., and Suits, J.C. (1964). Effect of electron concentration on magnetic exchange interactions in rare earth chalcogenides. Phys. Rev. Lett. 13: 18–21. 43 Kasuya, T. and Yanase, A. (1968). Anomalous transport phenomena in Eu-chalcogenide alloys. Rev. Mod. Phys. 40: 684–696. 44 Methfessel, S. (1965). Potential applications of magnetic rare earth compounds. IEEE Trans Mag. 1: 144–155. 45 Furdyna, J.K. (1988). Diluted magnetic semiconductors. J. Appl. Phys. 64: R29–R64. 46 Dietl, T. and Ohno, H. (2014). Dilute ferromagnetic semiconductors: physics and spintronic structures. Rev. Mod. Phys. 86: 187–251. 47 Ohno, H., Chiba, D., Matsukura, F. et al. (2000). Electric-field control of ferromagnetism. Nature 408: 944–994. 48 Chen, L., Yang, X., Yang, F.H. et al. (2011). Enhancing the Curie temperature of ferromagnetic semiconductor (Ga,Mn)As to 200 K via nanostructure engineering. Nano Lett. 11: 2584–2589. 49 Pulizzi, F. and Chambers, S. (2010). Is it really intrinsic ferromagnetism? Nat. Mater. 9: 956–957. 50 Pulizzi, F. and Samarth, N. (2010). A model ferromagnetic semiconductor. Nat. Mater. 9: 955–956. 51 Dietl, T. (2010). A ten-year perspective on dilute magnetic semiconductors and oxides. Nat. Mater. 9: 965–974. 52 Zhao, J.H. (2016). Is it possible to create magnetic semiconductors that work at room temperature? Chin. Sci. Bull. 61: 1401–1406. 53 Chen, N., Zhang, Y.Q., and Yao, K.F. (2017). Transparent magnetic semiconductors from ferromagnetic amorphous alloys. Acta Phys. Sin. 66: 176113. 54 Chen, N., Li, H.P., Hirata, A. et al. (2017). Transparent magnetic semiconductor with embedded metallic glass nano-granules. Mater. Design. 132: 208–214. 55 Inoue, A. (2000). Stabilization of metallic supercooled liquid and bulk amorphous alloys. Acta Mater. 48: 279–306. 56 Johnson, W.L. (1999). Bulk glass-forming metallic alloys: science and technology. MRS Bull. 24: 42–56. 57 Greer, A.L. (1995). Metallic glasses. Science 267: 1947–1953. 58 Coey, J.M.D., Venkatesan, M., and Fitzgerald, C.B. (2005). Donor impurity band exchange in dilute ferromagnetic oxides. Nat. Mater. 4: 173–179. 59 Inoue, A., Shen, B.L., Koshiba, H. et al. (2004). Ultra-high strength above 5000 MPa and soft magnetic properties of Co–Fe–Ta–B bulk glassy alloys. Acta Mater. 52: 1631–1637. 60 Inoue, A., Shen, B.L., Koshiba, H. et al. (2003). Cobalt-based bulk glassy alloy with ultrahigh strength and soft magnetic properties. Nat. Mater. 2: 661–663. 61 Chen, N., Fang, K.X., Zhang, H.X. et al. (2019). Amorphous magnetic semiconductors with Curie temperatures above room temperature. J. Semicond. 40 (8): 081510. https://doi.org/10.1088/1674-4926/40/8/081510.

283

284

8 Semiconductor Thin Films for Information Technology

62 Lucas, M.S., Bourne, W.C., Sheets, A.O. et al. (2011). Nanocrystalline Hf and ta containing FeCo based alloys for high frequency applications. Mater. Sci. Eng. B 176: 1079–1084. 63 Fang, K.X., Zhang, Y.Q., Zhang, Z.J., and Chen, N. (2019). Crystallization-induced semiconductor-metal transition in an amorphous CoFeTaBO magnetic semiconductor nanocomposite. J. Alloys Compn. 797: 606–611. 64 Bader, S.D. and Parkin, S.S.P. (2010). Spintronics. Annu. Rev. Condensed Matter. Phys. 1: 71–88. 65 Song, C., Cui, B., Li, F. et al. (2017). Recent progress in voltage control of magnetism: materials, mechanisms, and performance. Prog. Mater. Sci. 87: 33–82. ´ I., Fabian, J., and Das Sarma, S. (2004). Spintronics: fundamentals and 66 Žutic, applications. Rev. Mod. Phys. 76: 323–410. 67 Awschalom, D.D. and Flatté, M.E. (2007). Challenges for semiconductor spintronics. Nat. Phys. 3: 153–159. 68 Weisheit, M., Fähler, S., Marty, A. et al. (2007). Electric field-induced modification of magnetism in thin-film ferromagnets. Science 315: 349–351. 69 Cui, B., Song, C., Mao, H.J. et al. (2016). Manipulation of electric field effect by orbital switch. Adv. Funct. Mater. 26: 753–759. 70 Zhang, Y.Q., Zhao, S.F., Song, C. et al. (2018). Electric-field control of ferromagnetism in a Co–Fe–Ta–B amorphous alloy, mater. Design 143: 65–71. 71 Nakano, M., Shibuya, K., Okuyama, D. et al. (2012). Collective bulk carrier delocalization driven by electrostatic surface charge accumulation. Nature 487: 459–462.

285

9 Glass Transition in Organic Semiconductor Thin Films Han-Nan Yang 1 and Zheng-Hong Lu 1,2 1 Yunnan University, Center for Optoelectronics Engineering Research, Department of Physics, No. 2, Cuihu North Road, Kunming, Yunan Province 650031, China 2 University of Toronto, Department of Materials Science and Engineering, 27 King’s College Circle, Toronto, ON M5S 1A1, Canada

9.1 Introduction Organic semiconductor thin-films are key building blocks for fabricating future generation of organic optoelectronic devices such as organic light-emitting diodes (OLED). Already OLED technology has been industrialized in making vivid flat-panel displays for a variety of electronic gadgets such as smart phones and televisions. The structure of OLED comprises a stacked of multiple organic molecular thin-films deposited either by physical vapor deposition (PVD) or by solution-casting. Each of these layers plays a different role and collectively they dictate the performance of organic optoelectronic devices [1–5]. In stark contrast to inorganic semiconductors, each organic molecular is topologically flawless and therefore amorphous packed organic films do not need lattice matching with substrates. This unique amorphous property enables fabrication of OLED on any type of substrate including flexible plastics [6, 7]. Amorphous materials, also frequently referred to as glasses, are a special class of materials characterized by their liquid-like disorder structure. The most common route to form a glass is by quenching a molten liquid directly into a non-crystalline solid without crystallization. In this case, before forming a glass, the super-cooled state is formed, that is, a system in metastable equilibrium [8–10]. In general, a glass transition temperature (T g ) in a non-crystalline solid is defined by a distinct inflection point in the specific volume change as the temperature is varied. Because of the complexity in defining molecular packing in an amorphous network, the glass transition physics remains largely elusive. It is known that glass transition is not a typical phase transition, as it does not involve abrupt changes in any observable physical properties. Up to now, numerous theories have been proposed to describe the glass transition processes and the characteristic glass transition temperature T g , which is frequently defined by changes in viscosity, or in free volume or in other observable material properties [11]. Because the glass transition Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications, First Edition. Edited by Yujun Song. © 2021 WILEY-VCH GmbH. Published 2021 by WILEY-VCH GmbH.

9 Glass Transition in Organic Semiconductor Thin Films

is somewhat a convoluted process involving dynamics and thermodynamics, some second-order physical properties of amorphous materials at around T g will change discontinuously. These physical properties include viscosity, coefficient of thermal expansion (CTE), heat capacity, optical refractive index, etc. Thus, the T g of an amorphous material is generally determined by a temperature at which some of these physical properties show a significant deviation from below to above this temperature. One of the key challenges in OLED technology is thermal stability. For example, for applications requiring very high brightness in harsh working environments such as outdoor displays and automobile displays, where the working temperature could be higher than 400 K in summer, thermal stability of organic materials become a major issue [12]. It is generally known that T g is a critical factor dictating a device’s thermal stability, device degradations, molecular orientation in thin-films, and thin-film interlayer diffusions [13–16]. In particular, the material having the lowest T g in an OLED device determines the thermal-failure temperature of the device [17, 18]. Figure 9.1a shows abrupt operation failures, defined by a sudden drop in luminance, of OLEDs as operating temperatures are increased to the catastrophic failure temperature (T cf ). The hole transport material in these devices have the lowest T g of all organic thin-films used. Figure 9.1b shows a strong linear correlation between T g and T cf . Engineering organic thin-films for higher glass transition temperature is thus important for developing future generation of robust organic devices. Now let’s discuss briefly methodology in measuring glass transition temperature. For bulk materials, T g is typically measured by differential scanning calorimetry (DSC) method. For thin-film materials, ellipsometer has been used to measure T g by monitoring changes in film thicknesses, i.e. volume change. As mentioned above, there are two types of organic thin-film semiconductors: polymeric thin films that are normally deposited by spin coating and small molecular thin films that Tg (K) 390

Tcf

1.0

380

0.8

400

410

TPD: C60 (0 wt%) TPD: C60 (10 wt%) TPD: C60 (30 wt%) TPD: C60 (50 wt%)

410

400 Tcf (K)

370

0.6 0.4

360 390

0.2 0.0

350

NPB: C60 (0 wt%) NPB: C60 (10 wt%) NPB: C60 (30 wt%)

Pure TPD TPD:C60 in 50 wt%

300

320

340

360

380

400

340 340

Temperature

(a)

390

380

Tcf (K)

Normalized luminence (a.u.)

286

(b)

350

360

370

380

380 390

Tg (K)

Figure 9.1 (a) Electroluminance versus temperature of OLEDs having pure TPD and TPD:C60 as hole transport materials, respectively. (b) T cf versus T g plots of OLEDs materials made with various organic:C60 composite hole transport materials and glass transition temperatures. Source: Yang et al. [18]. © 2018 AIP Publishing.

9.2 Determination of Glass Transition Temperature in Organic Thin Films

are deposited by PVD. First, the T g of polymeric thin-films has been successfully determined by measuring an inflection point at their thickness versus temperature plot, i.e. a disruptive change of CTE. Using this method, the relationship between variables such as molecular weight and glass transition temperatures of polymeric thin-films have been studied [19–21]. The work on ellipsometry study of glass transition in PVD small molecular thin-film organic semiconductors and their composite thin films has recently become available. Organic composite semiconductor thin-films is a material system in which two organic constituents are mixed to achieve a specific functionality, and composite thin-film is a common practice in making organic devices [22, 23]. For example, two component host-dopant systems with doping concentration ∼10–20 wt% are widely used in fabricating OLEDs and solar cells. In this chapter, the T g of pure organic thin films and organic composite thin films will be discussed. The composite organic thin films will include: (i) organic–organic composite thin films [24]; (ii) fullerene nano-organic composite thin films [18].

9.2 Determination of Glass Transition Temperature in Organic Thin Films Several archetypical organic molecules used in OLEDs and solar cells are selected for variable temperature spectroscopic ellipsometry studies. These organics are N,N ′ bis(1-naphthalenyl)-N,N ′ -bis-phenyl-(1,1′ -biphenyl)-4,4′ -diamine (NPB), N,N ′ -bis (3-methylphenyl)-N,N ′ -bis(phenyl)benzidine (TPD), 1,3,5-tris(1-phenyl-1Hbenzimidazol-2-yl)benzene (TPBi), 4,4′ ,4′′ -tris(N-3-methylphenyl-N-phenylamino)triphenylamine (m-MTDATA), 1,3,5-tri(m-pyrid-3-yl-phenyl)benzene (TmPyPB), 4,4′ , 4′′ -tris(carbazol-9-yl)-triphenylamine (TCTA), and C60 . The chemical structures of these molecules are shown in Figure 9.2. Thin films of these molecules are formed by condensation of these molecular vapors on silicon substrates under ultra-high vacuum, i.e. PVD. Each type of molecular vapor is generated by thermal sublimation of the molecule from solid powder placed in a vacuum furnace or referred to as Knudsen cell. For a composite thin film, two types of vapors are simultaneously generated and condensed on silicon substrates. The weight percentage in the films is controlled by controlling the Knudsen cell temperature. To ensure accuracy and reproducibility, vapor flux is monitored in real-time by quartz crystal microbalance inside the PVD chamber. The thickness of organic thin-film is kept at 100 nm for all samples. The thin film deposition rate was carried out at 0.1 nm/s under ∼10−7 Torr vacuum. The spectroscopic ellipsometry measurements were performed at a 75∘ incident angle using a polarized light at wavelengths varying from 400 to 800 nm. The ellipsometry data acquisition time for each temperature point is 15 minutes. The substrate temperature was calibrated by a thermocouple placed directly on the silicon. The temperature ramping rate is 2 K/min. The temperature is scanned from room temperature to a temperature slightly above the T g . The schematic diagram of the variable temperature ellipsometry measurement set-up is shown in Figure 9.3. The thin-film

287

288

9 Glass Transition in Organic Semiconductor Thin Films N

N

N

N

N N N

TPD

NPB

N

N

N

N

N

N

TmPyPB

N

N

N N

N N N

TPBi

Figure 9.2

N

m-MTDATA

TCTA

C60

The chemical structures of various organic molecules.

Light source

Detector Polarization generator

Polarization analyzer 75°

Thin-flim

Thermal tape Substrate Thermal stage

Figure 9.3

Schematic set-up for variable temperature ellipsometry measurement.

thickness at a given temperature is obtained by modeling analysis of the ellipsometry parameters (Ψ and Δ). Figure 9.4a shows a percentage change in thickness of pure NPB thin film as a function of temperature. Here, a disruptive change in thickness is observed near its T g obtained by DSC measurement of NPB in powder form. This disruptive change in thickness indicates that the glass transition of the NPB film behaves in a manner similar to that of a crystal melting. This abrupt change in NPB film thickness is significantly different from that of traditional polymeric thin-films. As a comparison, the green dashed line in Figure 9.4a shows thickness variation of a polymeric polymethyl methacrylate (PMMA) thin-film at around its glass transition temperature using data reported in Ref. [21]. For all small molecular thin-films formed by PVD method, an abrupt change in thickness occurs within a narrow temperature range (∼3 K, in the case of NPB thin film). To determine the glass transition temperature, three linear regression fittings of three distinct thickness-temperature plot regions can be used. As shown in

9.2 Determination of Glass Transition Temperature in Organic Thin Films

0.08

0.06

(d–d0)/d0

(d–d0)/d0

0.06 0.04

0.02

0.04 0.02 0.00

0.00 310

330

350

370

390

410

300

430

Temperature (K)

330

360

390

420

450

480

Temperature (K)

(b)

(a) 0.06

Before

After

(d–d0)/d0

0.05 0.04 0.03 0.02 0.01 0.00

500 μm 230 300 320 340 360 380 400 420 440

(c)

Temperature (K)

(d)

Figure 9.4 (a) NPB thickness versus temperature plot. The solid red solid lines are the linear regression of the experimental data in three distinct temperature regions. The dashed line is for polymeric PMMA thin film replotted using data from Ref. [21]; (b) NPB:TPD composite thickness versus temperature plot. The solid red lines are the linear regression of the experimental data in three distinct temperature regions. The dashed line is for polymer blend PS:TMPC film replotted using data taken from Ref. 19; (c) NPB:C60 composite thickness versus temperature plot. The solid red lines are the linear regression of the experimental data in three distinct temperature regions; (d) NPB film surface images taken before and after the ellipsometry measurement shown in (a). Source: (a, b) Yang et al. [24]. © 2018 Elsevier, (c, d) Modified from Yang et al. [24]. © 2018 Elsevier.

Figure 9.4a, three linear lines provide a fairly good fit to the experimental data. The highest temperature used for fitting the data is a temperature above which there is a significant decrease in thickness (Figure 9.4a). This sudden decrease in thickness is caused by the delamination in organic thin film, as shown in Figure 9.4d. The intersections of these three lines are used to determine the onset and end of glass transition process in PVD thin films. The mid temperature between onset and end transition temperature is then used to define the glass transition temperature of PVD thin-films. The T g of the NPB thin films is determined to be 379.6 K, which is slightly higher than 373.2 K of NPB in solid powder measured by DSC. The T g of PVD-deposited NPB thin film is expected to be slightly higher than that of NPB in the powder form because molecular packing in a thin film is not completely random. This type of material is referred to as ultra-stable glass that typically requires a normal glass to age over thousands of years [8, 25]. The origin of this ultra-stable glass forming, i.e. higher than bulk transition temperature, is related to the formation process

289

290

9 Glass Transition in Organic Semiconductor Thin Films

Table 9.1 T g of various organic molecules in thin-film forms and in powder forms [27–31].

Organic molecule

T g (K) (in thin-film form)

T g (K) (in powder form)

TPD

342.4 ± 1.1

338.2 [27]

m-MTDATA

356.5 ± 1.2

348.2 [28]

TmPyPB

354.7 ± 1.2

352.2 [29]

NPB

379.5 ± 1.2

373.2 [27]

TPBi

397.9 ± 1.2

395.2 [30]

TCTA

428.4 ± 1.3

424.2 [31]

Source: Yang et al. [24]. © 2018 Elsevier.

of these organic thin-films. During PVD deposition, lateral diffusion of molecules on a film surface combined with molecular conformational packing may lead to the formation of somewhat ordered packing microstructure within the film. This results in enhanced kinetic stability and lower enthalpy [26]. This ordered molecular packing may also explain that thin-film organic semiconductors exhibit a disruptive change in thickness or volume at T g , similar to the melting process of a crystal solid. Table 9.1 compares the T g of various pure PVD organic thin films measured by ellipsometer (T g PVD ) to the T g of these organic molecules in powder forms reported literatures (T g powder ) [27–30, 32]. It is found that the T g of these organic thin-films are all slightly higher than their corresponding materials in powder forms. The relationship between T g powder and T g PVD is plotted in Figure 9.5. A linear regression fit to the experimental data yields the following equation: powder

TgPVD = Tg

+ 4.7

(9.1)

This equation indicates that the glass transition temperature of a one-component PVD thin-film can be simply calculated by the glass transition temperature of its starting source material in powder form. Now let’s examine the glass transition in organic–organic composite PVD thin films, Figure 9.4b shows a variation of NPB:TPD composite thickness as a function of temperature. The solid red lines are the linear regression of the experimental data in three distinct temperature regions. The dashed line is for polymer blend PS:TMPC film re-plotted using data taken from Ref. [19]. Again, the glass transition process in small molecular organic–organic composite thin film shows a disruptive change in thickness, in sharp contrast to the polymeric composite films. For organic nano-composite thin-film system. The fullerene C60 is used as a nano filler for organic matrix. Figure 9.4c plots the NPB:C60 composite thickness versus temperature plot [18]. Interestingly, the nano-composite thin film also shows a disruptive change in thickness at around glass transition temperature. The solid red

9.3 Model for Predicting Glass Transition Temperature of Organic–Organic Composites

440

420

TPD TmPyPb m-MTDATA NPB TPBi TCTA

Tg PVD

400

380

360

340

340

360

380

400

420

440

Tg powder

Figure 9.5 T g powder versus T g PVD plot based on data from Table 9.1. The red dashed line is a linear regression fit to the experimental data.

lines are the linear regression of the experimental data in three distinct temperature regions. In summary, all three types of PVD organic thin-films show a unique disruptive change in film thicknesses as a function of temperature. This disruptive thickness change can then be used to define the glass transition temperatures of the organic semiconductor thin films.

9.3 Model for Predicting Glass Transition Temperature of Organic–Organic Composites Two component host-dopant systems with doping concentration ∼10–20 wt% are widely used in fabricating OLEDs and solar cells. To develop a model for predicting the T g as a function of constituent concentration, we selected three different groups of Host:Dopant combinations: NPB:TPD (x wt%), NPB:TPBi (x wt%), and NPB:TCTA (x wt%). The doping concentration x wt% in each group is varied with x = 0, 5, 25, 50, 75, 90, 95, and 100. Using the PVD T g definition, we have successfully determined the T g of each sample. As shown in Figure 9.6, these data exhibit a similar linear relationship from 0 to 100 wt% doping concentration. This indicates

291

9 Glass Transition in Organic Semiconductor Thin Films NPB:TPBi

NPB:TPD

Tg (K)

292

380

400

370

395

360

390

NPB:TCTA 430 420 410 400

385

350

390

380

340 0

20

40

60

80

100

380 0

Concentration (wt%)

(a)

20

40

60

80

100

0

Concentration (wt%)

(b)

20

40

60

80

100

Concentration (wt%)

(c)

Figure 9.6 The measured T g (shown in open symbols) of composite mixture films are plotted against dopant concentration. The theoretical data computed using the Fox equation are shown as dashed lines. Source: Yang et al. [24]. © 2018 Elsevier.

that there is no or little interaction between two constituent organic molecules in the composite thin-films. To quantify the result, we try to fit the experimental data using the Gordon–Taylor equation developed for polymers [31]. Tgm =

w1 ∗ Tg1 + k ∗ w2 ∗ Tg2 w1 + k ∗ w2

(9.2)

𝜌 ∗𝛥𝛼

where k = 𝜌1 ∗𝛥𝛼2 , T gm is the T g of the mixture material. w1 , w2 , T g1 , T g2 , 𝜌1 , 𝜌2 are 2 1 the weight fraction of doping concentration, glass-transition temperature and density of component 1 and 2, respectively. As these organic components have similar densities, Eq. (9.2) will be simplified as: w w 1 = 1 + 2 (9.3) Tgm Tg1 Tg2 This equation is also known as the Fox equation. It is widely used to predict the T g of polymeric mixtures. The small molecular organic molecules used in organic semiconductors are relatively rigid and have similar molecular weight [33, 34]. We now apply Eq. (9.3) to experimental T g of composite organic films. The dashed lines shown in Figure 9.6 are the computed lines using Eq. (9.3). As shown in Figure 9.6, each theoretical line shows a good agreement with the experimental data. In conclusion, the Fox equation is proven valid in predicting T g of organic–organic composite semiconductor thin films.

9.4 Model for Predicting Glass Transition Temperature of Nano-organic Composites Another type of composite thin-film system is the interacting composite thin-films. As a model system, we chose the hole transporting molecule doped with nano-carbon C60 to establish a rule for predicting T g . In this system, there is an interaction between the hole transporting molecules and C60 [35]. As we have discussed in the previous section, for organic–organic composite systems, the FOX equation is valid to predict the T gm of the non-interaction composite thin

9.4 Model for Predicting Glass Transition Temperature of Nano-organic Composites

films. However, the T g of C60 composite (T gc ) thin films with various C60 doping concentration clearly deviate from the Fox equation. This indicates that the C60 composite thin film is different from the classical two-component organic thin-film system. The Fox equation can effectively predict the T g of the organic–organic small molecular thin-films is because there is little molecule–molecule interaction in that system. In contrast, there is an interaction between the hole transporting molecules and C60 . Thus, we have to develop a new equation to quantify the T g of these C60 composite thin-films. Figure 9.7 shows that T g increases linearly with increasing C60 doping concentration. Thus the linear regression analysis is used and the results are shown as the (a)

40

TPD:C60

ΔTg (K)

30 20 10 0 (b) 50

NPB:C60

ΔTg (K)

40 30 20 10 0 (c) 60

ΔTg (K)

50

m-MTDATA:C60

40 30 20 10 0 0

10

20

30

40

50

Concentration (wt%)

Figure 9.7 (a) TPD:C60 , (b) NPB:C60 , and (c) m-MTDATA:C60 thin films as a function of weight concentration. The dashed lines are the linear regression fits to the experimental data. Error bars represent measurement error. Source: Yang et al. [18]. © 2018 AIP Publishing.

293

294

9 Glass Transition in Organic Semiconductor Thin Films

Table 9.2 The 𝜅, 𝛼, T g , and M of organic–C60 composite thin-films.

C60 composite thin films

M (host molecule) (g/mol)

T g (host molecule) (K)

𝜿

𝜶 (10−3 )

TPD:C60

516.7

342.4

0.77

1.49

NPB:C60

588.7

379.5

0.92

1.56

m-MTDATA:C60

789.0

356.5

1.19

1.51

Source: Yang et al. [18]. © 2018 AIP Publishing.

dashed lines in the figure. From this linear-fit data analysis, the T g of these C60 composite thin films is found to follow a simple linear sum equation: Tgc = 𝜅 ⋅ wD + Tg0

(9.4)

where T gc is the glass transition temperature of the C60 doped organic thin-films, 𝜅 is a fitting parameter, wD is the weight fraction of C60 doping concentration, and T g0 is the glass transition temperature of the pure organic thin film. The 𝜅 values of these C60 composite thin films are listed in Table 9.2. This linear relationship suggests that there is an interaction between organic molecules and C60 molecules. As C60 molecule is a very strong electron acceptor, there is a strong π–π interaction between the benzene ring in the organic molecules and C60 . As a result of this interaction, a permanent dipole between organic–C60 molecules may tangle the motion of the constituent organic molecules and thus increases T gc . Equation (9.4) shows that T g depends only on κ for different C60 composite thin films. Therefore, κ is an indicator of the strength of the interaction between C60 and organic molecules. Based on this argument, we speculate that 𝜅 may be somewhat related to molecular weight. Based on experimental data, we discover that κ and M follows a simple relationship 𝜅 =𝛼⋅M

(9.5)

where M is the molecular weight, 𝛼 = 1.5 × 10−3 is a constant determined experimentally. The calculated 𝛼, 𝜅, and M of these three C60 composite thin films are given in Table 9.2. Then, a universal formula for calculating T g of C60 –organic composite thin films can now be given by a simple mathematical equation with only one variable, wD Tgc = 𝛼 ⋅ M ⋅ wD + Tg0

(9.6)

Therefore, equipped with this equation, the T gc of all organic–C60 composite thin-films can now be quantitatively computed by the C60 concentration. To illustrate the validity of Eq. (9.6), the ΔT gc (T gc − T g0 ) now can be calculated for various composite thin films. Figure 9.8 plots the computed theoretical ΔT gc versus the measured experimental ΔT g . The excellent fit indicates Eq. (9.6) can be used effectively to predict all T g of C60 composite thin films.

9.5 Summary

70

60

Experimental ΔTg (K)

50

40

30

20 TPD:C60

10

NPB:C60 m-MTDATA:C60

0 0

10

20

30

40

50

60

70

Theoretical ΔTgc (K)

Figure 9.8 The plot of ΔT g versus ΔT gc . ΔT gc is the theoretical value calculated by Eq. (9.6), and ΔT g is the experimental value. Source: Yang et al. [18]. © 2018 AIP Publishing.

9.5 Summary Thermal stability of amorphous organic thin-film plays an important role in determining the performance of organic optoelectronic devices. In a device, the organic film having the lowest glass transition temperature T g dictates the maximum operating temperature of an OLED product. Synthesis of high T g organic molecules without disrupting other optoelectronic properties and manufacturability is quite difficult. Thus, organic–organic and nano-organic composites provide practical thin-film engineering methods to boost glass transition temperatures of existing compounds. In this chapter, glass transition temperatures of pure and composite organic PVD thin-films are discussed. Experimentally, variable temperature spectroscopic ellipsometer is shown to be a reliable method for measuring T g of organic thin-films. All types of PVD thin-film systems exhibit disruptive changes in thicknesses at their glass transition temperatures. Thus, T g can be defined as the mid-point temperature of this disruptive change regime. For one component PVD organics, the thin-film glass transition temperature can be calculated by adding 4.7 K to the glass transition temperature of its raw organic powder. For non-interaction organic–organic composite thin-films, the inverse sum Fox equation is shown to provide an excellent prediction of the composite T gm . For interacting nano-organic composite thin-films,

295

296

9 Glass Transition in Organic Semiconductor Thin Films

a simple universal linear sum equation is shown to provide an accurate prediction of the glass transition temperature at any given composition.

Acknowledgments Financial support for this work is provided by the National Natural Science Foundation of China (Grant Nos. 11774304 and 11804294) and by the Natural Science and Engineering Research Council of Canada.

List of Abbreviations CTE DSC OLED PVD

coefficient of thermal expansion differential scanning calorimetry organic light emitting diode physical vapor deposition

References 1 Forrest, S.R. (1997). Ultrathin organic films grown by organic molecular beam deposition and related techniques. Chemical Reviews 97: 1793–1896. 2 Walzer, K., Maennig, B., Pfeiffer, M., and Leo, K. (2007). Highly efficient organic devices based on electrically doped transport layers. Chemical Reviews 107: 1233–1271. 3 Chang, Y.L. and Lu, Z.H. (2014). Organic light emitting diodes. In: Wiley Encyclopedia of Electrical and Electronics Engineering (ed. J. Webster), 1–15. London: Wiley. 4 Adachi, C. (2014). Third-generation organic electroluminescence materials. Japanese Journal of Applied Physics 53: 060101. 5 Zhao, W., He, Z., Lam, J.W. et al. (2016). Rational molecular design for achieving persistent and efficient pure organic room-temperature phosphorescence. Chem 1: 592–602. 6 Yagi, I., Hirai, N., Miyamoto, Y. et al. (2008). A flexible full-color AMOLED display driven by OTFTs. Journal of the Society for Information Display 16: 15–20. 7 Gu, G., Burrows, P.E., Venkatesh, S. et al. (1997). Vacuum-deposited, nonpolymeric flexible organic light-emitting devices. Optics Letters 22: 172–174. 8 Swallen, S.F., Kearns, K.L., Mapes, M.K. et al. (2007). Organic glasses with exceptional thermodynamic and kinetic stability. Science 315: 353–356. 9 Forrest, J., Dalnoki-Veress, K., Stevens, J., and Dutcher, J. (1996). Effect of free surfaces on the glass transition temperature of thin polymer films. Physical Review Letters 77: 2002.

References

10 Boucher, V.M., Cangialosi, D., Alegría, A., and Colmenero, J. (2017). Reaching the ideal glass transition by aging polymer films. Physical Chemistry Chemical Physics 19: 961–965. 11 Debenedetti, P.G. and Stillinger, F.H. (2001). Supercooled liquids and the glass transition. Nature 410: 259–267. 12 Scholz, S., Kondakov, D., Lüssem, B., and Leo, K. (2015). Degradation mechanisms and reactions in organic light-emitting devices. Chemical Reviews 115: 8449–8503. 13 Shibata, M., Sakai, Y., and Yokoyama, D. (2015). Advantages and disadvantages of vacuum-deposited and spin-coated amorphous organic semiconductor films for organic light-emitting diodes. Journal of Materials Chemistry C 3: 11178–11191. 14 Ohisa, S., Pu, Y.J., Yamada, N.L. et al. (2015). Molecular interdiffusion between stacked layers by solution and thermal annealing processes in organic light emitting devices. ACS Applied Materials and Interfaces 7: 20779. 15 Smith, A.R., Lee, K.H., Nelson, A. et al. (2012). Diffusion – the hidden menace in organic optoelectronic devices. Advanced Materials 24: 822–826. 16 Mcewan, J.A., Clulow, A.J., Nelson, A. et al. (2017). Dependence of organic interlayer diffusion on glass-transition temperature in OLEDs. ACS Applied Materials and Interfaces 9 (16): 14153–14161. 17 Tokito, S., Tanaka, H., Noda, K. et al. (1997). Temperature dependences of electroluminescent characteristics in the devices fabricated with novel triphenylamine derivatives. IEEE Transactions on Electron Devices 44: 1239–1244. 18 Yang, H.N., He, S.J., Zhang, T. et al. (2018). Nano-composites for enhanced catastrophic failure temperature of organic light-emitting diodes. Applied Physics Letters 113: 163301. 19 Pham, J.Q. and Green, P.F. (2002). The glass transition of thin film polymer/polymer blends: interfacial interactions and confinement. The Journal of Chemical Physics 116: 5801–5806. 20 Keddie, J.L., Jones, R.A., and Cory, R.A. (1994). Size-dependent depression of the glass transition temperature in polymer films. EPL (Europhysics Letters) 27: 59. 21 Richard, A. (1994). Interface and surface effects on the glass-transition temperature in thin polymer films. Faraday Discussions 98: 219–230. 22 Su, S.J., Sasabe, H., Takeda, T., and Kido, J. (2008). Pyridine-containing bipolar host materials for highly efficient blue phosphorescent OLEDs. Chemistry of Materials 20: 1691–1693. 23 Holmes, R., Forrest, S., Tung, Y.-J. et al. (2003). Blue organic electrophosphorescence using exothermic host–guest energy transfer. Applied Physics Letters 82: 2422–2424. 24 Yang, H.N., He, S.J., Zhang, T. et al. (2018). Glass transition temperatures in pure and composite organic thin-films. Organic Electronics 60: 45–50. 25 Singh, S., Ediger, M.D., and de Pablo, J.J. (2013). Ultrastable glasses from in silico vapour deposition. Nature Materials 12: 139.

297

298

9 Glass Transition in Organic Semiconductor Thin Films

26 Dalal, S.S., Walters, D.M., Lyubimov, I. et al. (2015). Tunable molecular orientation and elevated thermal stability of vapor-deposited organic semiconductors. Proceedings of the National Academy of Sciences 112: 4227–4232. 27 Inada, H., Yonemoto, Y., Wakimoto, T. et al. (1996). Organic electroluminescent devices using novel starburst molecules, 1,3,5-tris[4-(3-methylphenyl-phenylamino)phenyl] benzene and 4,4′ ,4′′ -tris(3-methyl-phenylphenylamino) triphenylamine, as hole-transport materials. Molecular Crystals and Liquid Crystals 280: 331–336. 28 Su, S.J., Chiba, T., Takeda, T., and Kido, J. (2008). Pyridine-containing triphenylbenzene derivatives with high electron mobility for highly efficient phosphorescent OLEDs. Advanced Materials 20: 2125–2130. 29 Mayr, C. and Brütting, W. (2015). Control of molecular dye orientation in organic luminescent films by the glass transition temperature of the host material. Chemistry of Materials 27: 2759–2762. 30 Kuwabara, Y., Ogawa, H., Inada, H. et al. (1994). Thermally stable multilared organic electroluminescent devices using novel starburst molecules, 4,4′ ,4′′ -tri(N-carbazolyl) triphenylamine (TCTA) and 4,4′ ,4′′ -tris(3-methylphenylphenylamino) triphenylamine (m-MTDATA), as hole-transport materials. Advanced Materials 6: 677–679. 31 Hancock, B.C. and Zografi, G. (1994). The relationship between the glass transition temperature and the water content of amorphous pharmaceutical solids. Pharmaceutical Research 11: 471. 32 Usluer, O., Demic, S., Egbe, D.A. et al. (2010). Fluorene–carbazole dendrimers: synthesis, thermal, photophysical and electroluminescent device properties. Advanced Functional Materials 20: 4152–4161. 33 Fox, T.G. and Flory, P.J. (1950). Second-order transition temperatures and related properties of polystyrene. I. Influence of molecular weight. Journal of Applied Physics 21: 581–591. 34 Fox, T.G. and Loshaek, S. (1955). Influence of molecular weight and degree of crosslinking on the specific volume and glass temperature of polymers. Journal of Polymer Science 15: 371–390. 35 Yuan, Y., Grozea, D., and Lu, Z. (2005). Fullerene-doped hole transport molecular films for organic light-emitting diodes. Applied Physics Letters 86: 143509.

299

10 Thermoelectric Films for Electricity Generation Metin Yurddaskal 1,2 , Melis Yurddaskal 3 , Ozan Yilmaz 1 , and Serdar Gultekin 1 1 Dokuz Eylul University, Engineering Faculty, Department of Metallurgical and Materials Engineering, Adatepe District Dogus Street No: 207/I, Buca, Izmir 35390, Turkey 2 Dokuz Eylul University, Center for Fabrication and Application of Electronic Materials, Adatepe District Dogus Street No: 207/I, Buca, Izmir 35390, Turkey 3 Manisa Celal Bayar University, Engineering Faculty, Department of Mechanical Engineering, Yunusemre District, Muradiye, Manisa 45140, Turkey

10.1 Introduction Most of the human activities require a power source. Fossil energy sources are limited and will disappear. Electricity is one of the essential energies of modern life. We need electricity to manage our daily routine such as transportation, communication, heating, food storing, etc. The need of electricity has increased since its discovery. Our modern life activities inevitably depend on electricity; therefore, solutions to generate electricity lies in renewable sources. Until today there have been several kinds of renewable energy sources that we used like wind, water, sunlight, etc. There is also one more relatively new renewable energy source – global warming. It is possible to use global warming in a positive way. It gives us a good opportunity to generate electricity, thanks to thermoelectricity. Heat that we radiated to do other task may be used to generate electricity. The solution lies in devices capable of recovering energy from the environment surrounding the device or the user. Even human body radiates heat due to metabolic activities. In addition, many other devices that could be used as heat sources to generate electricity include air conditioners, exhaust systems in vehicles, and industrial processes. Thermoelectric generators are able to use the waste heat radiated from these devices, which are not used, to produce heat to generate electricity. Recently, one of the most promising solutions has been considered as direct conversion of heat into electricity via advanced thermoelectric materials (TE). Over the last couple of years, much achievement has been maintained in their dimensionless figure of merit (ZT), which therefore has a word in calculating conversion efficiency of TE devices. It can be said that ZT is associated with different interlocked factors such as the Seebeck coefficient, electrical conductivity, and thermal conductivity, in which it is very odd that these factors are interrelated in bulk TE materials, Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications, First Edition. Edited by Yujun Song. © 2021 WILEY-VCH GmbH. Published 2021 by WILEY-VCH GmbH.

300

10 Thermoelectric Films for Electricity Generation

which means if you modify one, it has effects on the others. In this chapter, some TE fundamentals will be introduced, and a thorough examination will be made in inorganic and organic TE materials in order to see if there are any improvements and to see is there is a future for them. It seems inevitable to be sustainable technology called thermoelectric effect since there is a high demand for energy all over the world and much concern regarding climate and diminishing fossil fuels, showing effort for using, converting, and recovering sources. The main principle of TE conversions can be seen described elsewhere. Among the advantages of this technology are as follows: (i) when compared with others, TE is more silent and reliable since no movement is involved in the process; (ii) they are simple, compact, and safe; (iii) it is a green technology since no heat, no gas, or no chemical disposals are produced during the process; and (iv) it is convenient when it comes to working in remote outer areas.

10.2 Thermoelectricity The origin of thermoelectric effect dates back to 1822. Thomas Seebeck observed that when a temperature gradient was applied to the junction of two different materials (metal or semiconductor), the mobile charge carriers at the hot end tend to diffuse to the cold end, resulting in electric current [1]. This mechanism is just like the temperature sensors in Figure 10.1. Seebeck effect “α” can be described by voltage differences caused by difference of temperature: ΔV , (𝜇V∕K) (10.1) ΔK Unaware of Thomas Seebeck’s invention, Jean Peltier discovered in 1834 that heat could be generated if the electric current was applied to the junction of two materials. He also stated that it was possible to absorb heat by changing current direction [2]. Even if the inventions of Thomas Seebeck and Jean Peltier were related to each other, it would be injustice to say that these two inventions are not independent. In 1838, Heinrich Lenz produced a setup made out of bismuth (Bi) and antimony (Sb) shown in Figure 10.2. In order to observe temperature changing at the junction of Bi and Sb, he used a drop of water. He was able to freeze the water droplet by applying electric current. When he reversed the direction of the current, the ice melted. In this respect Heinrich Lenz created the first thermoelectric device. In 1848, William Thomson (Lord Kelvin) came up with new formula that defined the Peltier coefficient (𝜋) using Seebeck coefficient (𝛼): 𝛼=

𝜋 = 𝛼T, (V)

(10.2)

where T represents the temperature at the junction of the materials [3]. Thomson took his work a step further and revealed that there might be a third effect. This effect states that an electrical conductor cools or heats with electric current if it is exposed to temperature gradient. Thermoelectric studies were continued by other scientists John William Strutt and Edmund Altenkrich [4–6]. Especially, Altenkrich revealed considerable studies. One of the most important properties of

10.3 Overview of Inorganic and Organic Thermoelectrics for Thin Films

Electric energy generated

Generated voltage

i

i

V

Material A

Load

Material A

Material B

Heart source (a)

Material B

Heart source (b)

Figure 10.1 An experimental design of Seebeck experiment: (a) temperature sensor and (b) thermoelectric generator.

thermoelectric materials is its figure of merit (Z), which defines the efficiency of thermoelectric materials. According to Altenkrich, good thermoelectric materials should have high Seebeck coefficient and low thermal conductivity (λ). Low thermal conductivity retains heat at the junction of two materials. In addition, another characteristic of good thermoelectric material is low electrical resistivity (ρ). Electrical resistivity needs to be as low as possible to obtain low Joule heating. The figure of merit of a thermoelectric material can be assigned by the following formulas: σ Z = α2 , [K−1 ] (10.3) λ where “σ” is electrical conductivity that is the inverse of resistivity “ρ ” : l (10.4) ρ = R , [Ω − cm−1 ] A Because the temperature can affect the figure of merit, “T” should be considered and the formulation modified as below: T T or ZT = α2 σ (10.5) ZT = α2 𝜌𝜆 λ

10.3 Overview of Inorganic and Organic Thermoelectrics for Thin Films 10.3.1 The Seebeck Effect As we mentioned before, low thermal conductivity and high electrical conductivity are the most important parameters for high Seebeck coefficient. However, these

301

302

10 Thermoelectric Films for Electricity Generation

Drop of water ice

Drop of meltwater

H2O H2O Bismuth (Bi)

Antimony (Sb)

Bismuth (Bi)

Heart absorbed

Heart released

Electric current (a)

Figure 10.2

Antimony (Sb)

Electric current (b)

Experimental setup of the Peltier effect: (a) cooling and (b) heating.

three parameters are difficult to be all best, because they are function of carrier concentration, which cannot be varied independently. As the carrier concentration increases, the Seebeck coefficient decreases, and thermal and electrical conductivities increase [7]. In this respect carrier concentration should be at optimal value to maximize the power factor α2 σ (Figure 10.3). The sophisticated relationships between characteristic parameters of the materials complicate the approach of adjusting carrier concentration; alone it is difficult to improve the figure of merit. In the last few decades, to increase power factor and reduce thermal conductivity, great developments have been carried out. It is possible to divide thermoelectric materials into three sections depending on their optimal working temperature: low temperature, middle temperature, and high temperature. These temperatures are typically below 400 K, between 600 K and 900 K, and above 900 K, respectively. Bi2 Te3 -based thermoelectric materials belong to low temperature. PbTe-based materials belong to middle temperature. SiGe-based materials belong to high-temperature thermoelectric materials. Since the emergence of the Seebeck effect in 1821, thermoelectric materials have been divided into three generations according to ZT values [8]: ZT values 0 for p-type semiconductors and 𝛼 n < 0 for n-type semiconductors. The Seebeck coefficient, also known as entropy per charge carrier, is generally correlated with the density of energy levels. While undoped semiconductors have the Seebeck coefficient values on the order of mV/K, the Seebeck coefficient values of doped semiconductors decrease typically, and it even reaches the order of a few μV/K for highly doped materials [37]. The unit of the Seebeck coefficient is usually expressed as V/K (μV/K or μV/∘ C). It has been found that the Seebeck effect shows only in a material called a thermocouple consisting of a combination of two different materials. Although both ends of the same material actually have intrinsically Seebeck coefficient, they do not possess a Seebeck effect as a result of symmetry. The Seebeck effect, however, is a bulk property and does not depend on the specific arrangement of the tip or material, nor the specific method of joining them [38]. Conductive materials have different Seebeck coefficients or thermoelectric sensitivities. This Seebeck coefficient can be positive or negative. For example, the Seebeck coefficient of iron is 19 μV/∘ C at 0 ∘ C, which means that a positive Seebeck voltage (or thermoelectric emf) of 19 μV is induced in iron for every 1 ∘ C difference in temperatures near 0 ∘ C. As an example with a different Seebeck coefficient, constantan (a copper–nickel alloy) has a thermoelectric emf of −35 μV/∘ C at 0 ∘ C. The semiconductors are more promising for the construction of thermocouples since the Seebeck coefficient is generally greater than 100 μV/K, while that of metals are 10 μV/K or less. Looking the relationship between thermoelectric voltage and temperature, the relationship is linear only for small changes in temperature, and it becomes nonlinear if the temperature difference is increased [7]. It is therefore necessary to specify the temperature at which the thermoelectric coefficient is being used [39]. A load must be connected to the thermoelectric material (or series of materials) in order to convert the voltage produced by the Seebeck effect into electricity, which will result in an electrical current. In order to maximize the electrical energy generated, it is necessary to minimize both the electrical losses due to Joule heating and the thermal losses caused by the heat flow between the hot and cold junctions. Therefore, the thermoelectric material should exhibit high electrical conductivity (𝜎) and low thermal conductivity (k). The high electrical conductivity and the low thermal conductivity required by the high Seebeck coefficient, when combined with a dimensionless number called thermoelectric figure of merit (ZT), allow comparison of the yields of different thermoelectric materials as given in Eq. (10.6).

10.3 Overview of Inorganic and Organic Thermoelectrics for Thin Films

10.3.2 The Peltier Effect Thirteen years after the discovery of the Seebeck effect, Jean Charles Athanase Peltier, a French watchmaker, invented the second of thermoelectric effects in 1834. Peltier has found that an electric current creates a minor heating or cooling effect, depending on the direction of its passage through the thermocouple. Since the Joule heating effect is also present, it is difficult to see the Peltier effect in metallic thermocouples. When the current is passed in one direction, it is sometimes difficult to determine whether heating or cooling is greater. If an arrangement as in Figure 10.6 is used, the Peltier effect can be demonstrated, theoretically, by connecting the galvanometer or with a direct current source and placing a small thermometer on the thermocouple junction [40]. In this effect, cooling effect is seen in one junction, and the temperature increases in the other junction, when the electric current passes through a circuit consisting of two different conductive materials. This makes the Peltier effect inverse of the Seebeck effect. Let us take the example that copper conductor wires are connected to both ends of the battery as in Figure 10.7 and the circuit is completed with bismuth between these copper wires. It was observed that when the circuit was closed, a temperature gradient occurred as predicted by the Peltier effect phenomenon. At the point where the current passes from copper to bismuth, the temperature increases, while at the point where current passes from bismuth to copper, the temperature decreases. Also, symbolic representation of the Peltier effect is shown in Figure 10.8. Figure 10.6 An experimental design to demonstrate the Peltier effects.

Heat source or thermometer Conductor A

Conductor B

Galvanometer or electric current source

Bismuth

Cold Junction

Hot junction

Copper

+– Current

Figure 10.7

Voltage

Schematic representation of the Peltier effect.

307

308

10 Thermoelectric Films for Electricity Generation

Cold side Isolator (ceramics)

p

n

Conductor (copper)

p

p

n n p p Semiconductor of p-type Semiconductor Hot side of n-type –

Figure 10.8

n +

Symbolic representation of the Peltier effect. I

Figure 10.9 A schematic illustration explaining the Peltier Effect.

Material A

Qp = αIT

Reject Material B Absrorb

I

In other words, the Peltier effect can explain the temperature difference that is generated by the voltage applied between two electrodes connected to a semiconductor material. In this effect, if an electrical current is passed through the junction of two different materials, heat is generated or absorbed, as can be seen in Figure 10.9, depending on the direction of the current. In this figure, 𝛼 shows the Seebeck coefficient. In fact, this effect is due to the difference between the Fermi energy of the two materials. When a current (I) is passed through a circuit as in Figure 10.10, heat is generated at the upper junction (at T 2 ) and is absorbed at the lower junction (at T 1 ). The amount of Peltier heat (Q) absorbed per unit time at the lower junction can be given by Q = ΠAB I = (ΠB − ΠA ) I

(10.8)

where ΠA and ΠB are the Peltier coefficients of the each material and ΠAB is the coefficient of the entire thermocouple. As the names imply, the Peltier coefficient of

10.3 Overview of Inorganic and Organic Thermoelectrics for Thin Films

Figure 10.10 Demonstration of the application of the Peltier effect for cooling experiment.

Active cooling p

n

Heat rejection



+

I

the p-type semiconductors is typically positive (though not above ∼550 K), whereas that of the n-type semiconductors is negative. The major advantage of the Peltier effect is that it can create effectively heating or cooling without moving parts; thus, the probability of malfunction is much lower than conventional heating or cooling systems. They are also virtually maintenance-free. The devices operating using the Peltier effect are very quiet and can operate at temperatures as low as −80 ∘ C (−176 ∘ F). The Peltier effect can work successfully even at microscopic levels where conventional cooling systems would not work. In applications, Peltier devices comprising n-type and p-type materials are connected electrically to each other in series and thermally in parallel. Figure 10.10 demonstrates a Peltier device operating in cooling mode.

10.3.3 The Thomson Effect At first, the dependence of the Seebeck and Peltier phenomena on one another was not immediately realized. Later, the Thomson effect, called the third thermoelectric effect, discovered by William Thomson (later known as Lord Kelvin) in 1854 is a combination of Seebeck and Peltier effects [21]. Thomson found that the relationship between the two effects should have an additional effect. The theoretical definition of this effect is that the homogeneous conductor carrying an electric current between two points with different temperatures either emits or absorbs the heat depending on the direction of the current and material, which is additively the Peltier heating. For example, if a copper wire carrying a constant electric current is exposed to external heating in a short portion while the remaining portion remains colder, the heat is absorbed from the copper, as the conventional current approaches the hot spot, and the heat is carried away from the hot spot. The heat absorbed or emitted is called the Thomson heat (Q) and is given by the equation Q = 𝜌J 2 − 𝜇 ∗ J

dT dx

(10.9)

where 𝜌 is the resistivity of the material, J is the current density, 𝜇* is the Thomson coefficient, and dT/dx is the temperature gradient through the conductor. The term 𝜌J 2 is called irreversible Joule heating always produced as a current in a wire, and

309

310

10 Thermoelectric Films for Electricity Generation

QThomson.A Wire A

QPeltier.AB

TL

I



+

TH

QPeltier.AB

QThomson.B

Wire B

Figure 10.11

Schematic demonstration of the Peltier effect and the Thomson effect.

the second term is Thomson heating whose sign changes according to the direction of the current [41]. Thomson also examined the relationship between these three effects thermodynamically, demonstrating that the electrical Seebeck effect is due to a combination of thermal Peltier and Thomson effects. Despite the Thomson effect being small, it leads to a very useful and important relationship called the Kelvin relationship, with Seebeck and Peltier effects. When the current flows as shown in Figure 10.11, the heat is absorbed in wire A because of the negative temperature gradient and released in wire B due to the positive temperature gradient. The Thomson heat is proportional to the electric current and the temperature gradient shown schematically in Figure 10.11 and Eq. (10.9). The Thomson coefficient is the only thermoelectric coefficient that can be measured directly for individual materials. The heat absorption is a proof of an electromotive force moving in the same direction as the current, which means that electrical energy is supplied to the circuit. Therefore, the Thomson effect for iron would lead to a flow from heat to cold regions in iron. In addition to iron, many metals such as bismuth, cobalt, nickel, and platinum exhibit the same property called a negative Thomson effect. That is, in these metals, the high-temperature portion has a lower potential than a low-temperature portion, and therefore, heat energy is absorbed when the current flows from a high temperature point to a low temperature point. The opposite of this situation explains the positive Thomson effect. The potential at high temperature is considered to be higher than a section at low temperature. When the new current flows from a low-temperature point to a high-temperature point, the heat energy is absorbed. Metals showing positive Thomson effect are copper, silver, zinc, antimony, cadmium, etc. Lead is a metal with zero Thomson effect. In some metals, changes in temperature or crystal structure can reverse the effect sign [21]. The Thomson coefficient is positive for materials with positive Thomson effect. The term in Eq. (10.9) is simply the Joule heating that is always released when a current pass through an imperfect conductor. There is no relationship with the Thomson effect, but it is included in the equation for integrity. Studies have shown that the total Thomson emf along a conductor depends only on the temperatures of

10.4 Classification of Thin Film Thermoelectric (TE) Materials

the both ends and is not in any way dependent on a particular shape in which the temperature gradient changes.

10.4 Classification of Thin Film Thermoelectric (TE) Materials In this part of the book, after making a general classification for thermoelectric films, general characteristics of these films and some numerical values are mentioned. In this context, the experimental data obtained from the studies we have collected from the literature have been presented to the reader.

10.4.1 Inorganic Thermoelectric Thin Films 10.4.1.1 Bi–Te-Based Superlattices

The performance of Bi—Te materials has been tried to be improved by adding dopant material to crystal lattices in the last 20 years. Compared with pure Bi2 Te3 and Ag-doped BiTe3 , the results showed that the doped structure had lower thermal conductivity. It also has a low power factor. Thanks to the 2% silver dopant added to Bi2 Te3 , ZT value of 0.77 was obtained at 450 K temperature [30]. Similar studies have been carried out for materials produced in thin film form that showed great performance [30, 42]. In some similar studies that used low-temperature growth process in metal–organic chemical vapor deposition (MOCVD) technique to prepare Bi2 Te3 film with a thickness of 10 Å p-type Bi2 Te3 /Sb2 Te3 super lattices and n-type Bi2 Te3 /Bi2 Te2 .83Se0.17 super lattices, thin films showed ZT value of ∼2.4 and ∼ 1.4 at 300 K [30, 43]. In addition, there are also studies in the literature of BiTeSb thin film that have been formed by methods such as flash evaporation (FE) method [30]. However, it can be said that Bi—Te-based films are relatively more expensive than other conventional material. Recently, Zn- and Cu-based thermoelectric films, which can be produced at a lower cost than Bi—Te-based thermoelectric films, have started to attract attention [19, 44–46]. 10.4.1.2 Cobalt Oxide-Based Thin Films

Cobalt oxide-based thin films exhibit a high power factor in the in-plane direction [47, 48]. The NaCo2 O4 , an example of single-crystal thin film, showed the strength of 100 μV/K at 300 K [49]. In addition to these investigations, Ca3 Co4 O9 has been found to be used in terms of thermopower at room temperature. The single crystal of Ca3 Co4 O9 showed resistance of 10–40 mΩ cm, 125 μV/K thermopower, and 0.04–0.16 mW /(m K2 ) power factor, in order [50]. Bendable and flexible nanostructured Ca3 Co4 O9 thin films showed high power factor above 1 × 10−4 W/(m K2 ). They also did not lose significant performance when bended at high temperature. Porosity is one of the important properties for the thermoelectric futures. Nanoporous structures in the Ca3 Co4 O9 caused selective scattering phonons. In this way thermoelectric performance could be improved, yielding a power factor of 2.32 × 10−4 W/(m K2 ) near room temperature. The lowest electrical resistivity for CaCo4 O9 is around 7 mΩ cm [51].

311

312

10 Thermoelectric Films for Electricity Generation

Furthermore, by using simple dry transfer, it is possible to transfer these materials from their initial substrates to arbitrary polymer platforms. This is one of the important signs that the potential for wearable applications is quite high. In addition, it has been reported in some studies that the layered calcium cobalt oxides and strontium cobalt oxides synthesized from sodium cobalt oxide starting material exhibit high power factor values. In these studies, the highest power factor values for layered Ca0.33 CoO2 and Sr0.29 CoO2 at 300 K were found to be 9 × 10 −4 W/(m K−2 ) and 3.5 × 10−4 W/(m K−2 ). This value for Ca0.33 CoO2 is 25% higher than that of Ca3 Co4 O9 [52–54]. In thermoelectric power generation applications, the power factor for a specified ZT value is considered to be a more important material property than thermal conductivity [55]. Therefore, in applications where low power is used, it is preferable to obtain high power factors before reducing thermal conductivity. 10.4.1.3 Zn-Based Thin Films

ZnO thin films can be used in a wide range of applications in energy-related area. One of them is thermoelectric film studies. Among thermoelectric materials, ZnO materials got great attention with its high Seebeck coefficient and high thermal stability. As in Bi–Te-based electrical materials, doping method is usually used in Zn-based thermoelectric materials to improve performance [56, 57]. ZnO thin films have a number of advantages to engineers compared with bulk materials as they can be produced at nanoscale. In the case of thermoelectric materials, major improvements have been achieved in Bi–Te-based materials with nano-sized defects to control phonon scattering [12, 30]. The main problems of Bi–Te-based materials are their decomposition at temperatures above 600 ∘ C and the limitation of possible practical applications [58]. Nano-defects in bulk oxide materials such as ZnO are randomly distributed, and it is difficult to control their density and size. ZnO thin films are good candidate to solve this kind of problems [58]. Studies have shown that ZnO thermoelectric properties can be improved by doping Al into its crystal structure. Besides aluminum, titanium, antimony, and nickel are used frequently as doping element [59–63]. 10.4.1.4 Cu-Based Thin Films

No matter whether it is bulk or thin film, Cu-based materials are of considerable interest for thermoelectric applications [64–68]. In example of study about Cu2 Se bulk materials, it has been reported reaching the highest ZT value of 2.4 at 1000 K [69]. This superior ZT value of Cu2 S material is associated with its low latency thermal conductivity [64, 65]. Already bulk Cu2 S materials exhibit better thermoelectric properties than Cu2 S materials produced in the form of thin films. A relatively new approach, the flexible Cu2 S, has recently begun to be tested. According to one of these studies, the Cu2 Se powder was dissolved in the organic solution, making an ink solution. This ink solution was then deposited on a flexible substrate using the wet coating method. The Cu2 S material produced in this way showed a power factor of 0.62 mW/(cm K) at 684 K [44]. Furthermore, thermoelectric materials that are capable of operating at low temperatures, such as room temperature, will provide

10.4 Classification of Thin Film Thermoelectric (TE) Materials

a significant improvement for different application such as wearable and portable devices. Another Cu-based thermoelectric materials is Cul. Recent research explored that γ-CuI has low thermal conductivity as well as high hole conductivity and Seebeck coefficient. The thickness of the films produced in this study was around 200–300 nm while the ZT values were obtained as 0.21 at 300 K [19]. The conductivity of the p-type Cul film obtained by the sputter method was 156 S/cm, while the Cul film with iodine doped was 283 S/cm [66].

10.4.2 Organic-based Thin Film TE Materials Organic polymers also have low thermal conductivity. We can also refer to these materials in the form of semimetallic polymers. Due to the low thermal conductivity of the polymers, they exhibit significant advantages over other conventional materials in the thermoelectric material fields [70–73]. The abundance of polymers and their light weight, flexibility, and low cost are also other advantageous aspects [74]. But one of the biggest disadvantages of the polymer materials is their low efficiency. Efficiency of such materials is tried to be increased by doping. Various polymers of p-type and n-type can be produced by this method. Several examples are polyacetylene, polyaniline, polypyrrole, polythiophene, and poly(3,4-ethylenedioxythiophene) [75–79]. 10.4.2.1 Polyacetylene and Polyaniline

In a study, iodine-doped polyacetylene film showed better conductivity than pure film. Its conductivity value was increased from 3 × 103 S/cm to 1 × 105 S/cm [80]. Metal chlorides FeCl3 , MoCl5 , NbCl5 , and ZrCl4 can also be added to the structure of this polymer to increase the efficiency of it. According to a study, conductivity measurements made at different temperatures, FeCl3 -doped polyacetylene film showed the best result with 3 × 105 S/cm conductivity at 220 K [81]. However, the polyacetylene material has a significant disadvantage, which is its instability under atmospheric conditions. Another remarkable polymer material is the composite of Bi0.5 Sb nanoplate-doped polyaniline. This material has a relatively high Seebeck coefficient, and therefore it possesses a power factor of 16.5 × 10−8 to 84.4 × 10−8 W/(m K2 ) at 400 K [82]. Polyaniline material, on the other hand, stands out as a material that can meet the stability requirement under atmospheric conditions and is produced on a large scale at low cost [83, 84]. The composite material consisting of polyaniline and camphorsulfonic acid (CSA) exhibits good electrical conductivity. This conductivity can go up to metallic levels with regard to the doping level. For example, a high conductivity value of 300 S/cm could be obtained in doping at 60% [85]. 10.4.2.2 Poly(3,4-ethylenedioxythiophene)

Poly(3,4-ethylenedioxythiophene) polymer is another remarkable material. This polymer can exhibit high electrical conductivity along the casting plane, especially when mixed with polystyrene sulfonate (PEDOT: PSS) [86]. Thermal and electrical

313

314

10 Thermoelectric Films for Electricity Generation

in-plane conductivities of drop-cast films characterized by 20 μm are 1.0 W/(m K2 ) and 500 S/cm, respectively. Drop-cast dimethylsulfoxide (DMSO)-mixed PEDOT: PSS films showed the characterization of the correlation between electrical conductivity and anisotropic thermal conductivity. As increasing electrical conductivity, anisotropic thermal conductivity tends to rise. The power factor of the DMSO posttreatment films is 65% increased when compare with DMSO added PEDOT: PSS films. When using DMSO posttreatment instead of DMSO addition in the PEDOT: PSS films, the power factor of the film is 65% increased. The increase in electrical conductivity contributes to the development of the power factor. While these changes occur, changes in the Seebeck coefficient remain at lower levels [87]. 10.4.2.3 Polypyrrole and Polythiophene

Polypyrrole films, unlike the aforementioned films, can exhibit high conductivity values at very low temperatures of about 0 K. This conductivity is higher than 300 S/cm [88]. High tensile stress-resistant polythiophene and poly(3-methylthiophene) nanofilms showed a value of 0.03 ZT at 250 K, which is higher than that of many other conducting polymers [89]. According to Sun et al., high conducting polymers tend to exhibit conductivity from carriers that are close to the Fermi level. However, their experimental result did not prove of their claim. A thickness of 100 nm of poly(3-hexylthiophene) thin film on Si substrate did not show that much performance. Their studies illustrate theoretical route for reaching a high Seebeck coefficient [90]. 10.4.2.4 Other n-Type Polymers

Examples of n-type organic thermoelectric materials include poly(nickel-ethylene tetrathionate) (poly[Ni-ett]). This material can reach its maximum ZT value of 0.3 at room temperature. Different production techniques are available for this material, for example, potentiostatic deposition method. Different production parameters lead to obtaining different optical power factor values. Increases in optical power factor up to 131.6 μW/(m K2 ) were observed with synthesis potential down to 0.6 V. This value was lower than that of at 1 or 1.6 V [91, 92]. Another n-type organic thermoelectric material is poly(benzobisimidazobenzophenanthroline) (BBL). Electron mobility and electrical conductivity values of this polymer are 0.1 cm2 /(V s) and 2.4 S/cm, respectively, with the field effect [93, 94]. Lastly perylene diimide (PDI)-based materials need to be referred to n-type thermoelectric materials. Their power factor can reach to 1.4 μW/(m K2 ) [95].

10.4.3 Inorganic–Organic Composite Thermoelectric Thin Film Materials The polymeric materials exhibit advantages such as low density, low thermal conductivity, easy reproducibility, and low cost and a number of disadvantages. Polymeric materials unfortunately do not perform well in terms of longevity. Their low electrical conductivity and ZT values also prevent them from being

10.5 Applications of Thermoelectric Materials

used in thermoelectric power generation applications. One of the important methods of dealing with such problems is to produce composite materials as in many applications. Composite materials, which exhibit the superior properties of inorganic and organic materials together, are promising. In this way, it becomes possible to eliminate the obstacles that prevent these materials from being used in thermoelectric applications. Recently, there has been an increase in the number of studies added to the literature with increasing interest in this subject [96, 97]. 10.4.3.1 Metal–Organic Frameworks

Metal-organic frames are a new type of thermoelectric materials. After observing the electrical conductivity of such composite materials, they became the subject of research for thermoelectric applications. One of the remarkable and common studied metal–organic frameworks is tetracyanoquinodimethane (TCNQ) and Cu3 (BTC)2 (BTC = benzene tricarboxylate) [98, 99]. This material draws attention with its higher Seebeck coefficient than some of the aforementioned materials. According to the measurement under the controlled temperature gradient, the Seebeck coefficient of this material is 375 μV/K [100]. Although the electrical performance of this material is not satisfactory, it is predicted that it can be improved in the following years and studies on this subject are continuing increasingly. 10.4.3.2 Carbon Nanotube–Polymer Composites

Carbon nanotube materials are one of the most remarkable technological materials that have attracted attention recently. Thanks to their superior electrical properties, it is possible to see them in many areas related to energy storage and thermoelectrics [101, 102]. The power factor values of the single-walled carbon nanotubes, which may also have elasticity, were 1.5 × 103 and 1.84 × 103 μW/(m K) for n-type and p-type, respectively [18]. When single-walled carbon nanotube (SWCNT)–polyaniline composite films were tested for thermoelectric measurements, a high power factor value of 217 μW/(m K) was found [103]. Polyaniline/carbon nanotube (CNT)/graphene multilayer film is another promising thermoelectric material. The power factor value of this material is 1.825 × 103 μW/(m K2 ) that is comparable with the others [104].

10.5 Applications of Thermoelectric Materials Bulk thermoelectric materials have widespread application fields. Power generation and cooling process can be counted as one. Nevertheless, their convenience in microscale applications is a bit demanding. If so, it is necessary to resort to thin film approach by which the films can be applied on the substrate. Equal conditions are difficult to maintain during the creation of films. Therefore, the properties of the materials can be distinct from their bulk counterpart. Consequently, thin film application rather than the bulk material usage is said to cause considerable trend upward in the figure of merit [30].

315

316

10 Thermoelectric Films for Electricity Generation

10.5.1 Thermoelectric Cooling One of the most important thermoelectric effects is the Peltier effect. It has many application fields such as refrigeration, cooling of electrical components, portable coolers, cameras, climate controlled jackets, spacecraft, satellites, etc. There are two main materials in thermoelectric devices; p-type and n-type, which are interconnected. The connections are made in electrical terms in series and in thermal terms in parallel. The main working principle of the process starts with applying a voltage to the lowest point of junction, and then free carriers start moving from the peak point of the junction downward. Therefore, the heat goes down to the bottom of the junction. Whether the device is effective or not is in relation with the figure of merit, as was described by Ioffe [105]. As can be seen from Eq. (10.10) [35], thermoelectric cooler performance of coefficiency can be calculated by 𝜑=

(sp − sn )ITC − kΔT − 12 I 2 R QC = P I[(sp − sn )ΔT + IR]

(10.10)

where Qc is the cooling rate, P is the power consumed, T C (T H ) is the cold side (hot side) temperature, I is the current, and R is total resistance. There come lots of conveniences with application of thermoelectric coolers such as being compact in volume, requiring minimal mend, having no mobile pats, and being trustworthy. Hence, they are convenient for micro-applications.

10.5.2 Thermoelectric Power Generation Seebeck effect is a way of generating useful power. Its main working principle is that it converts lost heat energy such as industrial plants, vehicles, boilers, stoves, or any kind of the heat-emitting item. Another usage of the effect is that it saves the lost heat from solar cells, powers, watches, and space vehicles [106]. The relationship of Seebeck coefficient, electrical conductivity, and thermal conductivity is shown in Figure 10.12. In this figure, 𝛼, 𝜎, 𝛼 2 𝜎, K e , and K l correspond to the Seebeck coefficient, electrical conductivity, power factor, and electronic and lattice parts of the thermal conductivity, respectively. The process of devices is to maintain a temperature gradient between junctions. The working principle is that while the heat source junction is keeping the temperature high, the heat sink junction draws the heat that keep the junction cold so that the source junction can be heated. Therefore, the outcome of this process is a maintained temperature gradient. As a result, charge carriers propagate from high temperatures to cold temperatures so that voltage or current can move through junctions [35]. Thermoelectric generators are a little less efficient than mechanical generators. Thus, it is essential to find the right materials to enhance TE efficiency.

10.5.3 Organic Inverter Circuit Presumably, an inverter can be appreciated as the simplest unit in digital circuitry. When compared with n-type materials, because of their higher field effect mobility

10.5 Applications of Thermoelectric Materials

Figure 10.12 The relationship of Seebeck coefficient and electrical and thermal conductivities for different materials.

Insulators

Semiconductors Metals

α

σ α2σ σ

α

opt

In(n) Ke

KI

Insulators

Semiconductors Metals

In(n)

and better intrinsic stability, p-type materials are highly preferred in designing organic thin film transistor (OTFT)-based inverters. Nevertheless, low static power consumption, high noise margin, and high gain and operational robustness are preferred properties of complementary organic inverters. Exhibiting mobility and threshold voltage is a must for p-type and n-type transistors. Nevertheless, we should remember that because n-type organic transistors have some drawbacks in high mobility, there is an arising need for an all-p-type organic inverter circuit. An all-p-type inverter also has some drawbacks. Low voltage swings, poor balance between pull-up and pull-down operations, higher power dissipation, and low noise margins are among the difficulties of those inverters. However, when we take those drawbacks into consideration, again there is an arising need for hybrid complementary circuits. Dodabalapur et al. [107] is the first person to maintain the technology of hybrid via adjusting n-type organic transistors with inorganic ones like a-Si: H TFT. All-p-type organic complementary inverters and hybrid inverters have different static and dynamic responses. The ideal value of Vdd/2 of noise margin can be maintained with a combination of pentacene and C60. It also has less propagation delay around 68% when compared with CuPc–F16CuPc combination. The main reason why it has less delay is that pentacene- and C60-based transistors have comparable mobility. On the other hand, when comparing CuPc-based transistors with F16CuPc-based ones, it can be seen that the former has 16 times higher mobility. Therefore, CuPc–F16CuPc combination brings about the highest delay ratio. Because F16CuPc has lower field effect mobility, the operational speed of transistors slows down and accordingly causes high delay ratio. The lowest propagation delay of 0.28 seconds

317

318

10 Thermoelectric Films for Electricity Generation

can be maintained with the hybrid inverter circuit pentacene–ZnO combination. This ratio is nearly %50 less than pentacene–a-Si: H. Meanwhile, all-p-type organics exhibit only an average ratio. The static performances of all-p-type organics are VoH = 8.0, VıH = 6.8, VoL = 0.2, and VıL = 4.4. That is to say, the upper voltage is lesser than the Vdd(10v) in contrast to lower voltage of 0.2 V, which is desirable. Nevertheless, it must be admitted that the noise margins of all-p-type organic inverters are not matched enough with silicon-based transistors.

10.5.4 Organic Light-Emitting Diode (OLED) The organic light-emitting diodes (OLEDs), which have been marketed up till now, are providing some benefits such as low- cost, large-area displays. They have proven their benefits in small displays such as mobile phones, programmable digital arrays, MP3 players, and digital cameras. The main reason why they have been chosen is that they have low costs, are lightweight and flexible, have modification advantages, have larger color scale, and have better efficiency. In addition, further developments can be made in color quality, sharp image, intensive background, larger sight angle, swift switching, and low voltage. In the manufacturing process of good quality flexible displays of electronic papers, these properties can be used, which is very much suitable since it matches with properties of OLEDs. An example of an OLED structure and its OTFT-based driving circuit can be seen in Figure 10.13. The OTFT1 shown in Figure 10.13 demonstrates the charging and discharging capacities. On the other hand, OTFT2 is the main driver. OTFT1 must have a high on current and low off current for effective charging and discharging; otherwise, a charge leakage might occur, which of course would require a large-size capacitor that should be charged as well. In general, a large on/off current ratio is necessary to operate such a large display. A high-performance pentacene-based TFT with a mobility ratio of 1.5 cm2 /(V s) can enable the display to point minimum a thousand lines and maintain minimum charge leakage in longer frame times [108]. The first active matrix organic light emitting diode (AMOLED), the first fully organic active matrix organic light emitting diode, is said by Zhou et al. [109] in 2006

V









+

+

+

+





+

+

Electron transort layer Recombination region Hole transport layer Anode

C OTFT1

OTFT2

OLED

Transparent V Data substrate

Glass

Figure 10.13



+

ITO

Light

– Alq3 ETL

CuPc HTL

VGate

Cathode

Silver

Output

A schematic representation of OLED driven by an OTFTs.

VCathode

GND

10.5 Applications of Thermoelectric Materials

to contain 48 × 48 bottom emission OLED pixels with two pentacene transistor per pixel. Li et al. [110] is said to be the first person to report a fully printed multilayer OLED manufactured via polymer inking and stamping technique. The process has been conducted via transferring a layer of PEDOT material on polyestersulfone (PES) substrate with a stamp of polydimethylsiloxane (PDMS) material. A yellow light is emitted from the bottom of the device, from the transparent indium tin oxide (ITO) substrate with a voltage of around 7 V. OTFT devices with a low temperature generally have low mobility that in turn impedes the designing process of pixel circuitry [111]. As a result, what is needed to turn the table in this process is a compensatory unit that is necessary for low mobility of OTFT circuits. Only after that, a constant driving current can be maintained that in turn provides with enough electrical operation for AMOLED display.

10.5.5 Organic Radio Frequency Identification Tags Nowadays, in areas like defense sector, medical, toll bridge, and supply chain management, there is a new player. Over the years, organic radio frequency identification (RFID) tags have raised much interest around scientific circles since they provide with lots of advantages such as low cost and flexibility. They are three times more inexpensive than silicon ones. According to researchers, they can replace the former technology in inventory identification [112–114]. In Figure 10.14, it can be seen that the unit consists of a transmitter/receiver, rectifier/modulator, and RFID tag (12). The principle of this RFID system follows some simple steps. Primarily, the signal moves toward two directions. On the one side, the signal is sent from transmitter to read a code from the tag, and on the other side, the code read is sent back with the same signal to the receiver.

10.5.6 Organic DNA Sensors In medical electronics field, DNA hybridization sensors with organic transistors is a groundbreaking technique. The purpose of these organic sensors is to detect and quantify the nucleic acids. These kinds of system have much importance since they provide with lots of potential in fields such as pharmacogenomic research, drug discovery, forensic analysis, and molecular diagnosis [115, 116]. Among the highly advanced nucleic acid detection system comes pentacene-based label-free DNA sensors. Furthermore, they are applied in viral, forensic, and genetic symptom examination and detection. Formerly used techniques for sensing DNA molecules were mostly based on expensive, complex optical methods, which have long processing time. On the other hand, this new electronic DNA hybridization technique seems to be the convenient technique since it has better selectivity and sensitivity at on inexpensive way. This new organic material-based DNA sensor has proven its worth for transforming a chemical binding event into electric current. In addition, they can also be measured, analyzed, and amplified readily when compared with conventional techniques [117].

319

320

10 Thermoelectric Films for Electricity Generation

Rectifier/Modulator

CCouple1

TAG A

GND

OTFT2

VRead VRec

Code Generator

OTFT1

Demodulator

CDecouple

CCouple2 Reader

–Vdd

Antenna

Clock generator (ring oseillator)

B Output shift register

Binary counterr (3-bit)

Output

Multiplexer (8*1)

Memory unit (ROM) D7D6D5D4D3D2D1D0Row 0 Line selector (8-bit) (3*8 decoder)

“ “ “ “ “ “

“ “ “ “ “ 7 Row

Figure 10.14 Schematic demonstrations of capacitive coupled organic rectifier/modulator (upper side) and transponder circuit of an organic RFID (lower side).

10.5.7 Limitations When they are compared with organic transistors, silicon-based transistors have higher operating speed, smaller size, environment stability, and durable performance. Therefore, over the last 10 years, they have strong position in the electronic sector. However, the mind started to blow to other direction, since, despite their drawbacks, organic semiconductors have extended their market share. Nevertheless, there is much to be done so that organic semiconductors could become feasible and applicable.

10.6 Techniques of Thin Film Deposition for Thermoelectric Device A lot of thin film deposition techniques are said to be used. However, when deciding whether a technique will be beneficial or not, lots of different parameters must be taken into account such as melting point of material, temperature of deposition, and the bond between the substrate and the material.

10.6.1 Sputtering The main principle of sputtering technique is that an inert gas like Argon is taken to deposit atoms on the substrate by using high energy. To ionize the Argon, RF power

10.6 Techniques of Thin Film Deposition for Thermoelectric Device

Substrate and film growth Sputterin gas

Ar+

Sputtering target

Figure 10.15

The main principle representation of sputtering technique.

sources or direct current is preferred so that energy can be supplied to spread from single element targets or a mix of them. Once the films got deposited at different temperatures ranging from high to average, then they are annealed at high temperatures [118]. In this case, mixed powders from different compositions were taken as sputtering agents to prepare Bi2 Te3 films. As a result, it can be seen that electrical conductivity is enhanced under high-temperature conditions, while Seebeck coefficient plummeted. Another thing that can easily be seen is that annealing temperature of 300 ∘ C is the best condition if the highest power factor is to be targeted [119]. This system is shown in Figure 10.15.

10.6.2 Molecular Beam Epitaxy (MBE) If single crystals are to be deposited, then among the most qualified methods comes the molecular beam epitaxy (MBE). The process begins with separating the single effusion cells from the substrate with shutters or valves, and then evaporation is carried out to extract pure materials while the substrate is heated. The films are grown in epitaxial manner as long as there is a slow deposition rate. Optimization to achieve a certain purity level is maintained by benefitting from high vacuum chambers [21]. In a trial, Seebeck coefficient of 180 μ V/K is maintained by means of this technique to deposit Bi2 Te3 [120]. In another trial, a similar kind of output was maintained by using this technique where Seebeck coefficient is 184 μV/K and power factor is 1.6 mW/(K2 m) while (BiSb)2 Te3 was being deposited [121]. This process is demonstrated in Figure 10.16.

10.6.3 Metal–Organic Chemical Vapor Deposition (MOCVD) MOCVD is another technique where you can produce highly qualified n-type and p-type materials. This technique is maintained with metal hydrides and organic compounds and with the surface reaction between them that have necessary chemical elements. At the surface of the substrate, final pyrolysis of the constituent

321

10 Thermoelectric Films for Electricity Generation

Shutters switch beams on or off

Beams fire at substrate

Substrate

322

Effusion cells

Layers of atoms build up on substrate

Figure 10.16

Substrate heater

A schematic demonstration of molecular beam epitaxy.

chemicals takes place so that the required filming could be formed. The necessary pressure is the optimal rate needed for a successful reaction to occur. In addition, the necessary temperature must be between 300 and 500 ∘ C. However, the situation may change in relation to the reaction. Seebeck coefficient values under this technique’s circumstances are −210 and 100 μ V/K for Bi2 Te3 and Sb2 Te3 films in a row. Besides assessing values of Seebeck coefficient, resistivity values of Bi2 Te3 and Sb2 Te3 are 9 and 3.5 μΩ-m [122]. For maximum Seebeck coefficient, 115 μV/K,

CVD thin film growth

1

Main gas flow region

Gas phase reactions

Desorption of volatile surface reaction products

2

Transport to surface

3

Adsorbtion of film precursor

Figure 10.17 (MOCVD).

4

Surface diffusion

5

Redesorption of filim Nucleation and precursors island growth Step growth

6

Surface reactions

A schematic demonstration of metal–organic chemical vapor deposition

10.6 Techniques of Thin Film Deposition for Thermoelectric Device

Glass cover Anode

Zn foil

Cu foil

Electrolyte

Au contact

Cathode

Au contact Glass substrate

μA

Zn electrodeposited film

Au flash

Controlled current generator

I

+



V

Figure 10.18

A schematic representation of electrochemical deposition.

the deposition of Sb2 Te3 took place under 450 ∘ C. MOCVD system is represented in Figure 10.17.

10.6.4 Electrochemical Deposition (ECD) Electrochemical deposition (ECD) is a relatively cheap and suitable technique of producing thin films in comparison with vacuum-based techniques. Nevertheless, its applications are limited due to low-performance and low-segment materials. The components not only are smelted in nitric acid solution but also perform as chelating agent that helps in avoiding the deposit of irresolvable oxides. Deposition of n-type Bi2 Te3 was performed at a constant potential that is a regular three-electrode configuration [123]. While using ECD technique to deposit Sb2 Te3 films, the power factor was maintained at 0.57 mW/(K2 m) [124]. Electrodeposited Bi2 Te3 films at 50 mV showed a Seebeck coefficient of 51.6 μV/K and the power factor of 0.71 mW/(K2 m). However, electroplated Sb2 Te3 films at 20 mV showed a Seebeck coefficient of 52.1 μV/K and power factor of 0.17 mW/(K2 m) [125]. A schematic representation of ECD is shown in Figure 10.18.

10.6.5 Flash Evaporation (FE) When the components have various vapor pressures, it is flash evaporation that is preferred while depositing thin film alloys. The evaporation of the least volatile constituent of the alloy is maintained with a boat that has adequate pressure and temperatures. Crucial vapor pressure and temperature of components are not of particular importance, which is a very beneficial point of this technique. Rather than using sophisticated mechanical utensils or luxurious cleaning procedures or being contaminant, this technique is performed with simple evaporation materials and precursors [126]. Figure of merit ZT = 0.21 × 10−4 was yielded using this technique while evaporating (Bi2 Te3 )0.9 (BiSe3 )0.1 for n-type material powder and (Bi2 Te3 )0.25 (Sb2 Te3 )0.75 for p-type material powder. A schematic representation of flash evaporation is exhibited in Figure 10.19.

323

324

10 Thermoelectric Films for Electricity Generation

Powder vessel Substrate Glass chamber Guide Tungsten boat

Vacuum

Figure 10.19

A schematic representation of flash evaporation.

10.6.6 Thermal Evaporation Thermal evaporation is the process of heating solid materials inside a highly powerful vacuum chamber to the point where they can produce vapor pressure. In thermal evaporation, thermal energy is used to vapor the material in contrast to MBE kind of techniques with electron beam. While using evaporation technique, Seebeck coefficient and electrical conductivity of p-type Sb2 Te3 thin film were equal to 160 μV/K, 3.12 × 10−3 Ω cm and of n-type Bi2 Te3 thin film were equal to 200 μV/K, 1.29 × 10−3 Ω cm [127]. With the application of this technique, 4.9 mW/(K2 m) for Bi2 Te3 and 2.8 mW/(K−2 m) for Sb2 Te3 high power factors have been maintained [128]. A schematic representation of thermal evaporation is depicted Figure 10.20.

10.6.7 Pulsed Laser Deposition (PLD) While using this technique, the deposition of thin films is maintained with ablation of a couple of targets. A focused pulsed laser beam is necessary for this ablation. Vaporization of the materials from the target is maintained, and then deposition process of thin films under highly powerful vacuum conditions in the presence of gas took place [129, 130]. A schematic diagram of a typical pulsed laser deposition (PLD) setup is shown in Figure 10.21. The process starts with the striking of targets with a 45∘ angle by pulsed and focused laser beam under highly powerful vacuum conditions. After that, the ablation of the atoms and ions took place that subsequently are deposited. In the face of

10.6 Techniques of Thin Film Deposition for Thermoelectric Device

Substrate holder Substrate Deposited thin film Vaporized material

Evaporator

Target material

Chamber

Heater

Figure 10.20

A schematic representation of thermal evaporation.

Substrate Pl a

m

as

Ta rg et

e

um

Laser pulse

pl

Figure 10.21 A schematic demonstration of a typical pulsed laser deposition (PLD) method.

UHV chamber

an inert gas such as Argon or a gas like oxygen, deposition is performed, as long as oxide is to be formed. In general, attachment of substrates is in line with the target surface with a distance of 2–10 cm. In 1996, Dauscher was the first man who attempted to deposit Bi2 Te3 film with PLD [131]. With a thickness of 60 nm, laser energy ranging from 300 to 680 mj, laser intensity ranging from 2 to 10 J cm2 , and substrate temperature ranging from 20 to 500 ∘ C, thin films were deposited [132]. With various temperatures changing from 30 to 400 ∘ C in a deposition chamber that permits inert gas pressure to be taken under control between 100 Pa to atmospheric pressure, on silicon and mica substrates, Bi2 Te3 film raising was made possible [133]. At a substrate temperature ranging from 300 to 500 ∘ C with the usage of pulsed laser, on the surface of mica and aluminum nitride, the depositions of thin films of p-type Bi0.5 Sb1.5 Te3 , n-type Bi2 Te2.7 Se0.3 , and n-type(Bi2 Te3 )90 (Sb2 Te3 )5 (Sb2 Se3 )5 were maintained [134]. Organic transistors have wide range of application fields such as organic inventors, ring oscillators, analog circuits, solar cells, and sensors [135–139]. Their main usage

325

326

10 Thermoelectric Films for Electricity Generation

field is backplane driver in organic display circuits. In addition, they can also be used in organic radio frequency systems as rectifier/modulator unit.

10.7 Conclusion and Future Trends Thermoelectric devices have a wide range of personal electronic applications such as wearable devices (watches or those that can be attached to clothes) and biomedical materials (drug and equipment used to monitor vital signs). This up-to-date technology can provide us benefits. However, their development depends on cost-effective mass production techniques and performance optimization. A good thermoelectric material must have a high electrical conductivity and low thermal conductivity [140]. If electrical conductivity is to be increased, then what must be done is to enhance carrier concentration and/or improve carrier mobility. Enhancing material structure or fabrication processes improves mobility or concentration. Using the superlattice, quantum well structure enhances not only mobility but also Seebeck effect [141]. Quantum confinement is the key feature to enrich figure of merit; consequently nanowires and dots can be incorporated into the scope. Besides the technique mentioned above, using Si ions, various doses over a multilayer thermoelectric material are among the most beneficial ways of carrier concentration enrichment, thus, reducing thermal conductivity when considering quantum structures. In addition, ion usage is said to boost up electronic intensity of quantum structures, thus supporting electrical conductivity and Seebeck coefficient. With enhanced conductivity comes a certain default high thermal conductivity. As far as Seebeck effect is concerned, high thermal conductivity is something faulty. As for the solution, the one that makes sense is to separate them. That is, conductivities must be distinct from each other. Phonon glass electron crystal (PCEC) is a structure; among the processes that can be carried out, there are alloying and site substitution. However, it should not be forgotten that it can only be done with isoelectric elements such as Bi2 Te3 with Sb2 Te3 or Bi2 Se3 . As a result of this process, p-type Bix Sb1−x Te3 and n- Bi2 Sey Te1−y are formed. The interesting thing is that disorder can be introduced into a complex lattice structure as a different technique [7]. As was mentioned above, PCEG material is a perfect material, and it can be created by adding defected atoms in the empty space of the material structure. It does not have any impact on conductivity because the disorder is situated in the empty space and the transmittance occurs outside the empty space. Another beneficial side of it is that defected atoms suppress thermal conductivity. Decoupling process of both conductivities by spin Seebeck effect (SSE) is another way. The main working principle of this effect is that current can be obtained by converting thermally generated spin. You benefit from spin–orbit interaction [142]. Low thermal ferromagnetic materials with highly conductive properties are preferable when generating charges. In Figure 10.22, images respectively show the following: ●

Thermoelectric device with traditional Seebeck effect.

List of Abbreviations and Symbols Longitudinal spin Seebeck effect LSSE) Metalic film Conductor

Ese

∇T

● ●

Flexible device

Z

∇T

Magnetic M Y insulator X Enlarging area

Flexible substrate

Metalic film a1 – Ferromagnetic device ∇T

∇T

Figure 10.22



Coating method

Iy

Serial cnnection

Thermocuple



Developing LSSE devices

Ix

i3

Module structure

Fundamental element

Seebeck effect

Magnetic insulator

Multilayer stack

Ferromagnetic metal

An example of device structure for longitudinal spin Seebeck effect (LSSE).

Thermoelectric device with a module structure with Seebeck effect. Thermoelectric device with longitudinal spin Seebeck effect (LSSE). Thermoelectric device with a module structure LSSE. LSSE device structure.

The benefits of topological insulators have been proven useful when it comes to thermoelectric devices [143]. A mixture of Bi2 Te3 and Sb2 T3 is a perfect insulator. This material makes achieving a figure of merit ZT up to 1.86 feasible and possible, and that is much above traditional thermoelectric device [144]. In United States today, automobile manufacturers are considering using thermoelectric devices. Well-known firms such as Ford and GM are being funded by the Department of Energy to manufacture systems for heating, ventilation, and air-conditioning (HVAC) [145]. Their production leads to a wider range of industrial and electronic applications of thermoelectric devices. To sum up, thermoelectric devices for electricity generation have proven their worth not only in personal applications but also in industrial electronics. Nevertheless, there is much way to go further until they are more efficient and mass produced.

List of Abbreviations and Symbols BTC CSA DMSO ECD FE MBE MOCVD OLED OTFT

Benzene tricarboxylate Camphorsulfonic acid Dimethylsulfoxide Electrochemical deposition Flash evaporation Molecular beam epitaxy Metal–organic chemical vapor deposition Organic light-emitting diode Organic thin film transistor

327

328

10 Thermoelectric Films for Electricity Generation

PCEC PDMS PEDOT: PSS PLD RFID TE ZT 𝜋 𝛼 𝜆 𝜌 𝜎 k Q

Phonon glass electron crystal Polydimethylsiloxane Polystyrene sulfonate Pulsed laser deposition Radio frequency identification Thermoelectric materials Figure of merit Peltier coefficient Seebeck coefficient Thermal conductivity Electrical resistivity Electrical conductivity Thermal conductivity Thomson heat

References 1 Seebeck, T. (1822). Magnetische Polarisation Dermetalle und erze durch Temperature-Differenz. Berlin: Abband Deut. Akad. Wiss. 2 Peltier, J.-C. (1834). Nouvelles expériences sur la caloricité des courants électrique. Annales de Chimie Physique 56 (371): 371–386. 3 Thomson, W. (1857). 4. On a mechanical theory of thermo-electric currents. Proceedings of the Royal Society of Edinburgh 3: 91–98. 4 Rayleigh, L. (1885). XLIII. On the thermodynamic efficiency of the thermopile. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science 20 (125): 361–363. 5 Altenkirch, E. (1909). Über den nutzeffekt der thermosäule. Physikalische Zeitschrift 10 (560): 12. 6 Altenkirch, E. (1911). Elektrothermische Kälteerzeugung und reversible elektrische Heizung. Physikalische Zeitschrift 12: 920–924. 7 Snyder, G.J. and Toberer, E.S. (2011). Complex thermoelectric materials. In: Materials for Sustainable Energy: A Collection of Peer-Reviewed Research and Review Articles from Nature Publishing Group (ed. V. Dusastre), 101–110. World Scientific. 8 Zhao, L.-D., Dravid, V.P., and Kanatzidis, M.G. (2014). The panoscopic approach to high performance thermoelectrics. Energy & Environmental Science 7 (1): 251–268. 9 Zhang, X. and Zhao, L.-D. (2015). Thermoelectric materials: energy conversion between heat and electricity. Journal of Materiomics 1 (2): 92–105. 10 Pei, Y., Shi, X., LaLonde, A. et al. (2011). Convergence of electronic bands for high performance bulk thermoelectrics. Nature 473 (7345): 66. 11 Heremans, J.P., Dresselhaus, M.S., Bell, L.E., and Morelli, D.T. (2013). When thermoelectrics reached the nanoscale. Nature Nanotechnology 8 (7): 471–473.

References

12 Hicks, L. and Dresselhaus, M.S. (1993). Effect of quantum-well structures on the thermoelectric figure of merit. Physical Review B 47 (19): 12727. 13 Hicks, L. and Dresselhaus, M.S. (1993). Thermoelectric figure of merit of a one-dimensional conductor. Physical Review B 47 (24): 16631. 14 Hsu, K.F., Loo, S., Guo, F. et al. (2004). Cubic AgPbm SbTe2+m: bulk thermoelectric materials with high figure of merit. Science 303 (5659): 818–821. 15 Biswas, K., He, J., Blum, I.D. et al. (2012). High-performance bulk thermoelectrics with all-scale hierarchical architectures. Nature 489 (7416): 414. 16 Liu, H., Yuan, X., Lu, P. et al. (2013). Ultrahigh thermoelectric performance by electron and phonon critical scattering in Cu2 Se1–x Ix . Advanced Materials 25 (45): 6607–6612. 17 Zhao, L.-D., Lo, S.H., Zhang, Y. et al. (2014). Ultralow thermal conductivity and high thermoelectric figure of merit in SnSe crystals. Nature 508 (7496): 373. 18 Zhou, W., Fan, Q., Zhang, Q. et al. (2017). High-performance and compact-designed flexible thermoelectric modules enabled by a reticulate carbon nanotube architecture. Nature Communications 8: 14886. 19 Yang, C., Souchay, D., Kneiß, M. et al. (2017). Transparent flexible thermoelectric material based on non-toxic earth-abundant p-type copper iodide thin film. Nature Communications 8: 16076. 20 Sturm, C., Jafarzadeh, P., and Kleinke, H. (2019). 1.17 Thermoelectric nanomaterials. Comprehensive Nanoscience and Nanotechnology: 349. 21 Abdel-Motaleb, I.M. and Qadri, S.M. (2017). Thermoelectric devices: principles and future trends. arXiv preprint arXiv:1704.07742. 22 Sirusi, A.A. and Ross, J.H. Jr., (2017). Recent NMR studies of thermoelectric materials. In: Annual Reports on NMR Spectroscopy (ed. G.A. Webb), 137–198. Elsevier. 23 Riffat, S.B. and Ma, X. (2003). Thermoelectrics: a review of present and potential applications. Applied Thermal Engineering 23 (8): 913–935. 24 Cai, K.F., Mueller, E., Drasar, C., and Stiewe, C. (2004). The effect of titanium diboride addition on the thermoelectric properties of β-FeSi2 semiconductors. Solid State Communications 131 (5): 325–329. 25 Zide, J.M.O., Vashaee, D., Bian, Z.X. et al. (2006). Demonstration of electron filtering to increase the Seebeck coefficient in In0.53 Ga0.47 As/In0.53 Ga0.28 Al0.19 As superlattices. Physical Review B 74 (20): 205335. 26 Bian, Z. and Shakouri, A. (2006). Enhanced solid-state thermionic emission in nonplanar heterostructures. Applied Physics Letters 88 (1): 012102. 27 Snyder, G.J. and Toberer, E.S. (2008). Complex thermoelectric materials. Nature Materials 7: 105. 28 Zheng, J.-c. (2008). Recent advances on thermoelectric materials. Frontiers of Physics in China 3 (3): 269–279. 29 Hochbaum, A.I., Chen, R., Delgado, R.D. et al. (2008). Enhanced thermoelectric performance of rough silicon nanowires. Nature 451: 163. 30 Venkatasubramanian, R., Siivola, E., Colpitts, T., and O’quinn, B. (2001). Thin-film thermoelectric devices with high room-temperature figures of merit. Nature 413 (6856): 597–602.

329

330

10 Thermoelectric Films for Electricity Generation

31 Harman, T.C., Taylor, P.J., Walsh, M.P., and LaForge, B.E. (2002). Quantum dot superlattice thermoelectric materials and devices. Science 297 (5590): 2229–2232. 32 Gao, X., Uehara, K., Klug, D.D., and John, S.T. (2006). Rational design of high-efficiency thermoelectric materials with low band gap conductive polymers. Computational Materials Science 36 (1): 49–53. 33 Sootsman, J.R., Chung, D.Y., and Kanatzidis, M.G. (2009). New and old concepts in thermoelectric materials. Angewandte Chemie International Edition 48 (46): 8616–8639. 34 Nolas, G.S., Morelli, D.T., and Tritt, T.M. (1999). SKUTTERUDITES: a phonon-glass-electron crystal approach to advanced thermoelectric energy conversion applications. Annual Review of Materials Science 29 (1): 89–116. 35 Tritt, T.M. (2002). Thermoelectric Materials: Principles, Structure, Properties, and Applications (eds. K.H.J. Buschow, R.W. Cahn, M.C. Flemings, et al.). Elsevier Ltd. ISBN: 978-0-08-043152-9. 36 Willfahrt, A. (2014). Screen Printed Thermoelectric Devices. Linköping University Electronic Press. 37 Hofmann, A.I., Kroon, R., and Müller, C. (2019). Doping and processing of organic semiconductors for plastic thermoelectrics. In: Handbook of Organic Materials for Electronic and Photonic Devices (ed. O. Ostroverkhova), 429–449. Elsevier. 38 Rowe, D.M. (1995). CRC Handbook of Thermoelectrics. CRC Press. 39 Levin, E.M., Bud’Ko, S.L., and Schmidt-Rohr, K. (2012). Enhancement of thermopower of TAGS-85 high-performance thermoelectric material by doping with the rare earth Dy. Advanced Functional Materials 22 (13): 2766–2774. 40 Goldsmid, H.J. (2010). Introduction to Thermoelectricity, vol. 121. Springer. 41 Tong, X.C. (2011). Advanced Materials for Thermal Management of Electronic Packaging, vol. 30. Springer Science & Business Media. 42 Takashiri, M., Shirakawa, T., Miyazaki, K., and Tsukamoto, H. (2007). Fabrication and characterization of bismuth–telluride-based alloy thin film thermoelectric generators by flash evaporation method. Sensors and Actuators A: Physical 138 (2): 329–334. 43 Venkatasubramanian, R., Colpitts, T., O’Quinn, B. et al. (1999). Low-temperature organometallic epitaxy and its application to superlattice structures in thermoelectrics. Applied Physics Letters 75 (8): 1104–1106. 44 Lin, Z., Hollar, C., Kang, J.S. et al. (2017). A solution processable high-performance thermoelectric copper selenide thin film. Advanced Materials 29 (21): 1606662. 45 Fan, P., Li, Y.Z., Zheng, Z.H. et al. (2013). Thermoelectric properties optimization of Al-doped ZnO thin films prepared by reactive sputtering Zn–Al alloy target. Applied Surface Science 284: 145–149. 46 Fan, P., Zheng, Z.H., Li, Y.Z. et al. (2015). Low-cost flexible thin film thermoelectric generator on zinc based thermoelectric materials. Applied Physics Letters 106 (7): 073901.

References

47 Ohta, H. (2018). Thermoelectrics based on metal oxide thin films. In: Metal Oxide-Based Thin Film Structures (eds. N. Pryds and V. Esposito), 441–464. Elsevier. 48 Ravichandran, J. (2016). Thermoelectric and thermal transport properties of complex oxide thin films, heterostructures and superlattices. Journal of Materials Research 32 (1): 183–203. 49 Terasaki, I., Sasago, Y., and Uchinokura, K. (1997). Large thermoelectric power in NaCo2 O4 single crystals. Physical Review B 56 (20): R12685–R12687. 50 Masset, A.C., Michel, C., Maignan, A. et al. (2000). Misfit-layered cobaltite with an anisotropic giant magnetoresistance: Ca3 Co4 O9 . Physical Review B 62 (1): 166–175. 51 Paul, B., Björk, E.M., Kumar, A. et al. (2018). Nanoporous Ca3 Co4 O9 thin films for transferable thermoelectrics. ACS Applied Energy Materials 1 (5): 2261–2268. 52 Cushing, B.L., Falster, A.U., Simmons, W.B., and Wiley, J.B. (1996). A multivalent ion exchange route to lamellar calcium cobalt oxides, Cax CoO2 (x⩽ 0.5). Chemical Communications 23: 2635–2636. 53 Liu, J., Huang, X., Xu, G., and Chen, L. (2013). Thermoelectric properties of layered Sr0.29 CoO2 crystals. Journal of Alloys and Compounds 576: 247–249. 54 Liu, J., Huang, X., Li, F. et al. (2011). Low-temperature magnetic and thermoelectric properties of layered Ca0.33 CoO2 crystals. Journal of the Physical Society of Japan 80 (7): 074802. 55 Liu, W., Kim, H.S., Jie, Q., and Ren, Z. (2016). Importance of high power factor in thermoelectric materials for power generation application: a perspective. Scripta Materialia 111: 3–9. 56 Teranishi, T., Mori, Y., Hayashi, H., and Kishimoto, A. (2012). Thermoelectric property of polycrystalline aluminum-doped zinc oxide enhanced by micropore foaming. Journal of the American Ceramic Society 95 (2): 690–695. 57 Ghosh, C., Das, S., and Chattopadhyay, K. (2007). Enhancement of thermopower of Mn doped ZnO thin film. Physica B: Condensed Matter 399 (1): 38–46. 58 Mele, P., Saini, S., Honda, H. et al. (2013). Effect of substrate on thermoelectric properties of Al-doped ZnO thin films. Applied Physics Letters 102 (25): 253903. 59 Cheng, H., Xu, X.J., Hng, H.H., and Ma, J.J.C.I. (2009). Characterization of Al-doped ZnO thermoelectric materials prepared by RF plasma powder processing and hot press sintering. Ceramics International 35 (8): 3067–3072. 60 Cai, K., Müller, E., Drašar, C., and Mrotzek, A. (2003). Preparation and thermoelectric properties of Al-doped ZnO ceramics. Materials Science and Engineering B 104 (1–2): 45–48. 61 Park, K. and Ko, K. (2007). Effect of TiO2 on high-temperature thermoelectric properties of ZnO. Journal of Alloys and Compounds 430 (1–2): 200–204. 62 Park, K., Seong, J., and Nahm, S. (2008). Improvement of thermoelectric properties with the addition of Sb to ZnO. Journal of Alloys and Compounds 455 (1–2): 331–335.

331

332

10 Thermoelectric Films for Electricity Generation

63 Park, K., Seong, J., and Kim, G. (2009). NiO added Zn1−x Nix O (0≤ x≤ 0.05) for thermoelectric power generation. Journal of Alloys and Compounds 473 (1–2): 423–427. 64 Chen, X.Q., Li, Z., and Dou, S.X. (2015). Ambient facile synthesis of gram-scale copper selenide nanostructures from commercial copper and selenium powder. ACS Applied Materials & Interfaces 7 (24): 13295–13302. 65 Chen, X., Li, Z., Yang, J. et al. (2015). Aqueous preparation of surfactant-free copper selenide nanowires. Journal of Colloid and Interface Science 442: 140–146. 66 Chen, X.Q., Bai, Y., Li, Z. et al. (2016). Ambient synthesis of one-/two-dimensional CuAgSe ternary nanotubes as counter electrodes of quantum-dot-sensitized solar cells. ChemPlusChem 81 (4): 414–420. 67 Chen, X.Q., Li, Z., Bai, Y. et al. (2015). Room-temperature synthesis of Cu2−x E (E=S, Se) nanotubes with hierarchical architecture as high-performance counter electrodes of quantum-dot-sensitized solar cells. Chemistry A European Journal 21 (3): 1055–1063. 68 Tan, M. and Chen, X. (2011). Growth mechanism of single crystal nanowires of fcc metals (Ag, Cu, Ni) and hcp metal (Co) electrodeposited. Journal of the Electrochemical Society 159 (1): K15–K20. 69 Nunna, R., Qiu, P., Yin, M. et al. (2017). Ultrahigh thermoelectric performance in Cu2 Se-based hybrid materials with highly dispersed molecular CNTs. Energy & Environmental Science 10 (9): 1928–1935. 70 Lévesque, I., Bertrand, P.O., Blouin, N. et al. (2007). Synthesis and thermoelectric properties of polycarbazole, polyindolocarbazole, and polydiindolocarbazole derivatives. Chemistry of Materials 19 (8): 2128–2138. 71 Kim, G.H., Shao, L., Zhang, K., and Pipe, K.P. (2013). Engineered doping of organic semiconductors for enhanced thermoelectric efficiency. Nature Materials 12: 719. 72 Bubnova, O., Khan, Z.U., Wang, H. et al. (2013). Semi-metallic polymers. Nature Materials 13: 190. ´ ´ 73 Kolasinska, E. and Kolasinski, P. (2016). A review on electroactive polymers for waste heat recovery. Materials 9 (6): 485. 74 Bubnova, O. and Crispin, X. (2012). Towards polymer-based organic thermoelectric generators. Energy & Environmental Science 5 (11): 9345–9362. 75 Shi, K., Zhang, F., Di, C.A. et al. (2015). Toward high performance n-type thermoelectric materials by rational modification of BDPPV backbones. Journal of the American Chemical Society 137 (22): 6979–6982. 76 Dubey, N. and Leclerc, M. (2011). Conducting polymers: efficient thermoelectric materials. Journal of Polymer Science Part B: Polymer Physics 49 (7): 467–475. 77 Zhang, Q., Sun, Y., Xu, W., and Zhu, D. (2014). Organic thermoelectric materials: emerging green energy materials converting heat to electricity directly and efficiently. Advanced Materials 26 (40): 6829–6851. 78 Sun, Y., Qiu, L., Tang, L. et al. (2016). Flexible n-type high-performance thermoelectric thin films of poly(nickel-ethylenetetrathiolate) prepared by an electrochemical method. Advanced Materials 28 (17): 3351–3358.

References

79 Xuan, Y., Liu, X., Desbief, S. et al. (2010). Thermoelectric properties of conducting polymers: the case of poly(3-hexylthiophene). Physical Review B 82 (11): 115454. 80 Zuzok, R., Kaiser, A.B., Pukacki, W., and Roth, S. (1991). Thermoelectric power and conductivity of iodine-doped “new” polyacetylene. The Journal of Chemical Physics 95 (2): 1270–1275. 81 Park, Y.W., Yoon, C.O., Na, B.C. et al. (1991). Metallic properties of transition metal halides doped polyacetylene: the soliton liquid state. Synthetic Metals 41 (1): 27–32. 82 Guo, C., Chu, F., Chen, P. et al. (2018). Effectively enhanced thermopower in polyaniline/Bi0.5 Sb1.5 Te3 nanoplate composites via carrier energy scattering. Journal of Materials Science 53: 6752–6762. 83 Li, J., Tang, X., Li, H. et al. (2010). Synthesis and thermoelectric properties of hydrochloric acid-doped polyaniline. Synthetic Metals 160 (11): 1153–1158. 84 Sun, Y., Wei, Z., Xu, W., and Zhu, D. (2010). A three-in-one improvement in thermoelectric properties of polyaniline brought by nanostructures. Synthetic Metals 160 (21): 2371–2376. 85 Holland, E.R., Pomfret, S.J., Adams, P.N. et al. (1997). Doping dependent transport properties of polyaniline-CSA films. Synthetic Metals 84 (1): 777–778. 86 Liu, J., Wang, X., Li, D. et al. (2015). Thermal conductivity and elastic constants of PEDOT:PSS with high electrical conductivity. Macromolecules 48 (3): 585–591. 87 Luo, J., Billep, D., Waechtler, T. et al. (2013). Enhancement of the thermoelectric properties of PEDOT:PSS thin films by post-treatment. Journal of Materials Chemistry A 1 (26): 7576–7583. 88 Kemp, N.T., Kaiser, A.B., Liu, C.J. et al. (1999). Thermoelectric power and conductivity of different types of polypyrrole. Journal of Polymer Science Part B: Polymer Physics 37: 953–960. 89 Bao-Yang, L., Cong-Cong, L., Shan, L. et al. (2010). Thermoelectric performances of free-standing polythiophene and poly(3-methylthiophene) nanofilms. Chinese Physics Letters 27 (5): 057201. 90 Sun, J., Yeh, M.L., Jung, B.J. et al. (2010). Simultaneous increase in Seebeck coefficient and conductivity in a doped poly(alkylthiophene) blend with defined density of states. Macromolecules 43 (6): 2897–2903. 91 Di, C.-a., Xu, W., and Zhu, D. (2016). Organic thermoelectrics for green energy. National Science Review 3 (3): 269–271. 92 Sun, Y., Zhang, J., Liu, L. et al. (2016). Optimization of the thermoelectric properties of poly(nickel-ethylenetetrathiolate) synthesized via potentiostatic deposition. Science China Chemistry 59 (10): 1323–1329. 93 Babel, A. and Jenekhe, S.A. (2003). High electron mobility in ladder polymer field-effect transistors. Journal of the American Chemical Society 125 (45): 13656–13657. 94 Wang, S., Sun, H., Ail, U. et al. (2016). Thermoelectric properties of solution-processed n-doped ladder-type conducting polymers. Advanced Materials 28 (48): 10764–10771.

333

334

10 Thermoelectric Films for Electricity Generation

95 Russ, B., Robb, M.J., Brunetti, F.G. et al. (2014). Power factor enhancement in solution-processed organic n-type thermoelectrics through molecular design. Advanced Materials 26 (21): 3473–3477. 96 Du, Y., Shen, S.Z., Cai, K., and Casey, P.S. (2012). Research progress on polymer–inorganic thermoelectric nanocomposite materials. Progress in Polymer Science 37 (6): 820–841. 97 Wan, C., Gu, X., Dang, F. et al. (2015). Flexible n-type thermoelectric materials by organic intercalation of layered transition metal dichalcogenide TiS2 . Nature Materials 14: 622. 98 Sun, L., Miyakai, T., Seki, S. et al. (2013). Mn2 (2,5-disulfhydrylbenzene-1,4-dicarboxylate): a microporous metal–organic framework with infinite (−Mn–S−)∞ chains and high intrinsic charge mobility. Journal of the American Chemical Society 135 (22): 8185–8188. 99 Sheberla, D., Sun, L., Blood-Forsythe, M.A. et al. (2014). High electrical conductivity in Ni3 (2,3,6,7,10,11-hexaiminotriphenylene)2 , a semiconducting metal–organic graphene analogue. Journal of the American Chemical Society 136 (25): 8859–8862. 100 Erickson, K.J., Léonard, F., Stavila, V. et al. (2015). Thin film thermoelectric metal–organic framework with high Seebeck coefficient and low thermal conductivity. Advanced Materials 27 (22): 3453–3459. 101 Meng, C., Liu, C., and Fan, S. (2010). A promising approach to enhanced thermoelectric properties using carbon nanotube networks. Advanced Materials 22 (4): 535–539. 102 Luo, W., Shen, D., Zhang, R. et al. (2016). Germanium nanograin decoration on carbon shell: boosting lithium-storage properties of silicon nanoparticles. Advanced Functional Materials 26 (43): 7800–7806. 103 Wang, L., Yao, Q., Xiao, J. et al. (2016). Engineered molecular chain ordering in single-walled carbon nanotubes/polyaniline composite films for high-performance organic thermoelectric materials. Chemistry - An Asian Journal 11 (12): 1804–1810. 104 Cho, C., Stevens, B., Hsu, J.H. et al. (2015). Completely organic multilayer thin film with thermoelectric power factor rivaling inorganic tellurides. Advanced Materials 27 (19): 2996–3001. 105 Nolas, G.S., Sharp, J., and Goldsmid, J. (2013). Thermoelectrics: Basic Principles and New Materials Developments, vol. 45. Springer Science & Business Media. 106 Adroja, N., Mehta, S.B., and Shah, P. (2015). Review of thermoelectricity to improve energy quality. International Journal on Emerging Technologies and Innovative Research 2: 847–850. 107 Dodabalapur, A., Baumbach, J., Baldwin, K., and Katz, H.E. (1996). Hybrid organic/inorganic complementary circuits. Applied Physics Letters 68 (16): 2246–2248. 108 Jackson, T.N., Lin, Y.Y., Gundlach, D.J., and Klauk, H. (1998). Organic thin-film transistors for organic light-emitting flat-panel display backplanes. IEEE Journal of Selected Topics in Quantum Electronics 4 (1): 100–104.

References

109 Zhou, L., Wanga, A., Wu, S.C. et al. (2006). All-organic active matrix flexible display. Applied Physics Letters 88 (8): 083502. 110 Li, D. and Guo, L.J. (2006). Micron-scale organic thin film transistors with conducting polymer electrodes patterned by polymer inking and stamping. Applied Physics Letters 88 (6): 063513. 111 Chu, C.-W., Chen, C.W., Li, S.H. et al. (2005). Integration of organic light-emitting diode and organic transistor via a tandem structure. Applied Physics Letters 86 (25): 253503. 112 Myny, K., Steudel, S., Vicca, P. et al. (2010). Organic RFID tags. In: Radio Frequency Identification Fundamentals and Applications Design Methods and Solutions (ed. C. Turcu), 311–324. IntechOpen. 113 Subramanian, V., Fréchet, J.M., Chang, P.C. et al. (2005). Progress toward development of all-printed RFID tags: materials, processes, and devices. Proceedings of the IEEE 93 (7): 1330–1338. 114 Subramanian, V., Chang, P.C., Lee, J.B. et al. (2005). Printed organic transistors for ultra-low-cost RFID applications. IEEE Transactions on Components and Packaging Technologies 28 (4): 742–747. 115 Patolsky, F., Lichtenstein, A., and Willner, I. (2001). Detection of single-base DNA mutations by enzyme-amplified electronic transduction. Nature Biotechnology 19 (3): 253. 116 Fang, B., Jiao, S., Li, M. et al. (2008). Label-free electrochemical detection of DNA using ferrocene-containing cationic polythiophene and PNA probes on nanogold modified electrodes. Biosensors and Bioelectronics 23 (7): 1175–1179. 117 Roberts, M.E., Sokolov, A.N., and Bao, Z. (2009). Material and device considerations for organic thin-film transistor sensors. Journal of Materials Chemistry 19 (21): 3351–3363. 118 Bottner, H., Nurnus, J., Gavrikov, A. et al. (2004). New thermoelectric components using microsystem technologies. Journal of Microelectromechanical Systems 13 (3): 414–420. 119 Huang, H., Luan, W.-l., and Tu, S.-t. (2009). Influence of annealing on thermoelectric properties of bismuth telluride films grown via radio frequency magnetron sputtering. Thin Solid Films 517 (13): 3731–3734. 120 Charles, E., Groubert, E., and Boyer, A. (1988). Structural and electrical properties of bismuth telluride films grown by the molecular beam technique. Journal of Materials Science Letters 7 (6): 575–577. 121 Cho, S., Kim, Y., and Ketterson, J.B. (2004). Structural and thermoelectric properties in (Sb1–x Bix )2 Te3 thin films. Applied Physics A 79 (7): 1729–1731. 122 Giani, A., Boulouz, A., Pascal-Delannoy, F. et al. (1999). Growth of Bi2 Te3 and Sb2 Te3 thin films by MOCVD. Materials Science and Engineering B 64 (1): 19–24. 123 Lim, J., Snyder, G. J., Huang, C. K. et al. (2002). Thermoelectric microdevice fabrication process and evaluations at the Jet Propulsion Laboratory (JPL). 21st International Conference on Thermoelectrics. 124 Lim, S.-K., Kim, M.-Y., and Oh, T.-S. (2009). Thermoelectric properties of the bismuth–antimony–telluride and the antimony–telluride films processed by

335

336

10 Thermoelectric Films for Electricity Generation

125

126

127

128

129

130

131 132 133 134

135 136

137 138

139 140

electrodeposition for micro-device applications. Thin Solid Films 517 (14): 4199–4203. Kim, M.-Y. and Oh, T.-S. (2009). Electrodeposition and thermoelectric characteristics of Bi2 Te3 and Sb2 Te3 films for thermopile sensor applications. Journal of Electronic Materials 38 (7): 1176–1181. Hemanadhan, M., Bapanayya, C., and Agarwal, S.C. (2010). Simple flash evaporator for making thin films of compounds. Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films 28 (4): 625–626. Zou, H., Rowe, D.M., and Williams, S.G.K. (2002). Peltier effect in a co-evaporated Sb2 Te3 (P)-Bi2 Te3 (N) thin film thermocouple. Thin Solid Films 408 (1–2): 270–274. Carmo, J.P., Goncalves, L.M., Wolffenbuttel, R.F., and Correia, J.H. (2010). A planar thermoelectric power generator for integration in wearable microsystems. Sensors and Actuators A: Physical 161 (1–2): 199–204. Krebs, H.-U., Weisheit, M., Faupel, J. et al. (2003). Pulsed laser deposition (PLD)--a versatile thin film technique. In: Advances in Solid State Physics (ed. B. Kramer), 505–518. Springer. Geohegan, D.B., Chrisey, D.B., and Hubler, G.K. (1994). Pulsed laser deposition of thin films. In: Pulsed Laser Deposition of Thin-films (eds. D.B. Chrisey and G.K. Hubler), 59–69. New York: Wiley. Dauscher, A., Thomy, A., and Scherrer, H. (1996). Pulsed laser deposition of Bi2 Te3 thin films. Thin Solid Films 280 (1–2): 61–66. Zeipl, R., Pavelka, M., Jelínek, M. et al. (2003). Some properties of very thin Bi2 Te3 layers prepared by laser ablation. Physica Status Solidi C 3: 867–871. Bailini, A., Donati, F., Zamboni, M. et al. (2007). Pulsed laser deposition of Bi2 Te3 thermoelectric films. Applied Surface Science 254 (4): 1249–1254. Makala, R.S., Jagannadham, K., and Sales, B.C. (2003). Pulsed laser deposition of Bi2 Te3 -based thermoelectric thin films. Journal of Applied Physics 94 (6): 3907–3918. Mizukami, M., Hirohata, N., Iseki, T. et al. (2006). Flexible AM OLED panel driven by bottom-contact OTFTs. IEEE Electron Device Letters 27 (4): 249–251. Cantatore, E., Geuns, T.C., Gelinck, G.H. et al. (2006). A 13.56-MHz RFID system based on organic transponders. IEEE Journal of Solid-State Circuits 42 (1): 84–92. Briand, D., Oprea, A., Courbat, J., and Bârsan, N. (2011). Making environmental sensors on plastic foil. Materials Today 14 (9): 416–423. Takamiya, M., Sekitani, T., Kato, Y. et al. (2006). An organic FET SRAM with back gate to increase static noise margin and its application to braille sheet display. IEEE Journal of Solid-State Circuits 42 (1): 93–100. Tobjörk, D. and Österbacka, R. (2011). Paper electronics. Advanced Materials 23 (17): 1935–1961. Slack, G.A. and Materials, N. (1995). Performance limits for thermoelectric cooling. In: CRC Handbook of Thermoelectrics (ed. D.M. Rowe), 407–440. CRC Press.

References

141 Ismail, Y. and Alaskalany, A. (2014). Thermoelectric devices: cooling and power generation. arXiv preprint arXiv:1403.3836. 142 Uchida, K.-i., Adachi, H., Kikkawa, T. et al. (2016). Thermoelectric generation based on spin Seebeck effects. Proceedings of the IEEE 104 (10): 1946–1973. 143 Zhang, J., Feng, X., Xu, Y. et al. (2015). Disentangling the magnetoelectric and thermoelectric transport in topological insulator thin films. Physical Review B 91 (7): 075431. 144 Kim, S.I., Lee, K.H., Mun, H.A. et al. (2015). Dense dislocation arrays embedded in grain boundaries for high-performance bulk thermoelectrics. Science 348 (6230): 109–114. 145 Gonzalez, R. and Maronde, C. (2010). Automotive thermoelectric HVAC development and demonstration project. California Energy Commission 10, 10-10.

337

339

11 Template-assisted Fabrication of Nanostructure Thin Films for Ultrasensitive Detection of Chemicals and Biomolecules: Part A – Template-assisted Nanoimprinting Technology for Functional Thin Films Xiaomin Zhu 1 , Xinhua Chen 2,5 , Andrey A. Voronov 3,4 , Vladimir I. Belotelov 3,4 , and Yujun Song 1,5 1 University of Science and Technology, Center for Modern Physics Technology, Applied Physics Department, School of Mathematics and Physics, 30 Xueyuan Road, Beijing 100083, China 2 Zhejiang University, Key Laboratory of Combined Multi-organ Transplantation, Ministry of Public Health, Department of Hepatobiliary and Pancreatic Surgery, the First Affiliated Hospital, Hangzhou, Zhejiang, 310003, China 3 Lomonosov Moscow State University, Faculty of Physics, Leninskie Gory, Moscow 119991, Russia 4 Russian Quantum Center, 45, Skolkovskoye shosse, Moscow, 121353, Russia 5 Zhejiang Key Laboratory for Pulsed Power Technology Translational Medicine, Hangzhou Ruidi Biotechnology Company, Hangzhou 310000, China

11.1 Development of Template-assisted Nanoimprinting Technology Template-assisted (TA) lithography (LIGA) has developed as a powerful physical technique that enables the production of surface morphology-confined nanoparticles (NPs) and NP arrays with controlled shapes, sizes, and interparticle spacing. Several templates have been developed for these purposes, such as porous polymers, porous Al2 O3 foils, and nanosphere (NS) arrays (polymers and ceramics), resulting in various TA LIGA, correspondingly as porous polymers lithography (PP-LIGA), porous anodic aluminum oxide (AAO) LIGA, or nanosphere lithography (NSL) [1]. Our team has been working on the development and research of TA LIGA for the preparation of nanostructured thin films for many years and has made outstanding progress in NSL and porous anodized alumina TA fabrication processes. We have invented the methodology of single nanomaterials identification via multi-hierarchy arrayed microwindows (MHMW) assisted NSL and the AAO template-assisted in situ sol–gel transfer nanoimprinting method, which can obtain nanostructured array thin films on hard or soft substrates. This chapter focuses on three TA LIGA methods developed by our team. Combined with AAO technology, different nanoarrays, nanowires (NWs), and nanorods (NRs) were fabricated. Additionally, their application for detecting ultrasensitive chemical and biological molecules is also summarized.

Inorganic and Organic Thin Films: Fundamentals, Fabrication, and Applications, First Edition. Edited by Yujun Song. © 2021 WILEY-VCH GmbH. Published 2021 by WILEY-VCH GmbH.

340

11 Template-assisted Fabrication of Nanostructure Thin Films

11.2 Nanosphere Lithography (NSL) This section presents the recent progress in NSL for the controlled producing noble metal nanomaterials, including the size, shape, and surface morphology-controlled fabrication of noble metal NPs and nanoarrays. Then four distinct progresses in the development of NSL techniques are also showed: (i) fabrication of hierarchically ordered NW arrays on substrates by combination of NSL and porous AAO; (ii) identification of single NPs and nano-arrays by combining NSL and MHMW; (iii) fabrication of biosensing system based on the combination of the noble metal NPs and nanoarrays fabricated by NSL and microfluidic techniques; and (iv) synthesis of solution-phased NPs by transferring the NPs into solutions. In (ii) and (iii), the related 3D morphologies and arrangement dependent optical properties, and comparison between the numerical as well as experimental results are provided, revealing their intrinsic quantum mechanism. The research are fundamental pre-conditions for the discovery of novel properties and applications of noble metal NPs. Finally, issues and perspectives in the controlled fabrication of noble metal nanomaterials by NSL, and investigation of their 3D morphologies and arrangement dependent optical properties for future potential applications are also highlighted and discussed.

11.2.1 Size and Shape Controlled Fabrication of Nanomaterials via NSL NSs have been used to form uniformly arranged layers as templates to produce perfect triangle nanoprisms on substrates. The routine procedure for the production of triangular shaped nanoprisms, based on the NSL, is described in Figure 11.1 (a: cross-section view; b: top view). The hexagonal arranged NS monolayer is formed on the substrate by a coating process (e.g. dip-coating, rotating-coating, or spinning-coating) (Step 1a). The interstitial among any three adjacent NSs form triangular voids (Step 1b) as templates. The desired noble metal (e.g. Ag) is then deposited on the triangular interstitial among the NSs to form triangular Ag NPs (Step 2a, b). After NSs are released by sonication or other methods, surface-confined triangular Ag nanoprisms can be obtained (Step 3a). By this NSL process, uniform hexagonal-arrayed triangle nanoprisms can be fabricated on a variety of substrates (e.g. glass, mica, silica wafer, polymethyl methacrylate (PMMA), etc.). Step 3b is an atomic force microscope (AFM) image of Ag triangular nanoprisms fabricated by our group using a self-assembled monolayer of 300 nm polystyrene NSs as the template. The initial critical step in NSL is the formation of a uniform large-scale NS template. Both drop-coating and spin-coating can produce uniform templates on glass, silica wafer, or mica substrate. The uniformity of the NS template produced by drop-coating depends on the NS type and concentration, the hydrophilic properties of the substrate, the environmental humidity, temperature, and the drying speed. A monolayer colloidal polystyrene NS mask can be prepared by drop-coating of 3.0–4.0 μL, 2–15 times diluted NS solution (concentration 4.0 wt%) onto the glass

11.2 Nanosphere Lithography (NSL)

Step 1↵

Nanosphere monolayer

Triangular interstitials

(a) (b) Step 2↵

Triangle nanoprisms formed in the interstitial of nanospheres

(a)

(b)

Step 3↵ Triangle nanoprisms

(a)

(b)

100 nm

Figure 11.1 The NSL process for triangular NPs fabrication. Step 1a: The hexagonal arranged nanosphere mono layer is formed on the substrate by coating process. Step 1b: The interstitial among any three adjacent nanospheres form triangle shaped voids as templates. Step 2a,b: The Ag metal is deposited on the triangular interstitials among the nanospheres to form triangle shaped Ag NPs; Step 3a: The nanospheres is released by sonication or other methods, leaving the triangular Ag nanoprisms on the substrates, by this nanosphere LIGA process, the hexagonal arrayed uniform triangular nanoprisms can be fabricated on a variety of substrates (e.g. glass, mica, silica wafer, PMMA, etc.). Step 3b: The AFM image for Ag triangle nanoprisms fabricated by monolayer template from 290 nm polystyrene nanospheres in our group; these nanoprisms have very uniform edge length of 67 ± 4 nm (STDEV% of 6%) and thickness of 20.0 ± 1.0 nm (STDEV% of 5%). (a) Crosssectional view; (b) top 3D view. Source: Adapted in part from Song, Y. China Patent, CN200910085973.9; Haynes, C. L.; van Duyne, R. P., J. Phys. Chem. B 2001 105, 5599, Copyright [1999] American Chemical Society; (b) Song [1], doi: 10.5772/25037.

support and leaving them to dry overnight. A detailed procedure to fabricate the NS mask using drop-coating is described. The glass is cleaned by sonication with a mixture of sulfuric acid and hydrogen peroxide (3 : 1 = concentration H2 SO4 : 30% H2 O2 , volume ratio) at 80 ∘ C for 30 minutes and washed using sufficient nanopure water. Then, the glass is sonicated in a mixture of ammonia and hydrogen peroxide (5 : 1 : 1 = H2 O:NH4 OH (37%):30% H2 O2 , volume ratio) to increase the hydrophilic property on the surface. Finally, the glass is washed using sufficient nanopure water again and then stored in the nanopure water for future use. When drop-coating is to be performed, the glass is picked up from the nanopure water from one of its edges. The remaining water droplets on the glass substrate are removed by drying

341

342

11 Template-assisted Fabrication of Nanostructure Thin Films

0.5 μm

Figure 11.2 Nanosphere templates based on 290 nm spherical polystyrene nanospheres for Ag nanoparticle fabrication. Source: Reprinted from Song and Elsayed-Ali [1], Copyright (2010) Elsevier.

5 μm

on the opposite edge with filter paper. The substrate is then left flat in a clean Petri dish with a tilt angle of 3∘ –5∘ , 15 μL of PS NS solution is added on the surface of the glass substrate using a droplet. The water spreads over the whole glass to form a semi-ellipsoidal shaped water spot. The Petri-dish is left long enough for full evaporation. During evaporation, the temperature is kept at 18 ± 3 ∘ C and the humidity is kept 50 ± 5%. In our group, a near-uniform monolayer NS template can be prepared on almost the whole glass (18 mm diameter). Figure 11.2 shows one typical area of a near-uniform monolayer template over scale of 20 μm. From the magnified image, a selected area shown in the inset, no lattice defects can be observed. Using this template, uniform Ag nanoprisms can be fabricated by vapor deposition process. One typical area fabricated by our group is shown in Step 3b in Figure 11.1, where these nanoprisms have very uniform edge length of 67 ± 4 nm (STDEV% of 6%) and thickness of 20.0 ± 1.0 nm (STDEV% of 5%). We recently developed a modified NSL process to fabricate Ag NPs with controlled shapes on substrates. The modification in NSL is performed by thermally annealing the triangular nanoprisms, and sonication to remove weak tips, followed by removing debris and small broken parts around the NPs on the substrates. The detailed process is shown as the following: (i) releasing the NSs by immersing the cover slip into a 5% HCl solution for 30 minutes, then immersing the glass substrates into CH2 Cl2 for 30 seconds, and then sonicating for 20–60 seconds; (ii) The fabricated Ag nanoprisms on the glass substrates are annealed at 100–300 ∘ C for 1–5 hours; (iii) Then Ag nanoprisms are cleaned by immersing the glass cover slip into 5% HNO3 for 10–20 seconds to remove any surface contamination and dissolve debris around the NPs, and then washed by large amount of nanopure water. Comparing the AFM images in Step 3b in Figure 11.1 showing the NPs without above posttreatment, tip-round triangle nanoprisms, square-shaped, and trapezoidal Ag NPs (Figure 11.3) can be obtained via one or two of the above treatment. Thermal annealing results in much more uniform NP surfaces without the thin, weak tips, and edges (Figure 11.3a). From the magnified AFM plane image in the inset of Figure 11.3a and

11.2 Nanosphere Lithography (NSL) 50.0 nm

100.0 nm

50.0 nm

25.0 nm

50.0 nm

25.0 nm

0.0 nm 0.0 nm

0.0 nm

200 nm (a)

(c)

0.2

200 nm

200 nm (e)

0.2 0.4

0.2 0.4

0.6

0.4 0.6

0.8

0.6 0.8

1.0

0.8 1.0

(b)

(d)

(f)

1.0

Figure 11.3 Surface-confined Ag NPs with controlled shapes fabricated by the modified NSL process. (a) AFM image of triangular prism Ag NPs with round tips after thermal annealing at 200 ∘ C for 4 hours, cleaning by 5% nitric acid, and washing by nanopure water. (b) The 3D image of the triangular prism Ag NPs with round tips. (c) Flat trapezoidal Ag NPs after sonication to remove one tip, thermal annealing, cleaning by 5% nitric acid, and washing by nanopure water. (d) The 3D image of the trapezoidal Ag NPs with one snipped tip. (e) The quadrilateral or pentagon shaped Ag NPs after sonication intensively to remove two tips, thermal annealing, cleaning by 5% nitric acid and washing by nanopure water. Dashed circles: pentagonal Ag NPs with one sharp tip left; dashed squares: quadrilateral Ag NPs. (f) is the 3D image of the quadrilateral and pentagon shaped Ag NPs. Source: Song and Elsayed-Ali [7]. © 2010, Elsevier.

the 3D image in Figure 11.3b, the NPs still show triangular prism shape with round edges and little surface defects. Alternatively, if the NPs produced by NSL is sonicated for 30–45 seconds to remove a weak tip, anneal them at 200 ∘ C for 1–4 hours, then wash them with 5% nitric acid, the trapezoidal shaped NPs with round edges will be formed, as shown in Figure 11.3c,d. If the sonication time is increased to more than two minutes, the NPs will lose their two sharp tips and form quadrilateral or pentagon shaped NPs. After thermal annealing for 1–4 hours and washing with 5% nitric acid, their edges and corners become round, as shown in Figure 11.3e, which show quadrilateral NPs (in dashed squares) or pentagon (in dashed circles). The 3D AFM image in Figure 11.3f, shows NPs have round edges and corners. Clearly, even after thermal annealing, they are still showing prism shapes with increased thickness from their edges to centers according to their 3D AFM images. The work described above demonstrates that NSL, broadly defined to include angle-resolved nanosphere lithography (AR NSL) and some modified post treatment after deposition of the desired materials, is manifestly capable of creating far more than arrays of nanotriangles, nanodots as previously supposed. The progresses in NSL endow much potential in the size- and shape-controlled fabrication of NPs and nanoarrays, which gives NSL potential since the ability of NSL to synthesize

343

344

11 Template-assisted Fabrication of Nanostructure Thin Films

monodisperse, size- and shape-tunable NPs can be exploited to precisely investigate the size- and shape-dependent physiochemical properties of nanooptics and nanoarrays.

11.2.2 Multi-hierarchy Micro Windows (MHMW) for Single Nanostructure and/or Array Identification The physicochemical properties of nanomaterials significantly depend on their 3D morphologies (sizes, shapes, and surface topography), their surrounding media and spatial arrangement. Systematically and precisely correlating these parameters with the related physicochemical properties of specific single NPs or nanoarrays is a fundamental requirement for the discovery of their novel properties and applications, as well as for advancing the fundamental and practical knowledge required for the design and fabrication of new materials. The lack of effective means of fabricating recognizable 3D morphology-controlled NPs and nanoarrays and correlating their structure parameters with their physicochemical properties has been observed in different characterization techniques. It is an obstacle for studying the 3D morphology-dependent properties of individual NPs and nanoarrays. Current studies investigate the physicochemical properties of the NP ensemble, but not of a single NP. Effective methods for 3D morphology-controlled fabrication of nanomaterials, in order to correlate their 3D morphology of a single NPs or nanoarrays with their physicochemical properties, are also essential to address fundamental and practical questions related to the single NPs. An important research area in nanoscale plasmonic optics is single NP identification and characterization of their 3D morphologies and space orientation-dependent physicochemical properties. Recently, much attention has been given to the localized surface plasmon resonance (LSPR) of metal NPs because of their promising applications in plasmonic circuits, optoelectronic transducers, optical bioprobes, and surface plasmon resonance (SPR) interference LIGA. Since the plasmonic properties of metal NPs intrinsically rely on their size, shape, surface topography, crystal structure, interparticle spacing, and the dielectric environment around them. Methods to correlate their plasmonic properties with the above structural and environmental parameters have become one of the most rapidly developing research directions. In the precise investigation of the relationship between the LSPR properties and their 3D morphologies of specific NPs and nanoarrays, two methods have been developed recently: the in situ method and the spatial localization method [2]. The in situ method combines at least two different instruments together to conduct the structure and property characterization simultaneously: one can be used to characterize the 3D morphology (e.g. AFM or STEM) [3] of NPs and the other to characterize the LSPR-related optical properties of the same NPs (e.g. dark-field microscope and spectroscopy). The spatial localization method requires using markers to recognize the same single NP in different instruments. We have also developed one spatial localization method to precisely investigate the 3D morphology dependent LSPR properties of specific NPs and nanoarrays by the combination of NSL and traditional

11.2 Nanosphere Lithography (NSL)

ultraviolet lithography (UV-LIGA), where Ag NPs and nanoarrays can be fabricated by NSL in the preformed multihierarchy arrayed transparent microwindows on the substrates (e.g. glass cover slip) by the UV-LIGA. This technique permits easy characterization of the 3D morphologies of single NPs by AFM or SEM and their LSPR spectra using dark field optical microscopy and spectroscopy (DFOMS). It is also possible to investigate the local morphology dependence of the LSPR spectra of the single NPs and nanoarrays. In this method, MHMW are first fabricated on a glass cover slip using the standard photolithography, whose details are shown in Ref. [4]. Figure 11.4A,B shows one example of the designed MHMW (three tiers) and the typical final microwindows (Figure 11.4C) pattern after printing. The MHMW on the glass cover slip are used to identify the location and orientation of single NPs, whose tiers can be determined by the observed field at desired resolution. For example, in the first tier of the MHMW (Figure 11.4A), each local area can be discerned by marking its X and Y number, such as the shaded area X1–Y2. Then, in the second tier of the MHMW (Figure 11.4B), the scale can be reduced by M or N times and each local area can also be marked by x and y number. If this area is the sub-tier in the shaded area of the first tier, it can be labeled as X1–Y2–x3–y3. In a similar way, step by step, we can reach the last tier with several transparent microwindows available (Figure 11.4C), in which the desired NP can be made by different fabrication methods (e.g. electron beam LIGA or NSL). NPs less than 10 nm of different shapes synthesized by a wet-chemical process can be

b1

a1

c1 yn yn–1

YN

+++

+++

+++

yn–2

YN–1 +

+

+

+

+

+

+

+

+

+

+

+

R

y4

Y3 Y2

+++

Y1 X1 X2 X3 (A)

XM–1 XM

y3 +++ y2 y1 x1 x2 x3 x4 xm–1 xm (B)

D

S

T

H

(C)

Figure 11.4 The MHMW on the substrate (e.g. glass cover slip). (A) The first tier of the MHMW, each local area can be discerned by marking its X and Y number, such as the red-dashed square area of X1–Y2. (B) The second tier of the MHMW, whose scale can be reduced by M or N times, whose local area can also be marked by x and y numbers. If this net area is the sub-tier n the red area of the first tier, it can be labeled as X1–Y2–x3–y3. Step by step, the last tier with several uniquely shaped transparent windows can be reached. The open windows can be made with different shapes. (C) The nanoparticles can be fabricated on the micro-pattern by various methods (e.g. nanosphere lithography). In each window, the same nanoparticle can be identified by comparing the images taken by optical microscopy, AFM, or other microscopy methods. Finally, the structural parameters (size, shape, orientation, interparticle spacing, and thickness) can be correlated with their optical responses. Source: Reprinted from Song Y.; et al., Nanoscale 2011, 3, 31–44, Figure 7, copyright [2011] from the Royal Society of Chemistry.

345

346

11 Template-assisted Fabrication of Nanostructure Thin Films

immobilized by a routinely diluted deposition. Consequently, the same NP in each window can be identified by comparing the images taken by the optical microscope with those characterized by the AFM. Finally, in each window, the same NP can be characterized by different techniques (e.g. DFOMS and AFM) allowing correlation of its 3D morphology with its optical response.

11.2.3 Identification of Localized Surface Plasmon Resonance (LSPR) of Single Noble Metal Nanoparticles and/or Nanoarrays by MHMW-assisted NSL A typical example to identify NPs and nanoarrays using both AFM and DFOMS is illustrated in Figure 11.5. Triangular Ag NPs and hexagon-arranged nanoarrays fabricated on the surface of glass cover slips within the nearly circle-shaped microwindow can be identified and characterized using AFM (Figure 11.5a,b is the 3D AFM image of the dash-squared area in Figure 11.5a) and DFOMS equipped with a color camera (Figure 11.5c) and charge-coupled device (CCD) camera (Figure 11.5d). The CCD camera offers higher spatial resolution than the color camera, while the color camera provides the real color of individual Ag NPs that are generated by LSPR. The center of each individual NP in the optical images recorded by the CCD is located with a single-pixel resolution (each pixel can be 125 or 67 nm depending on the CCD resolution and equipment setup) by determining the address of the pixel with the highest intensity of the NP. The positions of individual NPs of interest (e.g. the circled one) within the microwindow in the optical images (Figure 11.5c,d) are then determined with a spatial resolution limited by the optical diffraction limit (200 nm) and an orientation angle resolution of about 1.0∘ . This approach allows us to correlate AFM images of individual NPs (as the one circled in each image) with the same NP shown in its corresponding optical image and to investigate its 3D morphology-dependent LSPR properties. Clearly, these triangular NPs in this window almost show the same scattering color (Figure 11.5c) and intensity contrast (Figure 11.5d). By comparing their scattering color images (Figure 11.5c) with their AFM images (Figure 11.5a,b) of these NPs, it is once again showing that NSL is powerful method in the fabrication of uniform triangular NPs and nanoarrays. We have used it to investigate size- and shape-dependent LSPR spectra of single Ag NPs by the analysis of the experimental results with the theoretical calculation (i.e. discrete dipole approximation (DDA) simulation). Figure 11.6 gives the AFM images of one specific triangle-shaped Ag NPs characterized by MHMW. The AFM image of the triangular silver NP shows that it has the edge length of 375–420 nm (Figure 11.6A) and the out-of-plane height of about 16.1 nm (Figure 11.6B). This NP shows multi LSPR scattering colors (Figure 11.6C), as further evidenced by its multi-mode LSPR peaks at 562.3, 659.9, and 759.6 nm (Figure 11.6D-b). The peak wavelengths, peak ratios, and full wave at half maximum (FWHM) at 562.3 and 659.9 nm from experiment are in good agreement with DDA simulation for its LSPR scattering (Figure 11.6D-c), as have been summarized together with other shaped NPs fabricated by the modified NSL in Ref. [4]. In general, the DDA simulation

11.2 Nanosphere Lithography (NSL)

1

1 μm

2

1 μm

3 4

(a)

(b)

1 μm (c)

1 μm (d)

Figure 11.5 One example for the identification of the specific nanoparticles and nanoarrays in different instruments via MHMW based on Ag triangle nanoparticles and nanoarrays fabricated by NSL in one nearly circle-shaped window. (a) The plan view of the hexagon-arrayed triangular Ag NPs in one circle microwindow scanned by AFM; (b) The 3D view of the hexagon-arrayed triangular Ag NPs marked in the large pink dash-square in (a); (c) The real scattering color of these hexagon arrayed triangular Ag NPs observed under a dark-field microscope; (d) The CCD images of the scattering light of these hexagon-arrayed triangle Ag NPs recorded by a CCD camera equipped in the dark-field microscope. The dashed circles in each image refer to the same specific particle and the dashed squares in each image refer to the same specific nanoparticle pair. Source: Y. Song, China Patent, Appl. No. CN200910085973.9.

shows best agreement with the experimental spectra for NPs; hence their shapes can be accurately modeled. However, it can also be seen that for wavelengths longer than 650 nm for the investigated NPs, the experimental result has a lower intensity than the simulation. By analysis the instrument errors and the wavelength dependent CCD quantum efficiency, these deviations are deduced by the precision in the shape construction during DDA simulations. From these results, it was also found that when the shapes and 3D morphology of the NPs became more complicated, the deviation between the DDA simulation and the experimental result increased. This is due to the geometrical deviation between the real NPs and

347

0.50 0.75 µm

11 Template-assisted Fabrication of Nanostructure Thin Films

0.25

H = 16.1 nm

20.0 (A)

–20.0

0

0

200 nm

(B) 2 N.M.intensity (a.u.)

348

562.3 1.5 540

659.0 662

1

759.6

0.5

(b)

(d) 1020 (c) (a)

0 400

200 nm (C)

600 800 1000 Wavelength (nm)

1200

(D)

Figure 11.6 (A) The AFM image for one single triangular shaped Ag nanoparticle with edge length of 375–420 nm; (B) the height mapping of the triangle shape Ag nanoparticle along the direction of the arrow in Figure 11.6A, showing the out-of-plane height of this nanoparticle of 16.1 nm; (C) the real scattering color image of this triangular Ag nanoparticle; (D-a) the LSPR absorption spectrum of this nanoparticle by DDA; (D-b) the LSPR scattering by experiment; (D-c) the LSPR scattering by DDA; (D-d) the LSPR extinction by DDA. In order to identify the location and the orientation of these positions around the NPs, the AFM image and color image were netted by dashed lines with each square unit of 125 nm × 125 nm after their distances and orientations are corrected. Source: (a, c) Adapted from references: Y. Song, China Patent, Appl. No. CN200910085973.9; (b, d) Song et al. [2]. Copyright [2011] from the Royal Society of Chemistry.

the regular species used in the calculations. If these two instrumental factors and the geometrical deviation of NPs are considered, the corrected experimental results will match with the DDA simulation very well. This result also confirms that our experimental method (DFOMS), based on the far field detection, preserves the ability to detect the near-field LSPR signal. This combined method based on the NSL and the MHMW also allows us to investigate the 3D morphology dependent tip–tip LSPR coupling of triangular NP pairs. The zoom-in AFM image for the detailed 3D morphology of one typical Ag nanoprism pair is shown in Figure 11.7A. The nanoprisms have almost the same edge size of 375 nm and maximum out-of-plane height 17.1 nm as shown in Figure 11.7B by the

11.2 Nanosphere Lithography (NSL)

nm

Section analysis

–20.0

0

20.0

typical height map along the arrowed tip–tip direction in Figure 11.7A. The real scattered color for the nanoprism pair, taken from dark-field microscopy, is shown in Figure 11.7C. Both of the nanoprisms in the pair give red color with different brightness, which might be due to variation in their surface roughness, slightly difference in the underlying surrounding dielectrics, and the focusing distance during image recording. The middle area between the two nanoprisms clearly shows more reddish color than the optical centers of the two nanoprisms. Using their CCD image (not shown here) for the location identification, together with that obtained by the DDA calculation of the nanoprism pair, the LSPR spectrum for the middle area of the two optical centers (representing the tip–tip-coupling) is recorded in Figure 11.7D. Based on the 3D morphology of the NP pair, the two nanoprisms can be treated as regular triangular nanoprisms with the bottom edge length of 375 nm, the top edge length of 125 nm, and out-of-plane height of 17.1 nm for conducting the DDA calculation

(A)

0

1.00

2.00

3.00

(B)

N.M. scattering (a.u.)

1.5 600 605 1

0.5

536 520 754

b a

0 450 (C)

550 650 750 Wavelength (nm)

850

(D)

Figure 11.7 Tip–tip coupling of one typical pair of triangular Ag nanoprisms in the arrays with interspacing of 103 nm is characterized using (A) AFM, (B) the scan of out-of-plane height of the two nanoprisms along the arrow direction in (A), (C) the color image taken from darkfield microscopy, (D-a) the LSPR scattering spectrum at the central locations (three pixels) of the source nanoprisms, and (D-b) the LSPR scattering spectrum of the pair calculated by DDA. Source: (a, c) Y. Song.

349

350

11 Template-assisted Fabrication of Nanostructure Thin Films

of the nanoprism pair. The recorded LSPR spectrum (Figure 11.7D-a) at the middle optical center of the two nanoprisms shows three distinct peaks: one strongest peak at 605 nm, one shoulder at 536 nm, and one secondary strong peak at 754 nm. By comparing the experimental result for the tip–tip coupling of the nanoprism pair with the DDA calculation (Figure 11.7D-b), it can be deduced that the peak at 605 nm represents the in-plane quadrupole resonances originated from the two source nanoprisms and the peak at 536 nm is from the out-of-plane quadrupole resonances of the two source nanoprisms. Although the DDA simulation does not show one distinct peak at 754 nm, our experiment result suggest one strong peak at this wavelength, which is probably from the strong tip–tip coupling. In order to reveal whether the peak at 754 nm is mainly from the tip–tip coupling or not, the LSPR spectra from the optical centers of the source nanoprisms, are recorded (not shown here), showing one strong peak at the same position. Generally, one can see that the peak positions and shape resonances for the two nanoprisms are almost the same, suggesting that the NSL process is very powerful in the fabrication of the nanoprisms with almost identical 3D morphologies and surroundings. Both of the two triangle nanoprisms do not give the peak at 754 nm as strong as the pair, confirming that the additional peak at 754 nm indeed is from the tip–tip coupling. However, previous investigations did not show additional strong peak due to tip–tip coupling. This significant coupling between the nanopair may be caused by the unique size of our particles that is just lying in the range of half wavelength of visible light, which can cause a strong long-range electrodynamic interaction among light and the collective electrons on the particle surfaces. In addition, our experimental observations show that nanoprism coupling does not affect the quadrupole mode in LSPR significantly, resulting in little shifts in the highest peak at 598–605 nm (the in-plane quadrupole mode). However, one additional peak (i.e. 754 nm) as compared with the in-plane quadrupole mode can be observed. This peak resulted from LSPR coupling is in good agreement with the prediction by the semi-analytical model by Schatz and coworkers [5]. In the present study, the edge lengths of the triangular nanoprisms are more than 𝜆/2𝜋 (64–128 nm), which is more than the critical scale in the semi-analytical model in the DDA. Therefore, the long-range electrodynamic interaction, not electrostatic effects, will be dominant in the LSPR of the two nanoprisms. The center-to-center interspacing of the two nanoprisms is 532 nm, more than the critical interspacing. As a consequence, the coupling will be mainly determined by the long-range radiative dipolar interactions (or radiative damping effects), and phase retard effects, resulting in one new peak with wavelength more than the highest peak for the two nanoprisms. Based on this combined method, we have investigated the distance dependent tip–tip coupling between triangular Ag nanoprism pairs with dimensions at the range of half wavelength of visible light and distance ranging from 100 to 400 nm. It has been found that the coupling peak wavelength increases and the coupling intensity decreases with the increased tip–tip distance, and finally the coupling disappears (no coupling peak) when the tip–tip distance is more than about 400 nm due to the coupling intensity becomes extremely low.

11.2 Nanosphere Lithography (NSL)

Generally, the combination of NSL and the MHMW fabricated by the routine UV-LIGA shows a powerful ability not only in the identification of NPs and nanoarrays but also in the precise investigation of the fundamental theory related to the 3D morphology dependent LSPR and LSPR coupling. In our study, the detector is far-field, while the DDA calculation is based on the near-field. Thereby, the results indicate that the near-field LSPR of single NPs and the coupling signals of nanoarrays can be detected by the far-field detector if the 3D morphologies of NPs or nanoarrays can be precisely accounted for in the DDA model.

11.2.4 Development of NSL for Nanomaterials Synthesis 11.2.4.1 Aqueous Phase Ag Nanoparticles with Controlled Shapes Fabricated by NSL

The chemical and physical properties (e.g. magnetic, catalytic, or optical) of NPs depend on their sizes, shapes, and spatial arrangements. Recent investigations have demonstrated that the optical properties of noble metal NPs significantly depend on their shapes and sizes. Therefore, the ability to productively control the shape of the fabricated NPs is of much interest. Chemical synthesis of noble metal NPs relies on the ability to reduce a metal salt in a controlled environment. An alternative to solution phase NP fabrication is to fabricate NPs on solid substrates and then release them into a solution. Recent progress in NSL has shown that this method provides a good template for shape-controlled fabrication of surface-confined NPs. This also allows for flexible institutionalization on the clean surface. van Duyne and coworkers have developed this process and used it to fabricate solution phase NPs in ethanol [6]. However, their results indicated that most of the NPs in the solution have nonuniform surface morphologies with truncated tips in addition to the presence of debris and some of the NPs attached together on the glass substrate surface causing the agglomeration of the released NPs. In addition, aqueous phase NPs are expected to be more biocompatible than those in ethanol. We present a modified NSL process to fabricate Ag NPs with controlled shapes on glass substrates. These modifications were made by thermal annealing of the triangular nanoprisms, sonication to remove weak tips, slight etching of the glass substrates under the NPs, removing debris and small broken parts around the NPs on the substrates. Thiol compounds were adsorbed on the NPs prior to dislodging them into water. These aqueous phase NPs show much more uniform shapes than those obtained by the traditional NSL process and give a narrower UV–Vis absorption peak [7]. Figure 11.8a shows AFM image of Ag NP array fabricated on a glass substrate using the established NSL process after removing the NS templates. The produced Ag NPs show hexagonally arranged triangular nanoprisms. The magnified image in the insert shows that for some NPs their tips and edges have many defects due to NP damage during liftoff by sonication. From the 3D images for some of these Ag NPs (Figure 11.8b), it can be clearly seen that the Ag NPs are triangular nanoprisms with increased thickness from edges to the center (30 ± 0.5 nm). However, as observed in the insert of Figure 11.8a, many of the NPs show rough surface morphology. NP tips

351

11 Template-assisted Fabrication of Nanostructure Thin Films

or edges form weak parts in the NPs which can easily break during sonication. This significantly reduces the shape uniformity of the NPs and produce debris in the solution. Transmission electron microscopy (TEM) images for these aqueous phase NPs show many small NPs present with prism shape (Figure 11.8c). The calculated edge size of the triangular prism NPs is ∼78 nm when using 290 nm diameter NSs. This size is consistent with that observed from the AFM images of unreleased NPs shown in Figure 11.8a,b. For the released NPs, the TEM image in Figure 11.8c shows the expected triangular prism shape (dashed ellipses) and some small irregular spherical shaped NPs (dotted circles). Their size distributions are shown in Figure 11.8d,e. For the triangular prism NPs, a broad size distribution of 74.3 ± 31.0 nm (in addition to some smaller triangular NPs with an edge size of ∼20–30 nm) is observed. The circular NPs had a size distribution of 10.0 ± 4.6 nm. The Ag NPs observed with one or two truncated tips (some of the dashed ellipse) are broken tips from the original NPs produced by NSL. Some of these triangular prism NPs have rough surfaces and their

100 nm 200 nm

0.2 0.4 0.6

1 μm

0.8

200 nm 1.0

(b) Frequency (counts)

10 8 6 4 2

40 20

2 4 6 8 10 12 14 20 30

140

110

80

50

Size (nm)

Size (nm)

(d)

0.4

0 20

0

(c)

60 Absorption (a.u.)

(a) Frequency (counts)

352

(e)

(f)

100 nm

0.35 0.3

605

0.25

352 100 nm 0.2 300 400 500 600 700 800 900 Wavelength (nm)

Figure 11.8 Ag NPs fabricated by the routine NSL process. (a) AFM image of an area of surface-confined triangular prism Ag NPs. The inserted magnified image shows the weak tips and rough surface morphologies. (b) 3D AFM image of the triangular prism NPs. (c) TEM images of the released aqueous phase Ag NPs show different kinds of NPs: some are the triangular prism Ag NPs with or without truncated tips (labeled with dashed ellipses and shown in the top-right inset) and some are the small NPs from debris produced during sonication (some irregular spherical debris labeled with dotted circles and some small triangular debris from broken tips of the triangular nanoprisms labeled by dashed squares). (d) Histogram of triangular shaped Ag NPs gives mean sizes of 74.3 ± 31.0 nm. (e) Histogram of the irregular spherical shaped Ag NPs gives mean sizes of 10.0 ± 4.6 nm. (f) The UV–Vis absorption spectrum for the aqueous phase NPs show two distinct peaks; one sharp peak at 352 nm from the combined contribution of the LSPR of the small Ag NPs and the high-mode LSPR of the triangular Ag NPs, another broad peak at 605 nm mainly from the dipole LSPR of the triangular Ag NPs. Source: Song and Elsayed-Ali [7]. © 2010, Elsevier.

11.2 Nanosphere Lithography (NSL)

50 nm

100 nm

Frequency (Counts)

thickness varies considerably over the triangular surface of the NPs. This roughness appears to be caused by NP damage during sonication. These size and shape features of the aqueous phase NPs are also reflected by their UV–Vis absorption spectrum, which shows two distinct absorption peaks centered at ∼352 and 605 nm, as shown in Figure 11.8f. We next investigated the shape integrity of the heat-treated NPs after releasing them into water. Figure 11.9a shows TEM image of the Ag NPs after thermal annealing without pre-sonication. Most of those Ag NPs show triangular shapes with round tips (dotted circles in Figure 11.9a) and some with snipped tips (dashed circles in Figure 11.9a). The inset is a magnified image of these NPs, clearly showing a triangular shape with round tips. The histogram for these Ag NPs (Figure 11.9b) gives a mean size of 39.6 ± 4.9 nm with much narrower size distribution of STDEV% = 12.4% than those obtained from surface-confined Ag NPs without any post-annealing (Figure 11.8d, STDEV% = 41.7%). Figure 11.9c is a TEM image for Ag NPs that were thermally annealed after removing two tips by sonication, whose histogram gives a mean size of about 33.9 ± 6.8 nm (Figure 11.9d), less than that for those triangular shaped NPs with round tips after post-annealing. Most of these NPs show quadrilateral shapes (dashed circles) or pentagon shapes as shown more clearly in the inset of Figure 11.9c. These NPs have a similar shape as those

15 10 5 0

35

40 45 Size (nm)

50

55

50 nm

Frequency (Counts)

(b)

(a)

16 12 8 4 0

25

30

35

40

45

50

55

60

65

Size (nm)

100 nm

(c)

30

(d)

Figure 11.9 TEM images of the aqueous phase Ag NPs after surface modification by thiol compounds and dislodging from the glass substrate. (a) Triangle Ag NPs with round tips. Dashed circles: Ag NPs with round tips; dotted circle: Ag NPs with slightly round tips. (b) Histogram of triangular shaped Ag NPs with round tips based on 45 NPs giving a mean size of 39.6 ± 4.9 nm. (c) Quadrilateral and pentagon-shaped Ag NPs. Dashed circles: some typical Ag NPs with quadrilateral shapes. (d) Histogram of quadrilateral and pentagon-shaped Ag NPs based on 45 NPs giving a mean size of 33.9 ± 6.8 nm. Source: Song and Elsayed-Ali [7]. © 2010, Elsevier.

353

11 Template-assisted Fabrication of Nanostructure Thin Films

476

672

(a)

Absorption (a.u.)

354

352

(b) 470

300

400

626

532

(c)

500

600

700

800

Wavelength (nm)

Figure 11.10 Optical absorption of Ag NPs. (a) Surface-confined Ag NPs before surface modification. (b) Surface-confined Ag NPs after surface modification with thiol. (c) Aqueous phase Ag NPs after releasing the surface-confined NPs into water. Source: Song and Elsayed-Ali [7]. © 2010, Elsevier.

observed by AFM images in Figure 11.3e,f. From the TEM images in Figure 11.9c, some of the NPs give less contrast in their central parts (NPs labeled by dashed circles). We believe that the lighter centers in these NPs are from a thinner center resulting from adhesion of the center of these NPs to the glass substrate during annealing. AFM observation of the glass substrate after removal of the NPs show debris forming hexagonal shaped arrangements. This observation is consistent with adhesion of the central part of the triangular prism NPs to the substrate. The UV–Vis absorption spectra of the surface-confined NPs change significantly before and after surface modification. Figure 11.10a shows the absorption spectrum of surface confined Ag NPs fabricated by NSL without any modification. It can be seen that the absorption spectrum for the Ag NPs yields two distinct peaks at 476 and 672 nm. Based on previous investigations on these NPs and their arrays [6, 8, 9], it is reasonable to expect that the absorption peak at 476 nm is primarily from the higher order mode SPR (e.g. quadrupole) of the NPs, and the peak at 672 nm is primarily from the dipole resonance of the NPs. We note that the higher order resonance peak has almost the same intensity as that for the dipole resonance, although the higher order modes are expected to be much weaker than the dipole resonance. Since the substrate is continuously covered by a hexagonally arranged array of Ag NPs with tip–tip distance less than 100 nm, we postulate that particle–particle coupling will contribute to the LSPR spectrum. This particle–particle interaction effect could be responsible for the observed spectrum. Before releasing the Ag NPs into water, their surfaces must be modified by a water-soluble stabilizer. We used the combination of 1-OT and 11-MUA. The surface modification of the Ag NPs was conducted using the routine thiol coupling chemistry. In order to confirm that the thiol compounds are linked on the Ag NPs, X-ray photoelectron spectroscopy (XPS) spectrum for the surface-confined Ag NPs after modification by 1-OT and 11-MUA was obtained. Before XPS was performed,

11.2 Nanosphere Lithography (NSL)

Ag3d5 Ag3d3 Ag3s Ag3p3

C2s

O1s

C1s

Ag4p S2p Si2p

Ag4s Ag3p1 1000

900

800

700

600

500

400

300

Si2s 200 100

0

Binding energy (eV)

Figure 11.11 XPS spectrum for the surface-confined Ag NPs after surface modification by 1-OT and 11-MUA. Source: Song and Elsayed-Ali [7]. © 2010, Elsevier.

the glass substrate with the modified Ag NPs was thoroughly washed by ethanol to remove the unbound thiol compounds (1-OT, 11-MUA, etc.). The obtained XPS spectrum in Figure 11.11 shows binding energy peaks for Ag3s (700 eV), Ag3p1 (608 eV), Ag3p3 (578 eV), Ag3d3 (376 eV), Ag3d5 (375 eV), Ag4s (334 eV), Ag4p (69 eV) in addition to peaks for C1s (292 eV), C2s (15 eV), O1s (537 eV), and S2p (170 eV) originating from the thiol compounds. The height of the Ag NPs before and after the thiol modification was also measured by AFM; it increased by ∼2.9 nm after the thiol modification. Since the calculated height of MUA is ∼1.7 nm, the adsorbate may be one layer formed by mono or double assembly of MUA and 1-OT molecules. The variation between the calculation and the AFM result may be resulted from the height increase of NPs during immersion in the thiol solution, similar to the solvent annealing effect [10–12]. These experimental results confirm that at least one layer of thiol compounds is bonded on the Ag NP surface. Since this adsorbed layer is only ∼2.9 nm thick, it is not expected to produce significant change to the surface dielectric properties of the Ag NPs. However, the silver–sulfur bonding between the NPs and the thiol compounds may affect the conductivity and the number of free surface electrons in the Ag NPs. The absorption spectrum for the modified Ag NPs (Figure 11.10b) shows a slight blue shift at the peak at 476 nm (to 470 nm) and a significant blue shift at 672 nm (to 626 nm) with reduced intensities. This spectrum was expected to give a red shift due to the increased dielectric constant from the adsorbed thiol compounds [13]. We attribute this blue shift to shape variation (e.g. increased height, round tips, smoother surface topography) during surface modification by immersion that was similar to solvent annealing which results in blue shift of LSPR since any solvent annealing has not been done on our NPs [10–12]. These variations are further observed by the slightly reduced NP size and round shapes observed in the TEM image in Figure 11.9 when compared with the AFM image in Figure 11.3. In addition, when the Ag NPs are covered by thiol groups, the surface free electron density may be reduced, leading

355

356

11 Template-assisted Fabrication of Nanostructure Thin Films

to weaker SPR in single NPs and SPR coupling among NP arrays [14]. This will also result in a blue shift of the LSPR peak and a reduced LSPR intensity. The UV–Vis absorption spectrum of the Ag NPs after release in water, shown in Figure 11.10c, was compared to surface-confined NPs. The aqueous Ag NPs give a main peak at 532 nm and a very weak peak at 352 nm. The main peak at 532 nm appears to be from LSPR by the triangular prism NPs with round tips and is blue-shifted from that obtained for NPs fabricated by the routine NSL and released from the surface that showed LSPR peak 605 nm. This is attributed to the reduced size and round tips. The peak at 352 nm in Figure 11.10c becomes much weaker and narrower than that for the aqueous Ag NPs released from the surface-confined Ag NPs as fabricated by routine NSL (shown in Figure 11.8f), obviously due to the shape variation of NPs and almost no small spherical shaped debris observed in the aqueous Ag NPs released from the surface-confined Ag NPs fabricated by the modified NSL (Figure 11.10a). By comparing the TEM images for the two kinds of Ag NPs, it can be deduced that the peak at 352 nm in Figure 11.10c is mainly from the out-of-plane quadrupole resonance of Ag nanoprisms with round tips according to previous investigation. The peak intensity ratio between the main peak at 532 nm and the weak peak at 352 nm for these NPs is ∼11.5 : 1 (after subtracting the background), which is much higher than that for the NPs obtained by the routine NSL and releasing process (1 : 3.6). Clearly, the number of small debris caused by sonication is greatly reduced using the modified NSL and releasing process. The modified NSL process favors the formation of uniform Ag NPs with round tips with significant reduction in Ag debris, as shown in Figure 11.9. In addition, 1-OT and 11-MUA can be substituted by the combination of 1-BT and TP, or MCH and MUA for more water-soluble NPs. 11.2.4.2 Ultrathin Nanopore Arrays with Uniform Size

The magneto-optical properties and the related proximity effect rely on the detailed morphology of each component and the dielectric properties of the media. The correlation between these parameters and their magnetic properties, SPR and magneto-optical properties has not been well characterized, even though some progress has been achieved. Particularly, research on how to utilize proximity effects among different components to improve the magneto-optical effect at low energy loss, critical for many applications, is just in its infancy. Most current studies have investigated ensembles of NPs in which it is difficult to guarantee homogeneity of composition, structure, and morphology by conventional preparation methods. This makes it difficult to identify the intrinsic proximity effects from the nano-structures and associated interparticle interactions, which are necessary for precisely addressing the relationship between structures and their properties for high performance. Therefore, it is essential either to characterize single nanostructures or to develop controlled fabrication processes for size homogeneity [15]. In this chapter, we introduce an NS TA multistep sputtering technique to fabricate multi-layered uniform nanoporous films, using nanoporous Ag/CoFeB/Ag films as examples. Structural parameter dependent magnetic properties, LSPR and light reflection tunable by a magnetic field at different wavelengths (magneto-optical

100 Counts (numbers)

Counts (numbers)

11.2 Nanosphere Lithography (NSL)

80 60 40 20 0

80 60 40 20 0 775 806 838 863 Diameter (nm)

934 952 967 988 1012 Diameter (nm)

2 μm

(a)

100

888

2 μm

(b)

Figure 11.12 SEM images of the monolayer nanosphere template (a) and the formed periodic nanoporous structure (b). The insets show the diameter histograms of nanospheres and nanopores. Source: From Song et al. [15]. Licensed under CC BY 3.0.

Kerr effect) in these hybrid nanostructures are experimentally investigated and theoretically analyzed. 11.2.4.3 Fabrication of Periodic Uniform Nanoporous Films with Controlled Layers

NSs with size dispersion below 4% can self-assemble into a uniform monolayer on a large scale (exceeding 5 mm × 5 mm) using a modified drop-coating process. Figure 11.12a shows a typical SEM image of the monolayer NS template with an NS diameter of 970 ± 19 nm. Using this template, uniform and periodic nanopore films with controlled layers can be fabricated by a multi-step magneto-sputtering process. As shown in Figure 11.12b, the resulting nanopores have a diameter of 840 ± 29 nm, and the thinnest pore wall is about 130 nm thick. The calculated area of the walls is around 37.5 V% of the total film. It is interesting to compare the uniform porous nanostructures shown in Figure 11.12 with our previous reports on the fabrication of separated triangular or square shaped NP arrays by thermal evaporation. The formation of a periodically arranged nanoporous film instead of separated NPs may be attributed to the effects from the high energy sputtered atoms/clusters having sufficient kinetic energy to penetrate the triangular holes between the NSs. This result provides us a new method to prepare surface confined uniform nanopore arrays using NS templates over a large scale, whose pore diameters and periods can be controlled by the diameter of the NSs. Results on the magnetic properties (Figure 11.13, Table 11.1) of these periodic nanoporous films suggest that enhanced magnetic response can be induced by the Ag layers and the nanopores even though they both hinder the rotation of magnetic domains away from the in-plane direction. The nanopores can result in a strong pinning of MDW, an increased local shape anisotropy and non-uniform spin distribution from the local dipolar field around the nanopores. The Ag layers mainly produce a strong interfacial magneto–electric coupling due to the increased conductivity, the pinning effect of Ag layers and the mixing of itinerant electrons with ferromagnetic d electrons at the Ag–CoFeB interface. Fabrication

357

11 Template-assisted Fabrication of Nanostructure Thin Films

80

(a) (b)

0 –40 –80

(a)

40 M (emu/g)

M (emu/g)

40

(a) (b)

20 0 –20 –40

–10 000

–5000

0 H (Oe)

5000

10 000

–10 000

(b)

–5000

0 H (Oe)

5000

10 000

40 20 M (emu/g)

0 –10

20 M (emu/g)

(a) (b)

10

–20 –10 000

(c)

–5000

0 H (Oe)

5000

0 –20 –40

10 000

(a) (b)

(d)

–10 000

–5000

0 H (Oe)

5000

10 000

10 (a) (b)

5 M (emu/g)

358

0 –5 –10

(e)

–10 000

–5000

0

H (Oe)

5000

10 000

Figure 11.13 Room temperature hysteresis loop of CoFeB film (A), nanoporous CoFeB film (B), Ag/CoFeB/Ag film (C), nanoporous Ag/CoFeB/Ag film (D), and nanoporous Ag/CoFeB/Ag without removing nanospheres (E). All nanoporous films have hexagonally periodic nanopores of circular shape. (a) In-plane magnetization (with the magnetic field parallel to the film surface); (b) out-of-plane magnetization (with the magnetic field perpendicular to the film surface). Source: From Song et al. [15]. Licensed under CC BY 3.0.

Table 11.1

| In-plane (IP) and out-of-plane (OP) magnetic properties of samples.

Magnetization direction parameters

Samples

IP Hc Oe

He Oe

Mr / Ms

OP HK Oe

Hc Oe

He Oe

Mr / Ms

HK Oe

CoFeB film

14

0

0.730

7

123

1

0.064

1958

CoFeB nanopore film

18

0

0.748

9

69

5

0.039

1487

0.213

965

Ag/CoFeB/Ag film

25

8

120

111

58

0.070

Ag/CoFeB/Ag nanopore film

35

20

0.182/ 221 0.101a)

122

91

0.091/ 1796 0.065a)

Ag/CoFeB/Ag nanopore film with nanospheres

32

13

0.173/ 290 0.088a)

95

59

0.092/ 1440 0.075a)

a) Right/Left: Mr./Ms. from the top/bottom half of the hysteresis loops. Source: From Song et al. [15]. Licensed under CC BY 3.0.

11.3 Anodic Aluminumoxide (AAO) Template-assisted Nanoimprinting

of nanopores in the CoFeB continuous thin film increases the field parallel to the film plane (IP) ferromagnetic character but reduces the field perpendicular to the film plane (OP) ferromagnetic character based on the changes in the squareness (remanence/saturation magnetization: M r /M s ) and incoercivity H c . However, from the M r /M s , H c and exchange bias H e of films with Ag layers, we infer that the formation of nanopores can increase the OP ferromagnetic character much more than the IP ferromagnetic character of the Ag/CoFeB/Ag film, due to the additional vertical magnetism caused by the much enhanced pore and Ag layer pinning effects perpendicular to the film plane. Therefore, Ag layers have the potential to improve the magnetic performance of the nanoporous magnetic film as a perpendicular recording media. According to the observed changes H c , H e and OP M r /M s , a nanoporous film with Ag layers has a strongly enhanced ferromagnetic character, exceeding that caused by the simple addition of the Ag layers or nanopores alone. Thus, there exists a synergistic effect between the nanopores and Ag layers in improving the ferromagnetic character of the magnetic film. In conclusion, a methodology based on NS templating and multistep magneto-sputtering process is successfully developed to fabricate surface confined uniform multilayered nanoporous arrays over a large scale. These hybrid nanostructures show unique magnetic and LSPR properties due to a synergistic interaction of periodic nanopores and Ag layers, compared with those films without the nanopores and/or Ag layers. This synergistic interaction can be used to improve the ferromagnetic character of the magnetic film or to tune their optical properties. By controlling the geometry of the formed nanostructures, it is possible to either enhance the light extinction by the synergistic effect of nanopores and surface magneto-plasmonic interference (such as in the nanoporous Ag/CoFeB/Ag film without removing NSs) or to foster the light transmission through the film (such as in the nanoporous Ag/CoFeB/Ag film) for different applications. The nanoporous Ag/CoFeB/Ag film shows strongly enhanced magneto-optic effects both in peak intensity and in peak wavelength changes as compared with the CoFeB film, the nanoporous CoFeB film and their precursors with NSs due to the magneto–-plasmonic interaction between the magnetic layer and the LSPR layer, endowing them with great potential for applications in high sensitivity magneto-optic sensing systems, such as optical transistor, circular isolator, active waveguides and gratings, or magnetic field sensors.

11.3 Anodic Aluminumoxide (AAO) Template-assisted Nanoimprinting 11.3.1 Introduce of AAO Template AAO templates have the advantages of low cost, high maneuverability, high temperature resistance, insulation, and highly ordered pore structure [16]. They have many applications in the field of nanomaterials, such as preparation of nanoarrays, nano-replication, preparation of quantum dots, etc. In 1953, Keller et al. first set up

359

360

11 Template-assisted Fabrication of Nanostructure Thin Films

the theoretical model of anodic alumina, becoming a fundamental explanation for the growth mechanism of anodic alumina [17]. In the past few decades, there have been new theories, such as field dissolution model [18], volume expansion theory, isofield strength model, viscous flow theory, and so on, which have provided rich theoretical knowledge for the development of AAO template and brought new results to the preparation technology and practical application of AAO template. AAO is thin film material with unique nano-porous structure, which consists of substrate, aluminum barrier layer and porous layer (alumina). The porous layer is composed of hexagonal stacked cells. The center of the cell is a cylindrical hole growing vertically along the surface. The general pore depth refers to the thickness of the AAO alumina layer, the pore channels are parallel to each other, by the bottom barrier layer to connect the pore channel and the unoxidized metal, as shown in Figure 11.14. AAO template is prepared by anodic oxidation of aluminum sheet in acid (sulfuric acid, oxalic acid or phosphate acid, etc.) electrolyte under appropriate conditions. A highly ordered nanoporous array film with self-organization can directly regulate the structure