Click to View SlideBook Version

Graphene Technology

Graphene Technology

Edited bySoroush Nazarpour and Stephen R. WaiteGraphene Technology

Edited by Soroush Nazarpour and Stephen R. WaiteGraphene TechnologyFrom Laboratory to Fabrication

Editors All books published by Wiley-VCH are carefully produced. Nevertheless, authors,Dr. Soroush Nazarpour editors, and publisher do not warrant theGroup NanoXplore Inc. information contained in these books,25 Montpellier Blvd including this book, to be free of errors.Montreal Readers are advised to keep in mind thatQC H4N 3K7 statements, data, illustrations, proceduralCanada details or other items may inadvertently be inaccurate.Stephen R. WaiteGraphene Stakeholders Association Library of Congress Card No.: applied for640 Ellicott Street, Suite 499Buffalo British Library Cataloguing-in-PublicationNY 14203 DataUnited States 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 <>. © 2016 Wiley-VCH Verlag GmbH & Co. KGaA, 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: 978-3-527-33833-7 ePDF ISBN: 978-3-527-68757-2 ePub ISBN: 978-3-527-68755-8 Mobi ISBN: 978-3-527-68756-5 oBook ISBN: 978-3-527-68754-1 Cover Design Formgeber, Mannheim, Germany Typesetting SPi Global, Chennai, India Printing and Binding Printed on acid-free paper

V Contents List of Contributors IX1 Graphene Technology: The Nanomaterials Road Ahead 1 9 Stephen R. Waite and Soroush Nazarpour1.1 Newly Discovered 2D Materials 11.2 Wonder Materials 21.3 The Rise of MPM 51.4 Addressing the Environment, Health, and Safety 71.5 The Nanomaterials Road Ahead 71.6 Can Graphene Survive the “Disillusionment” Downturn?1.6.1 Gartner’s Hype Cycle 91.6.2 Surviving the Trough of Disillusionment 101.6.3 Graphene and Batteries 111.6.4 Heat Management with Graphene 131.6.5 How Graphene Could Revolutionize 3D Printing 142 Graphene Synthesis 19 Siegfried Eigler2.1 Introduction 192.2 Definitions 202.2.1 Nomenclature and Structure 202.2.2 Polydispersity of Graphene 202.3 Characterization of Graphene by Raman Spectroscopy 222.4 Epitaxial Growth of Graphene from SiC 262.5 Graphene by Chemical-Vapor-Deposition 272.6 Delamination of Graphene from Graphite 312.6.1 Mechanical Cleavage of Graphite 322.6.2 Liquid Phase Exfoliation of Graphite – Stirred Media Mills 332.6.3 Liquid Phase Exfoliation of Graphite – Sonication 352.6.4 Liquid Phase Exfoliation of Graphite – Shear Mixing 362.6.5 Liquid Phase Exfoliation of Graphite Using Smart Surfactants 382.6.6 Electrochemical Exfoliation of Graphite 382.7 Wet-Chemical Functionalization and Defunctionalization 40

VI Contents Reductive Functionalization of Graphene 40 Oxidative Functionalization of Graphene 43 2.7.1 Generalized Synthesis of GO 45 2.7.2 Historical Development of the Synthesis of GrO 46 Structure of GO 48 GO as Precursor for Graphene 49 Synthesis of Nanographene from Small Molecules 52 References 57 2.8 Graphene Composites 63 3 Suman Chhetri, Tapas Kuila, and Naresh Chandra Murmu Introduction 63 3.1 Preparation and Properties of Graphene 65 3.2 Functionalization of Graphene 66 3.3 Covalent Modification 67 3.3.1 Non-Covalent Modification 70 3.3.2 Preparation of Graphene Polymer Composites 71 3.4 In Situ Polymerization 71 3.4.1 Solution Mixing 72 3.4.2 Melt Mixing 72 3.4.3 Other Preparative Technique 73 3.4.4 Characterization of Graphene-Polymer Composites 74 3.5 Properties of Graphene/Polymer Composites 77 3.6 Mechanical Properties 77 3.6.1 Thermal Properties 84 3.6.2 Electrical Properties 88 3.6.3 Dynamic Mechanical Properties 93 3.6.4 Application of Graphene Based Polymer Composites 94 3.7 Gas Barrier 95 3.7.1 Sensor 97 3.7.2 EMI Shielding 97 3.7.3 Flammability Reduction 99 3.7.4 Automotive and Aircrafts 99 3.7.5 Turbine Blades 100 3.7.6 Others 100 3.7.7 Conclusions and Outlook 101 3.8 References 102 4 Graphene in Lithium-ion Batteries 113 Cyrus Zamani 4.1 Introduction 113 4.2 Renewable Energies 114 4.3 Batteries, What are They? 115 4.4 Lithium-ion Batteries 116 4.5 Anodes, Cathodes, and Electrolytes 117

Contents VII4.6 Carbon Materials 1184.7 Graphite 1194.8 Graphene 1204.9 Graphene in Lithium-Ion Batteries 1214.10 Graphene in Anodes 1224.11 Graphene in Cathodes 1264.12 Graphene in Other Types of Lithium Batteries 127 Summary 127 References 1285 Graphene-Based Membranes for Separation Engineering 133 Luisa M. Pastrana-Martínez, Sergio Morales-Torres, José L. Figueiredo,5.15.2 and Adrián M.T. Silva5.3 Introduction 1335.3.1 Preparation of Graphene-Based Membranes 1345.3.2 Graphene-based Membranes for Separation Applications 1405.4 Gas Separation 140 Water Treatment 142 Conclusions 149 Acknowledgments 150 References 1506 Graphene Coatings for the Corrosion Protection of Base Metals 155 Robert V. Dennis, Nathan A. Fleer, Rachel D. Davidson, and Sarbajit Banerjee6.1 Introduction to Corrosion 1556.2 Bare Graphene as a Protective Barrier 1596.2.1 Some Electronic Structure Considerations at Graphene/Metal Interfaces 1596.2.2 Graphene as a Standalone Corrosion-Resistant Coating and Some Mechanistic Considerations 1626.3 Graphene Nanocomposites for Corrosion Inhibition 1646.4 Graphene/Metal Nanocomposites for Corrosion Inhibition 1686.5 Graphene/Ceramic Nanocomposites for Corrosion Inhibition 1716.6 Summary and Future Outlook 172 Acknowledgments 173 References 1747 Graphene Market Review 177 Marko Spasenovic7.1 Introduction 1777.2 Graphene Market: Past and Present 1787.3 Co-ordinated Market Initiatives 1847.4 Market and Application Projections 1857.5 Conclusion 186 References 187

VIII Contents Financing Graphene Ventures 189 Stephen R. Waite 8 Graphene Start-ups 190 The Art of Raising Capital 191 8.1 Shifting Financial Landscape for Graphene Ventures 199 8.2 The Graphene Financing Road Ahead 203 8.3 Summary 205 8.4 Appendix Nantero Case Study – The Funding and Evolution of a Nanomaterials Start-up 206 The Founding of Nantero 207 Series A: Financing Round 207 Post-Series A: Funding Evolution 208 Series B: Financing Round 208 Post-Series B: Funding Evolution 209 Series C: Financing Round 210 Post-Series C: Funding Evolution 210 Series D: Financing Round 212 Post-Series D: Funding Evolution 212 Series E: Financing Round 212 Summary 213 Index 215

IXList of ContributorsSarbajit Banerjee andTexas A&M UniversityDepartment of Chemistry Academy of Scientific andCollege Station Innovative Research (AcSIR)TX 77842-3012 CSIR-CMERI, CampusUSA Durgapur 713209 Indiaand Rachel D. DavidsonTexas A&M University Texas A&M UniversityDepartment of Materials Science Department of Chemistryand Engineering 580 Ross Street575 Ross Street College StationCollege Station TX 77842-3012TX 77843-3003 USAUSA andSuman ChhetriSurface Engineering & Tribology Texas A&M UniversityDivision Department of Materials ScienceCouncil of Scientific and and EngineeringIndustrial Research-Central 575 Ross StreetMechanical Engineering College StationResearch Institute TX 77843-3003Durgapur 713209 USAIndia

X List of Contributors José L. Figueiredo Laboratory of Separation and Robert V. Dennis Reaction Engineering Texas A&M University Laboratory of Catalysis and Department of Chemistry Materials (LSRE-LCM) 580 Ross Street Chemical Engineering College Station Department TX 77842-3012 Faculdade de Engenharia USA Universidade do Porto Rua Dr. Roberto Frias and 4200-465 Porto Portugal Texas A&M University Department of Materials Science Nathan A. Fleer and Engineering Texas A&M University 575 Ross Street Department of Chemistry College Station 580 Ross Street TX 77843-3003 College Station USA TX 77842-3012 USA Siegfried Eigler Friedrich-Alexander-Universität and Erlangen-Nürnberg (FAU) Central Institute of Materials and Texas A&M University Processes and Department of Materials Science Department of Chemistry and and Engineering Pharmacy 575 Ross Street Dr.-Mack-Str. 81 College Station D-90762 Fürth TX 77843-3003 Germany USA and Chalmers University of Technology Department of Chemistry and Chemical Engineering Kemivägen 10 SE-412 96 Göteborg Sweden

List of Contributors XITapas Kuila andSurface Engineering & TribologyDivision Academy of Scientific andCouncil of Scientific and Innovative Research (AcSIR)Industrial Research-Central CSIR-CMERI, CampusMechanical Engineering Durgapur 713209Research Institute IndiaDurgapur 713209India Soroush Nazarpour Group NanoXplore Inc.and 25 Montpellier Blvd MontrealAcademy of Scientific and QC H4N 3K7Innovative Research (AcSIR) CanadaCSIR-CMERI, CampusDurgapur 713209 Luisa M. Pastrana-MartínezIndia Laboratory of Separation and Reaction EngineeringSergio Morales-Torres Laboratory of Catalysis andLaboratory of Separation and Materials (LSRE-LCM)Reaction Engineering Chemical EngineeringLaboratory of Catalysis and DepartmentMaterials (LSRE-LCM) Faculdade de EngenhariaChemical Engineering Universidade do PortoDepartment Rua Dr. Roberto FriasFaculdade de Engenharia 4200-465 PortoUniversidade do Porto PortugalRua Dr. Roberto Frias4200-465 Porto Adrián M.T. SilvaPortugal Laboratory of Separation and Reaction EngineeringNaresh Chandra Murmu Laboratory of Catalysis andSurface Engineering & Tribology Materials (LSRE-LCM)Division Chemical EngineeringCouncil of Scientific and DepartmentIndustrial Research-Central Faculdade de EngenhariaMechanical Engineering Universidade do PortoResearch Institute Rua Dr. Roberto FriasDurgapur 713209 4200-465 PortoIndia Portugal

XII List of Contributors Cyrus Zamani University of Tehran Marko Spasenovic School of Metallurgy and Graphene Tracker Materials Engineering Center for Solid State Physics College of Engineering and New Materials University of Tehran Institute of Physics North Kargar Street Pregrevica 118 P. O. Box 14395-515 11030 Belgrad Tehran Serbia Iran Stephen R. Waite Graphene Stakeholders Association 640 Ellicott Street, Suite 499 Buffalo New York 14203 USA

11Graphene Technology: The Nanomaterials Road AheadStephen R. Waite and Soroush NazarpourA new paradigm is emerging for advanced nanomaterials and their use in com-mercial products. We call it “molecular precision manufacturing” (MPM), and itis evolving as a consequence of the need to develop new tools, new standards, newprotocols, and new processes (TSPPs) to foster the commercialization of nanoma-terials. Nanomaterials possess extraordinary properties, but harnessing these proper-ties for use in commercial products is challenging. The emerging MPM paradigmis required in order to realize the tremendous commercial potential of advancednanomaterials – both 2D and 3D – discovered over the past 25 years. The TSPPs associated with MPM have been in development for several decades.They combine activities that are critical to the use of advanced nanomaterials inproducts and applications: 2D materials, such as graphene, molybdenum disulfide,and boron nitride; and 3D nanomaterials, such as single-wall and multi-wall car-bon nanotubes (CNTs). Additionally, technologies have been developed to func-tionalize these advanced 2D and 3D nanomaterials to enhance their propertiesfor use in commercial products. We are at an early stage in the evolution of func-tionalized advanced 2D and 3D nanomaterials, but the research done thus far isencouraging.1.1Newly Discovered 2D MaterialsThe past decade has witnessed the discovery of several 2D nanomaterials, all ofwhich possess unique properties suited to various applications. These discoveriesinclude the following: Graphene: Single layer of carbon atoms only 1 molecule thick packed in ahexagonal lattice. Molybdenum disulfide (MoS2): When stacked, MoS2 looks and feels likegraphite. However, it is very different from graphene at the 2D level. Whilegraphene is a flat layer of carbon atoms, MoS2 is composed of molybdenumatoms sandwiched between two sulfur atoms. Unlike graphene, in its naturalGraphene Technology: From Laboratory to Fabrication,First Edition. Edited by Soroush Nazarpour and Stephen R. Waite.© 2016 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2016 by Wiley-VCH Verlag GmbH & Co. KGaA.

2 1 Graphene Technology: The Nanomaterials Road Ahead form it can serve as a semiconductor in transistors, making it appealing for use in electronics and solar cells. Scientists have been experimenting with combining the two materials to allow graphene to have transistor-friendly properties, but are now looking at using MoS2 on its own. It has properties similar to silicon, but requires the use of much less material and consumes less energy. Silicene: When silicon is reduced to a 1-atom-thick layer, it takes on a slightly squished-looking honeycomb structure similar to graphene. Like molybdenum disulfide, it can be used as a transistor in its natural form. Silicene also shares one of graphene’s especially interesting properties: electrons move through it at a very fast pace, as if they were massless. This means that silicene conducts electricity faster than any commercially available semiconductor. Because silicon is so ubiq- uitous in current electronics, silicene could be much easier to adopt than other 2D materials. It was only recently synthesized for the first time last year, so the research will take some time to mature. It also could turn out to be more difficult to make than graphene. Germanane: The element germanium has already been used as a semiconduc- tor, and actually formed the very first transistors in the 1940s. When reduced to a single layer of atoms, it forms a material known as germanane. Germanane con- ducts electrons 5 times faster than germanium and 10 times faster than silicon, which makes it ideal for creating faster computer chips. It is more stable than sil- icon and a better absorber and emitter of light. Manufacturers may also be able to produce it on existing equipment in large quantities, which would give it an advantage over emerging graphene manufacturing techniques. Our experience of working with 2D nanomaterials is limited, given their rela- tively recent discovery – in the case of graphene, as recent as 2004. Working with 2D materials presents a set of learning curves that require scaling even before the potential of such promising materials can be realized. The TSPPs associated with the emerging MPM paradigm are critical to the commercialization of products and applications using 2D nanomaterials and their 3D counterparts. Commercialization demands that one has a consistent and repeatable product available at a rational price, given the performance impact and value proposition. Creating the strongest composite in the world is of no value if its mechanical properties cannot be predicted or relied upon because of inconsistent materials or testing. Without these TSPPs, we are not likely to see the fruits anticipated with nanotechnology that many analysts have envisioned, given its vast potential in commercial applications. In the following text, we offer an overview of MPM and shed light on the promises and challenges associated with the emerging MPM paradigm. 1.2 Wonder Materials The ascent of MPM is associated with the discovery of “bulk” nanomaterials pos- sessing remarkable properties. We make the distinction between bulk materials

1.2 Wonder Materials 3and nanoscale elements of electronic and semiconductor devices, for example,which are created as sub-micron architectures using processes such as chemicalvapor deposition and epitaxial growth, but which are not “freestanding” materials. One of the early nanomaterial discoveries came from Rice University in themid 1980s, with the synthesis of fullerenes, commonly referred to as bucky-balls – hollow, spherical carbon structures that became an early impetus toresearch in novel carbon allotropes. The discovery led to more investigation inJapan on hollow tubes of carbon in the early 1990s and ignited great interestin single- and multi-wall CNTs. CNTs were seen to have a host of remarkableproperties that stimulated the interest of nanotechnology researchers all over theworld, and it was not long before patent filings on CNT-based applications beganto skyrocket. In 2004, researchers Andre Geim and Kostantin Novoselov from the Univer-sity of Manchester discovered graphene – another nanomaterial possessing trulyextraordinary properties. In 2010, Geim and Novoselov were awarded the NobelPrize in Physics for their discovery of this “wonder material,” which comprises asingle layer of carbon atoms only 1 molecule thick (hence its 2D classification)and packed in a hexagonal lattice. It is the thinnest material known to man, withan exceptionally high theoretical surface area (2630 m2 g−1). Atomically, it is thestrongest material ever measured, is extremely elastic (stretchable), and has excep-tional thermal and electrical conductivity, making it the substance a design engi-neer’s dreams. Understandably, graphene-related patent filings have risen significantly aroundthe world over the past several years. The United Kingdom is currently a hotbedof activity in graphene, with the University of Manchester acting as a magnet formillions of dollars of research funding. In 2013, the European Union created aFlagship to promote the development of graphene, committing 1 billion Euros infunding over a 10-year time frame. Entrepreneurial activity and investment asso-ciated with graphene has increased significantly. Technology stalwarts Samsungand IBM have been extremely active in patenting graphene-based applications.The Far East has been massively active not only in patent applications, but also ininvestment. Singapore, for example, boasts the highest level of graphene researchfunding as a percentage of GDP in the world. With the discovery of graphene in 2004, we have entered a new age of mate-rials and materials science. Since then, several other 1-atom or 1-molecule-thickcrystals have been isolated and tentatively studied. These materials range fromsemiconducting monolayers to wide-gap insulators to metals. This growing libraryof 2D materials opens the potential to construct various 3D structures with on-demand properties that do not occur naturally, but can be assembled “Lego-style”by stacking individual atomic planes on top of one another in a desired sequence(see Section 1.1). The discovery of new advanced nanomaterials – both 2D and 3D – over thepast 25 years has generated much excitement and hype, which is understandablein light of their remarkable properties. Today, the range of potential applicationsfor graphene and other 2D materials is limited only by one’s imagination. Yet, this

4 1 Graphene Technology: The Nanomaterials Road Ahead potential needs to be tempered by the kind of level-headedness that comes from experience working with advanced nanomaterials. In May 2013, Bayer Material Science (BMS) exited the CNT business and shut- tered its production plant, after many years of work and millions’ worth of invest- ment. BMS CEO Patrick Thomas noted that while the company remains con- vinced that CNTs have huge potential (they initially talked of over 3000 tons of output), their experience suggests that potential areas of application that once seemed promising from a technical standpoint are currently either extremely frag- mented or do not overlap with the company’s core products and spectrum of applications. At the time of exiting the business, it was reported that BMS had invested some $30 million to produce multi-walled CNTs with a facility that had a capacity of producing over 200 tons per year. Mitsubishi Corp. had a similar experience in the 1990s when it attempted to scale and commercialize fullerenes. While no pub- lic information has been made available, insiders indicated that as much as $60 million was invested and, to date, no commercial products realized. While sobering, the BMS experience holds many valuable lessons for those seeking to commercialize advanced nanomaterials. The commercialization of advanced nanomaterials, and nanotechnology in general, is unlike anything ever undertaken before. Successful commercialization of these advanced nanomateri- als requires new approaches, tools, and processes, and a great deal of what seems to be in short supply these days with investors: patience. Often, to satisfy the demands of investors, substantial claims are made on production volumes and estimated sales prior to evaluating the market and without exercising caution. Arriving at a pure material virtually free from the catalysts used in the produc- tion process was not as easy as expected. The challenge was compounded by the need to functionalize these materials; to aid dispersion, acids were often used (as that was all that was available then). High levels of functionalization required a vicious circle of excessive acid treatment, with higher resultant costs, waste streams, and structural degradation. Crucially, the effect of nanomaterials on the target medium is often not known or precisely predicted until it is attempted. Experience shows that taking a process from the lab (micro) level to the commercialization (macro) level is not easy, and in scaling up, the results can often be different from the lab-based results. This will affect commercial outcomes, possibly rendering a positive projected return to an uneconomic position. It is here that we encounter the classic case of over- promising and under-delivery, effectively stunting the market. Having to learn these important lessons the hard way is common in busi- ness – through failed multimillion-dollar investments, layoffs, plant sales, and closure. Yet, it would be foolhardy to extrapolate failures associated with the development of CNTs into the future, for the very success with advanced nanomaterials lies in these failures. Thomas Edison, Nikola Tesla, and Steve Jobs are just a few famous examples of innovators whose failures led to successes beyond their wildest dreams. Fostering a culture of acceptance of failure as a

1.3 The Rise of MPM 5learning process that moves one closer to success is crucial. How often is afailure seen as unacceptable, resulting in management changes that may not bejustified? Failure is instructive, and a large part of the innovation process. That said, it isimportant to respect the meme that insanity is doing the same thing over and overand expecting different results, as Einstein once observed. What we learn from thefailures of working with advanced nanomaterials is that traditional approachesand processes do not work, and something else is required. This is where MPMcomes in.1.3The Rise of MPMHumans have been figuring out how to turn various materials into useful productssince the Stone Age. While some of this knowledge scales into the commer-cialization of nanomaterials today, new learning curves are clearly required tobring advanced nanomaterials to the market in the form of new products andapplications. The BMS experience over the past decade with CNTs is a clear example. Nobodydisputes the theoretical properties of advanced nanomaterials such as CNTs andgraphene. These are well known. As Andrew Geim recently put it: Graphene isdead. Long live graphene! Hundreds of peer-reviewed scientific papers have beenpublished on the properties of graphene and other nanomaterials. The major issueassociated with these materials is not theory and properties, but practice andapplication. How do we turn their fantastic properties into useful and, in somecases, game-changing products? It is clear from the experience of BMS and others that traditional approachesto commercializing these materials are not effective. The emergence of MPM isdue to the shortcomings of these traditional approaches. We know that growth isa function of learning. After all, the cave man had access to all of the materials wehave today. What the cave man did not have was the propensity for learning thatcomes from having experienced failure and success. MPM embraces the learningcurves associated with bringing advanced nanomaterials to the market throughthe development of new processes, standards, tools, and technologies. There is no reason a priori to expect the earlier-described TSPPs associatedwith the successful commercialization of non-nanomaterials to be the same fornanomaterials. It is natural to want to apply the same tried-and-true TSPPs tocommercialize advanced nanomaterials. At the heart of MPM is the developmentof new TSPPs necessary for the proper characterization and functionalization ofadvanced 2D and 3D nanomaterials, together with its effect on the target matrixand down-stream processing. “Characterization” of nanomaterials is critical. Characterization involves theuse of sophisticated metrology tools and information technology that peer

6 1 Graphene Technology: The Nanomaterials Road Ahead down into the nano world and generate data that help us identify the type of nanomaterial being developed for commercialization. Manufacturers today might believe they are working with graphene because their supplier told them it was graphene, when in truth, characterization identifies the material as akin to “soot.” And there is a world of difference between graphene and soot. Knowing the kind of material one is using is paramount to the commercialization process. The way to know what type of material is being used is via characterization analysis. Characterization analysis enables material comparison and is a key component – and the foundation – of the MPM paradigm. A great deal of work is being done today by researchers at the National Phys- ical Laboratory (NPL) in the United Kingdom and elsewhere that is pushing the envelope of characterization analysis. NPL and others are pioneering new tech- niques that allow for more accurate assessment of nanomaterials, and even tools to enable real-time characterization of graphene. New types of metrology tools are being developed to foster characterization analysis of newly discovered 2D nanomaterials. Researchers at Lancaster University (LU) note that scanning probe microscopy (SPM) represents a powerful tool which, in the past three decades, has allowed researchers to investigate material surfaces in unprecedented ways at the nanoscale level. However, SPM has shown very little power of penetra- tion, whereas several nanotechnology applications would require it. The LU researchers are using other tools, such as ultrasonic force microscopy (UFM), in work with graphene and other 2D materials, including MoS2. UFM is a variation of the atomic force microscope (AFM) that overcomes the limitations of SPM in characterizing advanced nanomaterials such as graphene and other 2D materials. These new tools and techniques in development will give manufacturers the important data necessary to ensure that the correct material is being used in the manufacturing process. They also promise to foster quality control in a manner that has not existed previously. As producers in any industry know, quality con- trol is paramount to successful commercialization. Additionally, the creation of sophisticated models to assist in the development, design, and integration of these materials into devices and products relies heavily on the completeness and relia- bility of property data for these nanomaterials. Characterization work also facilitates the development of standards that are critical to the evolution of advanced nanomaterials. The term graphene today covers a family of different materials, including several-layer flakes, powders, liquid dispersions, and graphene oxide. Importantly, the correspond- ing properties and potential applications will vary depending on the type of material used. The other critical part of MPM is dispersion. The ability to consistently and uniformly disperse graphene in another material is important to realizing the outstanding properties of the material. Functionalizing graphene properly can enhance the strength, stiffness, and conductivity of the resulting composites, depending on the requirements and applications being targeted.

1.5 The Nanomaterials Road Ahead 71.4Addressing the Environment, Health, and SafetyAnother important component of the emerging MPM paradigm relates to theenvironmental, health, and safety (EH&S) procedures and protocols for advancednanomaterials. There have been a number of “scare stories” in the media aboutthe potential toxicity of various nanomaterials. Most of these fail to consider thefinal product form that nanomaterials actually take when introduced to the mar-ket, as well as the potential, or lack thereof, of their release into the environmentas nano-sized particles. Without a clear understanding of the full manufacturing cycle, product form,and disposal considerations, the limited information generated by current studiesis of little relevance. Additionally, lacking test standards and precise definitions, itis impossible to conduct credible, repeatable, and scientifically valid studies. Allof the characterization work that is going on behind the scenes with graphene andother 2D materials today is important to future EH&S studies. It is incumbent upon all in the nanomaterials community to collaborate onEH&S-related issues. The new characterization tools and techniques that havebeen developed and are being developed will help facilitate toxicity studies. Thereare groups of researchers today, such as the Arkansas Research Alliance, that areintent on doing credible nontoxicity research on graphene and other nanomateri-als that can be of benefit to all who wish to promote the responsible developmentof such materials. One way to minimize the EH&S effect and aid commercialization is to add thenanomaterials to a carrier in the form of a loaded masterbatch, which is then letdown (diluted) by a processor with the raw, untreated carrier material. This offerscontrollability; and once in a masterbatch, it can be handled without the need forexpensive nano-handling environments.1.5The Nanomaterials Road AheadWe are still at an early stage with the new MPM paradigm. The promise of nano-materials such as graphene and CNTs is great, but so, too, are the challengesassociated with successful commercialization. Several of the key challenges asso-ciated with commercializing nanomaterials-enabled products are being addressedthrough the development of the MPM paradigm. Again, considerable progress hasbeen made, but there is much more work to be done in terms of testing and dataanalysis. Companies seeking to work with graphene and other nanomaterials need toknow the type of materials they are using. Characterization analysis provides thisinformation and also helps to facilitate standards that are necessary for indus-try maturity and EH&S-related research. Additionally, companies need ways ofreliably producing materials to achieve their desired properties. Functionalization

8 1 Graphene Technology: The Nanomaterials Road Ahead assists greatly in this area, for without it, the inert carbon-based material will not want to disperse readily into a target medium. With respect to functionalization, it is also early days, but we see a great deal of potential as functionalization becomes commonplace among those commercializing advanced nanomaterials. It is clear from the lack of progress with CNTs thus far that there is a need for a paradigm such as MPM if we are going to realize the promise and potential of graphene and other nanomaterials. The excitement over these newly discovered nanomaterials is warranted, but again, those seeking to invest and innovate in this promising area need be mindful of the challenges associated with commercializing these materials. Key to progress on the commercialization front is close collaboration among suppliers and pro- ducers and a good deal of patience among all participants involved: the history of materials tells us that it can take years, and sometimes decades, before a new “wonder material” fulfills its promise and potential. Consider the evolution of materials such as aluminum and advanced ceramics. Aluminum was discovered in a lab in the 1820s. Like CNTs and graphene, the material was hailed as a wonder substance, with qualities never seen before in a metal. However, it proved expensive to make, and it was not until many decades later that it took off in the marketplace, when a new process using electricity was invented. Similarly, many of us remember the excitement surrounding advanced ceramics in the early 1980s, and the fever that developed with the discovery of high-temperature ceramic superconductors. The promise of ceramic engines, loss-free electrical transmission lines, and many other products that these material advances were expected to enable has remained unfulfilled. That said, the impact that these materials have had on our lives is nearly impossible to list – ranging from the mundane to the exotic and impacting transportation, communications, electronics, consumer goods, medical devices, and energy in ways that may be hidden but are enabling nonetheless. The road ahead for the development of applications and products using 2D and 3D nanomaterials is filled with tremendous opportunities and key challenges. There is also always a great deal of hype surrounding the discovery of new mate- rials, and experience teaches that hype often turns to disappointment before a wonder material’s potential is eventually fulfilled. In the main, those who earlier tried CNTs and failed remain willing to experiment with the likes of graphene and other nano materials as the desire to get a competitive advantage remains a key economic driver in a very competitive world. The emerging MPM paradigm discussed in this paper seeks to foster the accel- eration of the commercialization process of advanced nanomaterials and promote their responsible development. The TSPPs are designed to avoid corporations from being tempted to reach for instant volume in a desire for market dominance, growth, and profit. The investor community needs to be wary of those who claim volumes that are in the many tons, or hundreds of tons, without proving scale-up as well as process controls to ensure consistent quality production. For those who seek instant “glory,” the bear trap of failure through nonrepeatability looms large.

1.6 Can Graphene Survive the “Disillusionment” Downturn? 9 Despite the great deal of work ahead to realize the potential of these excitingmaterials, and despite some of the setbacks encountered over the past decade,we are encouraged by the progress we are making to bring these next-generation“wonder materials” to the market.1.6Can Graphene Survive the “Disillusionment” Downturn?Even if you are not familiar with the life cycle of emerging technologies, you havecertainly heard about technologies that generated lots of interest at an early stagebut a few years later are gone, having never really entered the marketplace. Many ofthese technologies showed outstanding results in the lab but were unsuccessful inmoving out of the lab into the real world. Most tech companies must pass throughthe ups and downs of their industry’s life cycle, but how they understand and reactto these cycles can make a big difference.1.6.1Gartner’s Hype CycleThe Hype Cycle is a branded graphical tool developed by the research and advi-sory firm Gartner ( for analyzing the maturity and adoption ofemerging technologies. Technology X (a shiny, life changing, and innovative tech) is introduced as thenext big thing (Technology Trigger) and everyone is talking about how it willchange our life (Peak of Inflated Expectations)! Then, as reality sets in, people real-ize that everything has not magically changed and disappointment sets in (Troughof Disillusionment). The shiny, new technology starts to look dull. As time goes by,smart people look at the real opportunities for the shiny new technology (Slopeof Enlightenment) and learn how to build solid businesses with the not-so-shiny-and-new thing (Plateau of Productivity). This is how technology X goes from thelab to the real market (Figure 1.1). The period of time from discovery to maturity is variable and depends on thetype of technology; for instance, it takes around 25–30 years for a new advancedmaterial to move through the cycle. The best recent example is CNT, graphene’ssister material, discovered in 1991. Today, after 23 years, the CNT industry isslowly moving up the “Slope of Enlightenment.” Graphene will pass through a verysimilar cycle, although the cycle time may be slightly faster since many grapheneplayers have learned from CNTs’ hurdles. Graphene was discovered in 2004, and the first generation of graphene pro-ducers, such as XG Science, Angstron Materials, and Vorbeck Materials, hadlaunched and introduced their first generation of products by 2008. Duringearly 2010, large corporations such as BASF (early adopters) showed interestin graphene and began to test first-generation products. Results were oftendisappointing due to problems with graphene dispersion, lack of batch-to-batchconsistency, and the lack of clear graphene standards.

10 1 Graphene Technology: The Nanomaterials Road AheadExpectations On the At the Sliding Into Climbing Entering rise peak the trough the slope the plateau Supplier Activity beyond proliferation early adopters Mass media Nagative press begins High-growth adoption hype begins phase starts: 20% to 30% Supplier consolidation of the potential Early adopters and failures audience has adopted investigate the innovation Second/thrid Frist-generation rounds of Methodologies and best products, high price, venture capital practices developing lots of customization funding needed Less than 5% of Third-generation products,Startup companies the potential audience out of the box, productfirst round of venture has adopted fully suitescapital fundingR&D Second-generation products, some servicesTechnology Peak of inflated Trough of Slope of enlightenment Plateau of trigger productivity expectations disillusionment Time Figure 1.1 Gartner’s hype cycle. In September 2010, Konstantin Novoselov and Andre Geim were delighted to receive the Nobel Prize in Physics for the discovery of graphene. This award resulted in broad media coverage, building up to mass media hype by early 2011. Media hype continues today as governments launch and build support for large science to industry programs. After few years of excitement and buildup, graphene is now “At the Peak” of Gartner’s Hype Cycle and sliding into the “Trough of Disillusionment” is happening. 1.6.2 Surviving the Trough of Disillusionment As scary as the “Trough of Disillusionment” appears, there are a few key strategies that graphene companies can employ to safely move through this stage: 1) Maintain low overheads. Growing too fast and burning cash at the “Peak of Inflated Expectations” stage has killed a lot of businesses. Access to capital becomes much harder as an industry moves into the “Trough of Disillusion- ment.” 2) Concentrate. Graphene companies need to focus on one target market in which their products provide the maximum value to their customers. Trying to chase all opportunities, across multiple industries, increases the burn and reduces the chances of success dramatically.

1.6 Can Graphene Survive the “Disillusionment” Downturn? 113) Be revenue-driven, rather than value-driven. Many potential markets (e.g., bio/health, aerospace) have a huge upside in terms of value, but time-to-market is long and regulatory hurdles exhausting; lots of cash will be burned before significant revenues are made. Unless graphene companies have a strong partner with deep pockets, short-term revenue opportunities must trump long-term, value-driven markets. Investors on the other hand, need to understand that the development of anadvanced material business is a long process and depends heavily on managementstrategy. Companies that can ignore the hype, and grow and generate revenuesduring the initial industry phases, will have the opportunity to create lasting, valu-able businesses. Reviewing the life cycles of other high-tech markets, such as thoseof solar cells and plastics, may help provide investors with key insights.1.6.3Graphene and BatteriesResearchers note that graphene can improve battery attributes like energy densityand form in various ways. Conventional battery electrode materials, as well asprospective ones, can be significantly improved when enhanced with graphene.Graphene’s unique traits, such as mechanical strength, electrical conductivity,large surface area, and lightness of weight can make batteries lighter and moredurable and suitable for high-capacity energy storage. Additionally, graphene canshorten charging times – a highly desirable feature for electric vehicles (EVs)and consumer electronics products. A battery’s lifetime is negatively linked tothe amount of carbon that is coated on the material or added to electrodes toachieve conductivity (Figure 1.2). Graphene adds conductivity without requiringthe amounts of carbon that are used in conventional batteries.1) Graphene, and in particular graphene oxide, has shown to be a valuable mate-rial for overcoming the hardest challenges presented in lithium–sulfur batteries.In summer 2014, researchers from Samsung’s Advanced Institute of Technology(SAIT) announced a novel way to extend the life of a lithium-ion battery (LIB)using a combination of silicon and graphene. SAIT fabricated anode material bygrowing graphene on the surface of silicon without forming silicon carbide. Thenew material has four times the capacity of commercial graphite. Researchers atSAIT note that the approach has the potential to increase the volumetric energydensity of LIBs by 1.8×. Key to the commercialization of this advanced graphene-enabled battery technology is the ability to manufacture carbide-free graphene inmass quantities. The biggest obstacle to realizing the full application of graphenetechnology today is the relatively high cost and low reliability for large-scale pro-duction and manufacturing. There is a great deal of work being done in SouthKorea by Samsung as well as by other researchers inside and outside large estab-lished corporations around the world to address this critical issue.1) For more, see Roni Peleg and Ron Mertens, Graphene Batteries Market Report, 2015.

12 1 Graphene Technology: The Nanomaterials Road Ahead Graphene batteries of the futureEnergy content (Wh kg−1) 400 350 300 NiCd NiMH Traditional Graphene Graphene 250 200 150 100 50 0 Lead acid Li-ion enhanced enhanced Li-ion Li-SFigure 1.2 2-D nanomaterials enable more powerful batteries. (Peleg and Mertens,The Graphene Batteries Market Report, 2015.) LIB is the most important type of rechargeable batteries. In such types of bat-teries, lithium moves back and forth between two electrodes, called cathode andanode, for charging and discharging. LIBs are common in many consumer elec-tronics and electric cars due to their relatively high energy density (the amount ofenergy stored in a unit of battery), low hysteresis (after charging and discharging,there is little loss of energy capacity), and a very slow loss of energy when not inused. LIBs consist of a lithium compound as cathode, spherical graphite as anode,and lithium salt as an electrolyte to allow lithium ion movement between the cath-ode and anode. Increasing the capacity of LIB is dependent upon better materialsfor cathode and anode. It should be noted that the combination of cathode, anode,and electrolyte is one cell; several connected cells are called a module and multiplemodules go together to make up a battery. Recently, news regarding the proposed Tesla battery Gigafactory has had animpact on the industries involved in the LIB supply chain, notably on naturalflake graphite junior miners. A large component of today’s LIBs is graphite and,for the proposed Tesla factory only, more than 300k metric tons/year graphitewould be needed. The news of the proposed Gigafactory has resulted in a boostin the graphite market, but graphite-based anodes are not at all adequate for thebattery performance that is required for EVs by 2030. By that date, most hybridelectric cars will have been converted to full electric cars running completely onbattery power and without any fossil fuel consumption. The replacement materialhas to radically improve the performance of existing batteries to provide longerrun times (a larger storage of energy), faster charge times, all with the smallestpossible weight and at the lowest possible added cost. Furthermore, the new bat-teries need to be long lasting (over 1000 cycles) and thermally stable (should not be

1.6 Can Graphene Survive the “Disillusionment” Downturn? 13over-heated during charging). Graphene is a leading candidate for the replacementmaterial. There are many studies and technical papers showing how graphene canimprove batteries. Its outstanding electrical and thermal conductivity enhancesthe activity of cathodes and prevents over-heating of the batteries. Recentfindings by researchers from Lawrence Berkley lab introduced lithium–sulphurgraphene compounds that generated twice the energy of current batteries andwere stable over 1500 cycles. Such batteries could enable EVs with an efficiencyof more than 500 miles on a single charge, which is what future electric cars need.Newer technologies such as Li–air batteries or supercapacitors could replaceLIBs as well. The future of the energy industry is largely dependent upon improved batteries.Such batteries will change our life drastically. In a matter of few years, gas stationswill be replaced by electric car charging stations and typical auto mechanicswill require new certification to repair electric cars. Further investigation ongraphene-enhanced batteries is absolutely crucial as graphene–silicon com-pounds have proved to be a potential replacement for spherical graphite as anodein LIBs and graphene oxide–sulphur compounds as cathode in lithium sulphurbatteries.1.6.4Heat Management with GrapheneMiniaturization of electronic systems and circuits is heavily restricted with heatdissipation challenges. Heat buildup reduces the efficiency of the electric motors,performance of CPUs, and lifetime of consumer products and batteries. Heat dis-sipation becomes even more challenging when flexibility and bendability of thefinal product is important. Metals are not a suitable candidate anymore and plas-tics are rapidly replacing them as they are cheaper and easier to shape and areweightless. However, plastics severely suffer from lack of thermal conductivity.The possibility of enhancing thermal conductivity of plastics (preferably by keep-ing them electrically insulating) is game changing. Graphene is proved to have thehighest thermal conductivity among all materials. Small loading of well-dispersedgraphene into plastics can enhance their ability to dissipate the heat tremendously.If such a loading level is lower than the percolation threshold, plastics stay insu-lating. Percolation in plastics starts by 0.2–0.5 wt% of graphene addition. Lowerconcentration of graphene is likely to be ineffective to change electrical conduc-tivity. Table 1.1 represents the impact of graphene addition upon the thermalconductivity of thermoplastics and thermosets. Having said this, selection of the optimized graphene loading is crucial. Concen-tration of graphene has to be finely tuned to an optimal value in order to achievethe best results. For instance, in case of poly(lactic acid) (PLA), the optimal con-centration of graphene was found to be 0.075% (Figure 1.3).

14 1 Graphene Technology: The Nanomaterials Road Ahead Table 1.1 NanoXplore graphene improves polymer thermal conductivity and effusivity.Material Thermal Improved Thermal Improved thermal effusivity thermal conductivity at conductivity (Ws0.5 m−2 K−1) effusivity 21–25 ∘C (W m−1 K−1)PLAa) 0.36 245% 714 112%PLA + 0.075 wt% 1.23 44% 1517 55%graphene 339% 142%PEb) 0.74 446% 888 166%PE + 0.1 wt% graphene 1.06 45% 1377 17%ABSc) 0.29ABS + 0.05 wt% 1.28 54% 643 15%graphene 1555Silicone rubber 0.23 80% 33%Silicone 1.24 572rubber + 0.2 wt% 1522graphene 0.382-Part epoxy potting 771compound 0.552-Part epoxy potting 905compound + 0.075 wt% 0.66graphene 1190Silicone heat transfer 1.02compound 1367Silicone heat transfer 0.21compound + 0.1 wt% 0.37 550graphene 730PolyurethanePolyurethane + 0.13 wt%graphenea) PLA stands for Poly(lactic acid).b) PE stands for Polyethylene.c) ABS stands for Acrylonitrile butadiene styrene.1.6.5How Graphene Could Revolutionize 3D PrintingLast year at the International Manufacturing Technology Show (IMTS) inChicago, one of the largest industrial trade shows in the world with more than100 000 visitors, 1900 exhibitors gathered in Chicago to showcase recent develop-ments in machines, tools, and manufacturing systems. Arizona-based automobilemanufacturer “Local Motors” stole the show by printing and assembling an entireautomobile, called the Strati, from scratch and live in front of spectators. Onthe other side of the world, a Chinese company “WinSun Decoration DesignEngineering” recently constructed a set of 10 single story, 3D-printed homesproduced in under 24 h. These homes, based upon cement-based prefabricatedpanels printed on a custom-built 10 x 6.6 m 3D printer, were assembled on site

1.6 Can Graphene Survive the “Disillusionment” Downturn? 17Table 1.2 NanoXplore graphene improves polymer mechanical properties.Material Ultimate tensile Tensile strain strength (MPa) at break (%)Base rubber compound 11.70 540.56Base rubber compound + 0.1 wt% graphene 12.99 663.18Acrylated monomers 19.5 0.71Acrylated monomers + 0.5 wt% graphene 30.6 1.23Additive manufacturing is a game changer for industry.material such as graphene will require significant effort and patience to perfectthe technology and achieve its promise. An example of such improvements are enhancing the mechanical properties ofrubber-based filaments and UV curable filaments. Table 1.2 represents some ofthe improvements. As we can see, there are many exciting opportunities that have arisen alongwith the discovery of graphene and other 2D nanomaterials in the past decade.This book provides an overview of some of the important ongoing research withgraphene and also highlights some of the commercial trends and related issuesassociated with financing companies innovating with 2D materials. The pastdecade has been one of intense research in 2D nanomaterials. As we have seenwith CNTs and other advanced materials, the commercialization cycle extendsout over a decade. Experience informs us that where there is great opportunityfor commercialization with 2D nanomaterials, there are also challenges and risksassociated with creating sustainable business models and successful companies.Based on their remarkable properties and ongoing research and developmenttrends highlighted in this book, we are optimistic about the commercializationprospects of graphene and other 2D nanomaterials-enabled products in the yearsahead.

192Graphene SynthesisSiegfried Eigler2.1IntroductionGraphene is exactly one carbon layer of graphite. The carbon atoms are arranged ina honeycomb lattice. However, the access to graphene is one hurdle for the devel-opment of applications based on graphene. In general, there are two approachesto graphene. One starts from graphite (top-down approach), the other one fromsmall molecules that are used to build up graphene (bottom-up approach). Notonly is the isolation of one layer of graphite a challenge but so also is the stabiliza-tion of the delaminated layers of graphene. If layers of graphene are not stabilizedthey tend to aggregate, forming a sort of graphite. While there is a stacking orderin graphite, restacked graphene exhibits no order of layers. Instead, the flakes ofrestacked graphene are randomly stacked and it seems likely that porous struc-tures will be formed, very similar to crumpled paper piles. The delamination of graphite to graphene can be performed mechanically usingadhesive tape, and mechanically in solvents by sonication, shear mixing, or ballmilling. In addition, the delamination of single layers of graphite can be facilitatedusing chemical methods that involve chemical functionalization and defunction-alization after processing. Another approach to graphene focuses on the synthesis of graphene on solidsusing a carbon source. The method of choice is most often chemical vapor deposi-tion (CVD) using a metal surface and small molecules, such as methane or acety-lene. The challenges in the CVD approach are that the substrate should meet somerequirements: the processing parameters such as temperature, gas mixture, andpartial pressure, as well as the subsequent processing of grown graphene. Untilnow, the synthesis of graphene has been demonstrated by both non-wet-chemicalmethods and chemical methods. However, processing and integrating graphenein either electronic devices or composite materials to obtain high performancematerials by using the unique properties of graphene, such as the high mobility ofcharge carriers, mechanical strength, conductivity, and transparency combinedwith flexibility is another challenge that must be overcome in the future.Graphene Technology: From Laboratory to Fabrication,First Edition. Edited by Soroush Nazarpour and Stephen R. Waite.© 2016 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2016 by Wiley-VCH Verlag GmbH & Co. KGaA.

20 2 Graphene Synthesis The challenge that all methods for the synthesis of graphene have in common is that the quality of graphene differs, which is due to the polydispersity of graphene. Thus, when talking about graphene, not only the synthetic method but also the resulting graphene in terms of the size of graphene, which can vary between nanometers, micrometers, and continuous films must be taken into account. Furthermore, the density and type of defects can vary strongly depending on the type of processing. Although graphene is a single layer, often more than one layer is obtained. In addition, graphene can be chemically modified or functionalized by covalent or non-covalent methods. The degree of functionalization is another polydispersity that should be taken into account when discussing graphene. 2.2 Definitions 2.2.1 Nomenclature and Structure Graphene is exactly one layer of graphite [1]. The C-atoms of graphene are arranged in a honeycomb lattice and the C–C bond distance in graphene is 0.142 nm. In addition, graphene is an infinite plane. However, real graphene exhibits edges that can either be zigzag or arm-chair (Figure 2.1) [2]. Bi-layer graphene consists of two layers of graphene stacked on each other. Gen- erally, it is not distinguished how they are stacked, but in principle they can be stacked as AA or AB. In AA stacked bi-layer graphene, the layers are stacked in a way that carbon atoms are on top of each other. In AB stacked bi-layer graphene, carbon atoms of the second layer are on top of the cavity of a six-membered ring of the first layer. Besides this ideal stacking, the layers can also be twisted to form a moiré pattern that is dependent on the twisting angle (Figure 2.1) [4]. Tri-layer graphene can be formed along these lines and up to 10 layers may be called few- layer graphene. However, as it is often difficult to distinguish between, say, 8 or 10 layers, the term multi-layered graphene can also be used. The properties of more than 10 layers of graphene are similar to those of graphite and therefore, the term graphite is used. However, the layer sequence may also be random, and is termed turbostratic graphite. Restacked graphite or also porous graphite is yielded when layers of graphene or few-layer graphene restack after delamination or par- tial delamination (Figure 2.1). 2.2.2 Polydispersity of Graphene In contrast to the fullerene C60, which is a molecule and a monodisperse material, graphene is a polydisperse material. Even if C60 is contaminated by C70 after synthesis, C60 can easily be purified by chromatography. Such simple procedures are not available for graphene and graphene-based systems. The following

2.2 Definitions 21 Zigzag(a) (c) (b) Arm-chair 600 μm (e) (f)(d)Figure 2.1 (a) Chemical structure motive (HR-TEM) image of one layer of graphene.of AB stacked graphite. (b) Graphene with (Reprinted from Ref. [3] with permissionzigzag and arm-chair edges. (c) Restacked from Macmillan Publishers Ltd: Nature Com-layers of graphene and few-layer graphene. munications, Copyright 2014.) (f ) Selected(d) A flake of natural graphite. (From Ref. moiré pattern of twisted bi-layer graphene.[2] with permission from Wiley-VCH Verlag (Reprinted from Ref. [4] with permission fromGmbH & Co, Copyright 2014.) (e) High res- Macmillan Publishers Ltd: Nature, Copyrightolution transmission emission spectroscope 2011.)equation expresses some of the possible aspects of polydispersity of grapheneand graphene-based systems [2, 5]:S∕s,dGn − (R)f ∕Af (2.1)S is the substrate, s the size of graphene, d the structural defect density ofgraphene within the carbon framework, G the graphene, n the number of layersof graphene, R the addend, f the degree of functionalization, A the non-covalentlybound molecules, and for excluded S, reactions are applied in dispersion. First of all, it is important to distinguish between graphene, few-layer graphene,and more layers. Therefore, graphene (G) is used in Eq. (2.1) with an index nindicating the number of layers. Graphene (G1) is used for graphene, and Gfew-layeror G3–10 indicates few-layer graphene. An infinite number of layers can beindicated either by G∞ or by Gn signifying graphite. Moreover, graphene may bestabilized or deposited on a solid support such as a surface. An often used surfaceis SiO2 grown on a Si wafer. Other surfaces are BN (boron nitride) or that of apolymer. The deposition of graphene on SiO2 can thus be indicated by SiO2/G1.Alternatively, graphene may be deposited on a surface of water, for example byetching copper away after growth of graphene on copper, which is indicated byH2O/G1. Another type of polydispersity is the size of flakes, which can be onthe nanometer-scale, micrometer-scale, or even centimeter-scale. Exemplarily,

22 2 Graphene Synthesis graphene grown on a copper foil of a size of 1 cm × 1 cm can be indicated as Cu/1 cmG1. Flakes of graphene, delaminated from graphite with a size of the flakes of about 5 μm and placed on a SiO2 wafer, are indicated as SiO2/5 μmG1. Another issue is about defects within the carbon framework of C-atoms. This means that the honeycomb lattice is disturbed. The density of these defects can be determined by the analysis of results obtained by Raman spectrascopy. Raman spectroscopy is very sensitive toward defects and the density of defects can be determined in a range of about 0.001% and 1–3%. It should be noted that not all types of defects in graphene can be identified by Raman spectroscopy and therefore, the determined density of defects must be seen only as a rough value [6]. The type of defects and the shape and size of defects are also difficult to determine and these types of polydispersity are currently under investigation in basic research. Structural defects can be due to missing atoms or mismatching arrange- ments of C-atoms, like five-membered rings or seven-membered rings. Raman spectroscopy however, also indicates sp3-hybridized carbon or clusters of sp3-hybridized carbon as defects. Thus, it is possible to determine the degree of functionalization by Raman spectroscopy, especially for graphene with an intact carbon framework [7]. The addition of functional groups to the lattice of graphene will lead to functionalized graphene, which reflects another type of polydispersity of graphene. The regiochemistry of addends on graphene is, until now, almost unexplored and therefore, the regiochemistry is normally not indicated. However, the trans-1,2-addition motive of functional groups is thermodynamically most favored. However, reactions with graphene placed on surfaces can only proceed from the upper side. In the systematic of Eq. (2.1), the degree of functionalization based on the available C-atoms is indicated. For every 20th C-atom being functionalized, for example, by OH, the degree of functionalization is 5% and therefore, is indicated by G1-(OH)5%. Furthermore, few layers of graphene can be exfoliated from graphite in stabilizing solvents such as N-methyl-2-pyrrolidone (NMP), which results in adsorption of NMP on the surface of graphene. Such a type of graphene is indicated by G1/NMP and if the amount of NMP is determined, for example, by elemental combustion analysis (determination of N-content), the mass % content can be indicated by an index (G1/NMP30%). It is of upmost importance to characterize synthesized materials in detail and the results must be indicated in a systematic way. Otherwise, it is difficult for the readers to realize what type of material was exactly produced. Therefore, the expression given in Eq. (2.1) makes it possible to understand the relationship between the structure of graphene and the determined properties. 2.3 Characterization of Graphene by Raman Spectroscopy Raman spectroscopic characterization of graphene is the most reliable method to determine the quality of graphene by identifying the density of defects within

2.3 Characterization of Graphene by Raman Spectroscopy 23 2D G D 2DIntensity (a.u.)GD + D″ 2D′ Frequency (cm−1)2DD ×K K K′ K K′ G D′ D + D″ D+D′ 2D′ 1500 2000 2500 3000 (b) Raman shift (cm−1) Γ(a) E π K K′ 1700 Γ MK 211 Energy (meV)EF Γ ΓΓ 1600 LO 198 186 π K 1500 174 Γ 1400 161 1300 TO 149 136 1200 124 1100 112 1000 TA LA 900 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Phonon wave vector (2π/a2)Figure 2.2 (a) Raman spectra of graphene the G, D, and 2D peak. (Adapted from Ref.(top) and graphene with defects (bottom) [6] with permission from Macmillan Publish-with the most relevant peaks labeled with D, ers Ltd: Nature Nanotechnology, CopyrightG, and 2D. (b) Illustrations of selected excita- 2013.)tion and emission processes responsible forthe C-framework. Raman spectroscopy of graphene can be performed by usingseveral laser wavelengths ranging from the blue to red region of the light spectrum. However, a green laser turned out to be the most popular one. Either 514 or532 nm of excitation is used. The Raman spectrum of graphene displays threemajor peaks that can be used to evaluate the quality of graphene. These peaks arethe G peak, D peak, and the 2D peak (Figure 2.2). The G peak at about 1580 cm−1is an allowed Γ-point phonon emission and its intensity scales with the numberof C-atoms probed by the laser spot. The D peak at about 1345 cm−1 is a resultof phonon emission that is forbidden by selection rules and is only possible dueto the activation by defects. These defects can be those within the carbon frame-work, such as missing atoms or substituted atoms. Alternatively, the transitions ofsp2-carbon to sp3-carbon also display a defect that can be probed by the D peak[7]. Furthermore, the D-peak position is dispersive, which means that its positionshifts with the excitation wavelength of the laser. The second order of the D peak,the so-called 2D peak, is the consequence of an allowed phonon process and the2D peak is the most intense peak in graphene. The shape of the 2D peak can be fit-ted with a single Lorentz function and the full-width at half-maximum (FWHM,Γ) is about 30 cm−1. For bi-layer graphene, the Γ2D > 40 cm−1, which allows dis-tinguishing between graphene and few-layer graphene [8]. As mentioned above,

24 2 Graphene Synthesis the D peak evolves with the introduction of defects. While the intensity of the G peak remains almost constant, the intensity of the D peak rises, which results in an increased ID/IG ratio. The maximum ratio possible is dependent on the exci- tation laser wavelength. For an excitation at 532 nm the ID/IG ratio increases to a maximum of about four. With further increase in the density of defects the ID/IG ratio declines until the ID/IG ratio is of about one. This declining of the ID/IG ratio is accompanied with line broadening. Thus, the ID/IG ratio follows a relation and therefore, an ID/IG ratio of two can be related to two different densities of defects. By taking the Γ of peaks into account it is possible to distinguish between both. By introducing defects, Γ2D increases from 30 cm−1, for a high quality of graphene, to about 50 cm−1, which relates already to a distance of defects of about 3 nm. The Γ2D can further increase to values up to 350 cm−1, which means that the 2D peak is only barely detectable. This also means that the structure of graphene is destroyed and several percentage values of defects are introduced within the carbon frame- work (Figure 2.3) [9, 10]. Intensity (a.u.) (About 3%) (About 0.03%) 1 nm 10 nm DG DG 2D 1500 2000 2500 3000 1500 2000 2500 3000 (a) 0.77% 10 nm (about 0.03%) LD = 2 nm ID IIG = 2.2 D 2D(G′) G 0.12% D′LD = 5 nm ID IIG = 2.9Intensity (a.u.) LD = 7 nm ID IIG = 1.6 0.062% LD = 14 nm ID IIG = 0.6 0.016% 1200 LD = 24 nm ID IIG = 0.2 0.005% (c)(b) 1600 2000 2400 2800 3200 Raman shift (cm−1) Figure 2.3 (a) Raman spectra of graphene [9] with permission from American Chemi- with different densities of defects. Left: cal Society, Copyright (2011).) (c) Illustration defect density about 1–3% (no sharp 2D of an idealized distance pattern of defects of peak); right: density of defects of 0.03%. 10 nm. (From Ref. [2] with permission from (b) Reference Raman spectra of graphene Wiley-VCH Verlag GmbH & Co, Copyright that relate to densities of defects between 2014.) 0.005% and 0.77%. (Reproduced from Ref.

2.3 Characterization of Graphene by Raman Spectroscopy 25 Since the quality of graphene is generally heterogeneous within a given sample,it is not sufficient to measure only one single spot spectrum of graphene.Therefore, scanning Raman microscopy was introduced, which allows thedetermination of the quality of graphene within a complete sample by applyingstatistical methods [7, 8]. Graphene, placed on a surface such as SiO2, allowsscanning Raman spectroscopy. Generally, silicon with 300 nm thick grownSiO2 is used. These substrates make effective Raman measurements possibleand allow visualizing graphene optically [11]. Using an optical microscopemakes it also possible to distinguish between graphene, bi-layer graphene, andfew-layer graphene by the blue contrast. Thus, films of SiO2/G1 can be used forstatistical Raman microscopy (SRM). Generally, an SRM increment is used onthe micrometer-scale. The useful increment depends on the size of the sampleor the size of the flakes of graphene. Moreover, SRM can be used for Ramanmaps displaying properties of a given sample, reflecting a microscopic technique.Therefore, a small increment of scanning is used for high resolution to visualizeinhomogeneity locally. After measurement, the ID/IG ratio of each spectrum andthe Γ values of the D, G, and 2D peak are evaluated. With this information onhand it is possible to illustrate the quality of a given sample of graphene. Thereare several useful representations for illustrating the quality of graphene and oneof them is the plot of the ID/IG ratio against Γ2D as depicted in Figure 2.4 [12, 13]. Graphene Graphene dominated by defects I2D/IG ID/IG 2.0 4.0 2.5 1.5 200 °C 1.0 3.5 Hydrazine 2.0 0.5 Vitamin C 3.0 HI/TFAID/IG 2.5 2.0 1.5 1.5 1.0 LD > 3 nm LD < 3 nm 1.0 0.5 30 40 50 70 100 150 200 250 2 μm 2 μm(a) Γ2D (cm−1) (b)Figure 2.4 (a) Plotted data of the statistical derived from the reaction of C8K (a donorRaman microscopic (SRM) analysis of films of graphite intercalation compound) and 4-graphene with different densities of defects;ID/IG versus Γ2D (full-width at half-maximum tert-butylphenyldiazonium tetrafluoroborate.(FWHM) of the 2D peak). Graphene derivedfrom graphene oxide using different reduc- Local variations of the functionalization areing agents. [12] – published by The RoyalSociety of Chemistry; (b) SRM maps of multi- visualized by plotting I2D/IG or ID/IG versuslayered films of functionalized graphene x,y-positions. (Adapted from Ref. [13] with permission from Macmillan Publishers Ltd: Nature Chemistry, Copyright 2011.)

26 2 Graphene Synthesis 2.4 Epitaxial Growth of Graphene from SiCA common material used in high-power electronics is silicon carbide (SiC). Thegraphitization of hexagonal SiC was already reported in 1961 [14]. In the earlystages of the growth of carbon layers from SiC, randomly stacked graphite wasmainly obtained. Silicon sublimates from SiC at high temperature and layers ofgraphene or even graphite are left behind. In recent years, the method developedand the number of layers of graphene can be controlled by the process parame-ters. The quality of such graphene can be very high, with crystallites approachinghundreds of micrometers in size [15]. Drawbacks of this method are not only the high costs of SiC wafers but also thehigh temperatures of above 1000 ∘C, or even 1500 ∘C that are required. The hightemperatures needed are related to the growth mechanism. Silicon from aboutthree layers of SiC must sublimate from SiC to form one layer of graphene and,for the formation of few layers of graphene, even more layers are needed. It mustbe taken into account that the first carbon layer remains covalently bound to theunderlying SiC and no sp2-network of C-atoms is formed [16]. Therefore, a sec-ond carbon layer must be formed to generate a real layer of graphene. The firstcarbon layer can be seen as a buffer layer. Consequently, Si atoms must becomemobile and because they cannot diffuse through a graphene layer they must dif-fuse to defect sites for sublimation. With respect to diffusion of Si atoms the rateof growth slows down, having control over the formation of layers [17]. High tem-peratures up to 1500 ∘C are necessary to make Si atoms sufficiently mobile andto minimize defect formation. Moreover, the pressure must be adjusted [18]. Itwas shown that treating the covalently bound carbon buffer layer on SiC by H2at 700 ∘C reduces the covalent bonds to form quasi free-standing graphene andhydrogen-terminated SiC [19].The Raman spectra of graphene grown from SiC are less straightforward to mea-sure because there is an overlay of the D and G peak with signals originating fromthe SiC substrate (Figure 2.5b). However, a subtraction of the SiC signals is possibleand the real spectra of graphene can be visualized [21].Moreover, graphene grown on SiC was transferred by separating the graphenefrom the substrate using a process based on a wet-chemical reaction producinggaseous species, such as oxygen from hydrogen peroxide. The oxygen intercalatesbetween the substrate and the graphene, which finally causes the delamination ofgraphene from the substrate. High-resolution transmission electron microscopywas performed on such transferred graphene layers and the honeycomb latticewas well visible, which demonstrates that high quality graphene can be producedfrom SiC [22].The synthesis of graphene from SiC is possible and may indeed be compatiblewith processes in the electronic industry. However, it must be taken into accountthat the process conditions must be very well controlled to generate pure grapheneor bi-layer graphene. Moreover, the size of perfect graphene patches may be lim-ited. In addition, the cost of energy is high and overcoming this problem remains

2.5 Graphene by Chemical-Vapor-Deposition 27 Si-face <1 μm SiC 2LG C Si 1LG Raman intensity (a.u.) SiC SiC G 2D D G′ C-face G3 G2 2LG 1LG(a) G1 1000 1500 2000 2500 3000 ∼3 μm SiC (b) Raman shift (cm−1)200 nm 2 nm(c) (d)Figure 2.5 (a) Illustration of the growth of subtraction, shown for graphene and bi-layergraphene on SiC. (Reproduced from Ref. [20] graphene [21]. (c) High resolution transmis-with permission from Macmillan Publish- sion electron microscope image of grapheneers Ltd: Nature Chemistry, Copyright 2009.) (overview) and (d) magnification that dis-(b) Top: Raman spectra of graphene and bi- plays the intact honeycomb lattice. (Repro-layer graphene on SiC, and bottom: spec- duced from Ref. [22] with permission fromtra obtained after SiC-background spectra The Royal Society of Chemistry.)challenging. Nevertheless, from the fundamental aspect of the study of graphene,bi-layer graphene, the formation of defects, and strain or dislocations, SiC growngraphene and bi-layer graphene remains a very likely preparation method [23, 24].2.5Graphene by Chemical-Vapor-DepositionCVD is one of the most popular methods to provide graphene on surfaces. Thereare many reports on synthetic production conditions and it turns out that thereare some general considerations that have to be taken into account to produce

28 2 Graphene Synthesis graphene on surfaces of high quality. Moreover, there are no generalizing reac- tion conditions available to predict the exact quality of graphene. Nevertheless, graphene growth catalyzed on metal surfaces from small molecules is a non-wet- chemical method to yield a high quality of graphene [25, 26]. Although there are some general obstacles in the preparation of graphene by the CVD approach, there are also some reports that provide methods to over- come the general problems of CVD graphene production. The number of layers of graphene can strongly vary depending on the metal surface used as well as on the preparation conditions [26, 27]. In addition, the growth mechanism changes upon changing the metal. A typical substrate for graphene growth is nickel, which can form nickel carbide at growth temperatures of around 1000 ∘C. The processing temperature and processing time is crucial for making graphene on nickel. Typi- cal conditions involve a pretreatment of the nickel surface at 1000 ∘C in hydrogen for a period of about 1 h. Different types of carbon sources can be used, such as methane, acetylene, or other gases. For methane, a 20 min growth is facilitated with a mixture of argon, hydrogen, and methane. Using this method yields not only graphene, but also few-layers. It is interesting to note that typical cooling rates range between 0.1 and 20 ∘C s−1. Upon cooling, nickel carbide decomposes and carbon segregates on the surface and forms graphene. However, the growth con- ditions must be well controlled to avoid producing excessive amounts of graphite. Beneath the preparation of few-layer graphene there are also grains in polycrys- talline nickel and they determine the size of the single crystalline graphene. When crystals of graphene grow, they will interact with each other and start to form continuous films. A general problem in this regard is the formation of non-six- membered ring structures and thus, grain boundaries are formed that limit the quality of films of graphene. The problem of grain boundaries not only exists for the growth of graphene on nickel, but also on other metals. Even single crystals of nickel have been used for the growth of graphene and it was found that much smoother layers of graphene can be obtained [28] (similar process, as depicted in Figure 2.6). It remains challenging to use nickel as the substrate for graphene production due to the preferred formation of few-layer graphene. Thus, other metals are currentlyCopper oxide 1000 °C, CH4/H2 Copper (b) (c)(a)Figure 2.6 Illustration of the formation pro- (c) patches of graphene merging formingcess of graphene grown on a catalytically grain boundaries. (Reproduced from Ref. [25]active metal surface, such as copper; (a) cop- with permission from The Royal Society ofper substrate with oxides on top; (b) growth Chemistry.)of islands of graphene and grains illustrated;

2.5 Graphene by Chemical-Vapor-Deposition 29more favored for the production of graphene. Many different metal surfaces, suchas copper, platinum, ruthenium, iridium, cobalt, rhenium, or germanium havebeen used. However, the most popular substrate is still copper, since the growthparameters can be adjusted to avoid the growth of few-layer graphene. Typicalcopper foils used are polycrystalline with a thickness of around 25 μm. The growthtemperature is about 1000 ∘C, and a carbon source, such as methane, is used in amixture with hydrogen. While the carbon source dissolves in nickel, the formationmechanism of graphene on copper differs [25, 26]. Copper does not form carbidesand carbon can only barely, if at all, dissolve in copper. The carbon source, suchas methane, reacts on the surface and the C–H bonds are cleaved. However, adirect interaction of methane with copper does not seem to be plausible and theadsorption of CHx species on copper is not feasible as well. In contrast to the directinteraction of methane with copper, the attention was turned to CuOx speciespresent on the surface of copper [29]. Copper forms species such as copper-I-oxide and under growth conditions the CuOx species are mobile on the surface.It was suggested that these oxo-species have a strong influence on the nucleationand growth of graphene on copper. In addition, it was demonstrated that the oxy-gen content in copper can indeed influence the kinetics of the growth of grapheneand that its content in copper is a critical parameter for that control. Polycrys-talline copper exhibits more active grains that can have a stronger interaction withoxygen compared to other surface atoms. Thus, adsorption or chemisorption ofoxygen atoms can passivate the grains of copper. Indeed, the amount of nucleationcenters could be reduced by pretreating copper with oxygen. The nucleation cen-ters were dramatically reduced by oxygen treatment of the copper foils and 0.01nucleation sites per square millimeter was found [30]. This finding resulted in asynthetic strategy allowing the growth of single crystals of graphene with diame-ters of about 1 cm. Moreover, it was reported that the growth of graphene changedfrom hexagonal structures to dendritic structures and the growth was reported toaccelerate due to oxygen species. It was found that the activation energy for thegrowth was reduced from 1.76 to 0.92 eV. It can be expected that C–H dissocia-tion is accelerated and CHx species can be generated at lower activation barriers.Thus, generated CHx species can easily be formed and diffuse rapidly to form C–Cbonds. Especially, the finding that very large crystals of graphene can be formedstrengthens the idea that the grain of the copper foil plays only a minor role. How-ever, the crucial drawback of CVD growth of graphene on copper is the differentthermal expansion coefficient of graphene and copper that causes wrinkling ofgraphene after cooling. These wrinkles also reduce the quality of the producedgraphene (Figure 2.7). Another process and synthetic strategy for the growth of graphene wasreported, which involves germanium as substrate [31]. Drawbacks such as thewrinkling of graphene upon cooling and grain boundaries that evolve whencrystals of graphene merge have been reported to be solved. Overcoming thesecrucial problems of graphene growth are breakthroughs in the technical produc-tion of graphene on surfaces. However, it must be kept in mind that the exactprocessing parameters are very critical in CVD growth and although a process

30 2 Graphene Synthesis(a)(b) (c)Figure 2.7 (a) Illustration of the transfer The Royal Society of Chemistry.) (c) (A):of graphene from the metal catalyst by image of graphene grown on a foil of cop-PMMA coating, metal etching, transfer on per and visualized by scanning electronthe desired substrate, followed by dissolu- microscopy (SEM); (B): magnification SEMtion of the polymer, leaving graphene back image of graphene on copper with wrinkles,on the substrate. (Reprinted from Ref. [26] grains, and steps; (C): transferred graphenewith permission from American Chemical on isolating SiO2/Si wafer; (D): transferredSociety, Copyright 2013.) (b) AFM images graphene on a glass slide. (Reprinted fromof graphene films transferred onto SiO2, Ref. [26] with permission from the Americandisplaying cracks and wrinkles. (Repro- Chemical Society, Copyright 2013.)duced from Ref. [25] with permission fromis described in detail, its reproducibility on other growth systems may need longoptimization. Consequently, reproducibility and full control of the process is atechnological question. The common strategy to grow graphene is based on the formation of isotropicnucleation sites that results in defective grain boundaries. The other one is basedon minimizing nucleation sites to grow single crystals of graphene. However, thisprocess is rather slow and maintaining a low density of nucleation sites is chal-lenging. The novel strategy followed on germanium (110) is based on anisotropicnucleation sites. This implies that graphene on germanium (110) starts to growin a favored direction. When all initial graphene crystals start to grow in thesame direction, the edges of graphene will fit perfectly when crystals start tomerge and therefore, grain boundaries are avoided. While graphene growthon germanium (111) results in isotropic growth, the anisotropic growth wasrealized on germanium (110). In addition, the germanium used was terminatedby hydrogen, which minimizes the interaction of germanium and graphene.Lowering the interaction forces make a transfer easier. Another benefit is due tothe ability to grow germanium (110) on silicon wafers, which minimizes costs.This approach also facilitates transfer of grown graphene to other substrates.Even the used germanium wafer was recycled for another growing process. Since

2.6 Delamination of Graphene from Graphite 31the thermal expansion coefficients of germanium and graphene are very similar,a rather smooth surface of graphene was yielded. The growth was accomplishedbetween 900 and 930 ∘C with methane as carbon source, diluted with 1–2 % ofhydrogen. Although growth methods for the synthesis of graphene have been developed,catalytically active metal surfaces are necessary. However, the desired substrateis mostly not a metal surface, but an insulating surface, such as SiO2 or a poly-mer/transparent polymer, targeting device fabrication. Therefore, it is necessaryto transfer the grown graphene from the metal surface to another one. The strate-gies most often used are either based on etching or based on using a polymer fortransfer. It is well known that graphene on, for example, copper can be suspendedon water using, for example, iron-III-chloride to solute the metal. After the cop-per dissolves, graphene floats on the surface of water and can be transferred ontoanother substrate, such as a silicon wafer with a thermally grown silicon dioxidesurface. Another method uses a polymer film, such as poly-methylmethacrylate(PMMA, Figure 2.7). The graphene binds to the PMMA and can be transferredonto the desired substrate. Subsequently, the PMMA is dissolved with acetone toleave the graphene on the surface. However, there are some drawbacks in bothmethods. The etching process is problematic, because heavy metals must be dis-solved, which is not environmentally friendly. Moreover, the process is not atom-ically economic and metal impurities remain on the graphene as contamination.The PMMA approach leaves polymeric residues on the graphene and removingthem is almost impossible. Another transfer method was reported by using a ther-mal release tape that is able to adsorb graphene and release it by heating to 100 ∘C[32]. This approach was also used for roll-to-roll production of graphene. The quality of CVD graphene is high enough for technical applications, evenwith grain boundaries and therefore, it is a technical question if costs, quality, andscaling of the process can meet the demands necessary for applications, such astransparent and conductive coatings or others. However, for future applications,such as spin electronics or ultra-high frequency applications, the requirements ofthe quality of graphene are absolute, with no defects at all [33].2.6Delamination of Graphene from GraphiteNatural graphite is available in large and almost infinite quantities. The fact thata single layer of graphite is graphene encourages scientists to use graphite asthe initial material to make graphene in large quantities. Surely, it is illusive tomake continuous films of graphene from graphite because of the flake limit. Atbest, films of flakes of graphene can be expected. Nevertheless, for basic researchand applications where diameters on the nanometer- or micrometer-scale areneeded, graphene from graphite is the method of choice. In the following sections,mechanical methods for the production will be highlighted. The advantages anddisadvantages of the approaches will also be discussed.

32 2 Graphene Synthesis 2.6.1 Mechanical Cleavage of Graphite The most prominent method to make graphene was introduced by Novoselov et al. [34]. They not only surprisingly succeeded in the isolation of graphene, but also placed it on a SiO2/Si substrate that allowed the evaluation of the outstanding physical properties of graphene. The method introduced is called the “scotch tape” method. An adhesive tape is used for the exfoliation of graphite down to the sin- gle layer of graphene. The procedure involves repeated pealing of few-layers of graphene from flakes of graphite (Figure 2.8). Then, the adhesive tape is placed on the SiO2/Si wafer and after gently pressing and slowly peeling off the adhesive tape from the wafer it is possible to deposit some flakes of graphene on the surface of the wafer. Flakes of graphene with lateral dimensions of several micrometers can be obtained and identified as graphene by Raman spectroscopy. It should be noted that the source of graphite is important and the right one must be identified with experience. Among graphene, bi-layer graphene and few-layer graphene are also deposited. However, bi-layer graphene and tri-layer graphene are also of interest in research [36, 37]. Graphene can be identified by a light blue contrast visible in an optical microscope if placed on 300 nm SiO2/Si wafer [11]. However, it is manda- tory to identify graphene as a single layer and the quality of graphene must be eval- uated also by Raman spectroscopy [6]. The advantage of peeling graphene off from graphite is due to its high quality. No grain boundaries are present and the lattice is free from defects within a very large area up to the micrometer-scale. Thus, thisSingle-layer graphene 1 μmFigure 2.8 Mechanical cleavage of graphene from graphite using adhesive tape. Grapheneis deposited on to SiO2/Si wafers and identified by optical microscopy and AFM. (Repro-duced from Ref. [35] with permission from The Royal Society of Chemistry.)

Breakage 2.6 Delamination of Graphene from Graphite 33 SI Delamination SI ViscosityFigure 2.9 Illustration of the adaption of Adaption of processing parameters improveprocessing parameters, such as stress inten- the delamination efficiency and reducesity or viscosity of the solvent, using also the formation of defects within the carbonbeads of different size. Optimization of the framework. (Reprinted from Ref. [38] withyield of flakes with a reasonable lateral permission from Springer Science+Businessdimension of hundreds of nanometer of Media, Copyright 2015.)graphene and few-layer graphene is possible.method of preparing flakes of graphene is favored for basic research to evaluatethe fundamental properties of graphene. However, for technical applications, thismethod is not efficient and therefore, other methods have been developed that arescalable (Figure 2.9).2.6.2Liquid Phase Exfoliation of Graphite – Stirred Media MillsScalable methods to prepare graphene are rare [39]. However, a facilitatedapproach is performed using ball milling, a grinding method with graphite ascarbon source. Beads that might consist of either polystyrene or denser oxidessuch as zirconoxide are typical grinding balls [40]. Hurdles in this approach are,on the one hand, avoiding breaking of flakes and, on the other hand, avoidingaggregation. Thus, processing parameters must be well controlled and a stabilizermust be used. ®Nonionic surfactants, such as TWEEN 80 that bear polyether groups areused in an advanced approach to prepare graphene and few-layer graphene [38].The stress intensity and specific energy input introduced by the balls must becontrolled to minimize defect formation and to increase the yield of few-layergraphene and graphene. Recently, it was demonstrated that a low-specific energyinput preferentially yields few-layer graphene, while high-energy input facilitatesbreaking of flakes to produce graphite nanoparticels accompanied with defectformation. Moreover, the solvent viscosity plays a crucial role in the delaminationof graphene flakes from graphite. It was shown that increasing the viscosityfrom 1 to 6 mPa ⋅ s enhances the delamination by shear and friction and reducesin-plane fracture. The dampening of beads in a more viscous medium wassuggested as reason for the more efficient delamination of graphite to grapheneand few-layer graphene. However, even if graphene was not exclusively produced,

34 2 Graphene Synthesis Feed 90 GS170 feed powder d (μm) v (rpm) (mPa s) After processing 80 MK 1Intensity (a.u.)D/G = 0.24 70 2D-FWHM (cm−1)2D-FWHM60 100 1000 1 G-FWHM 77.4 cm−1 50 100 1000 1.29 25.5 cm−1 100 1000 3.29 2D 100 1000 5.98D G D/G = 0.36 100 350 1 G-FWHM 2D-FWHM 100 350 1.29 20.4 cm−1 27.4 cm−1 1500 2000 2500 3000 0.0 0.5 1.0 1.5 2.0 2.5 D/G ratio Wavenumber (cm−1)(a) (b)Figure 2.10 (a) Raman spectra of graphite the processing parameter for the stirred(upper spectrum) and graphene (lower spec- media milling process. Parameters such as,trum). The FWHM (27.4 cm−1) of the 2D size of beads, viscosity or energy input werepeak is characteristic for graphene. (b) Anal- optimized. (Reprinted from Ref. [38] withysis of SRM data illustrating the quality of permission from Springer Science+Businessgraphene and few-layer graphene. The qual- Media, Copyright 2015.)ity of the product increased with improvingit must be kept in mind that few-layer graphene and graphene can be produced inlarge quantities. In addition, scale up to the liter-scale was demonstrated withoutinfluencing the quality of graphene and few-layer graphene. The product obtainedby such methods is heterogeneous and thus, it is important to analyze the numberof layers of graphene as well as the quality of graphene and few-layer graphene.As outlined earlier, statistical Raman microscopy (SRM) is a suitable methodto gain information about the density of defects (quality) by the interpretationof the ID/IG ratio, as well as by analyzing the FWHM of peaks, such as the2D peak at about 2700 cm−1. In Figure 2.10 typical Raman spectra of graphiteand delaminated graphene are shown. The overall quality of graphene, whichis dependent on processing parameters, is screened. A reasonable quality offew-layer graphene was achieved, characterized by the FWHM of the 2D peakbetween 55 and 60 cm−1 with an ID/IG ratio of about 0.3 only. A survey on ball-milling of graphite in solvents, as a method to preparegraphene and few-layer graphene (FLG), was presented in great detail [40]. Thefew-layer graphene yield was thoroughly analyzed by SRM. Surfactant-free FLGsuspensions were prepared by wet-media delamination of graphite in organic sol-vents. Among the investigated solvents, the highest FLG concentration was foundfor NMP in agreement with the lowest van der Waals interaction energy. Thegraphite delamination process in NMP was optimized by varying the delamina-tion tool, the delamination media size, and the stirrer rotation speed. Processingof graphite in NMP using a stirred media mill under optimized process conditions(100 μm ZrO2 beads, stirrer tip speed: 3.4 m ⋅ s−1) yields a suspension of few-layergraphene that is nearly free from graphitic impurities. The few-layer grapheneproduction rate and specific production rate is 73 mg ⋅ h−1 and 0.43 g l−1 ⋅ h−1,

2.6 Delamination of Graphene from Graphite 35respectively, which is higher when compared to ultrasound-assisted graphitedelamination (about 1–33 mg ⋅ h−1 and 2.6 × 10−3 –0.33 g l−1 ⋅ h−1, respectively)at comparable energy costs per gram of produced few-layer graphene. Thus,also from the economical point of, presented method is interesting for devel-oping an industrial few-layer graphene production process. The presentedparameter studies can be also regarded as a guideline for process design. Asexpected, the percentage of few-layer graphene in the products derived fromSRM agrees well with the percentage of few-layer graphene determined fromparticle thickness measured by AFM. Therefore, the distribution of Γ2D, whichis much easier and faster to measure than an AFM flake height statistics, alsogives information about the number of layers per particle. Evaluation of theΓ2D distributions reveals that produced few-layer graphene consists of mainlytri-layers.2.6.3Liquid Phase Exfoliation of Graphite – SonicationThere are manifold examples of sonicating graphite in either organic solventsor mixtures of water with surfactants [39]. However, in all examples, few-layergraphenes are yielded and graphene is only produced in small amounts. More-over, the sonication of graphite in solvents is hard to scale up and long periodsof time amounting to weeks are necessary for reasonable results [41]. Thus, agrowing content of graphite nanoparticles among nanometer-sized few-layergraphene is difficult to avoid. It was demonstrated that few-layer graphenes with flake diameters in therange of 100–500 nm are formed in quite large portions with a minor fractionof graphene. That is due to breaking of graphite flakes as a concurrent processto the delamination process. Density gradient ultracentrifugation is a methodsuitable for the analysis of the number of graphene layers of sonicated samples(Figure 2.11). A major part of such samples consists of few-layer graphene and acertain fraction of real graphene in minor quantities can be identified [43]. Ramanspectra of isolated fractions of graphene prove the presence of the number oflayers [42]. Typical surfactants used for exfoliation are depicted in Figure 2.12. They aregenerally based on the combination of a π-system and other functional groups thatprovide water solubility. In addition, polymers were used yielding water dispersedfew-layer graphene after a first exfoliation step in NMP or o-dichloro benzene[39]. The disadvantage of surfactants is that they are persistent on the surface ofgraphene and few-layer graphene and therefore, attention must be paid to the typeof surfactant used, which may be either a disadvantage or a benefit for the desiredapplication. Solvents of high boiling points were found to be suitable for the exfoliationof graphite with NMP, ortho-dichloro benzene (o-DCB) or dimethylformamide(DMF) as the most prominent ones. In addition, fluorinated aromatic molecules,such as pentafluorobenzonitrile, were used. A feature these solvents have in

36 2 Graphene Synthesis f4 4.5 nm(a) f4 f10 f16 f22 600nm 0.0 nm f28 4.5 nm f16 500nm 0.0 nm 500 nm Height 1.0 f4 (nm) f16 0.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Position (μm) (c)DGU Intensity (a.u.) f4 f10 f16 f22 f28 1500 2000 2500 3000 Raman shift (cm−1)(b) (d)Figure 2.11 (a) TEM image of graphene graphene, bilayer, and few-layer graphene. (c)and few-layer graphene after sonication of Fractions of stabilized graphene, separatedgraphite in N-methylpyrrolidone for 312 h. by DGU, with related AFM images and height(Reprinted from Ref. [41] with permission profiles. (d) Raman spectra of graphene frac-from the American Chemical Society. Copy- tions prepared by DGU, F4 fraction relates toright 2011.) (b) Polydisperse surfactant- graphene. (Reprinted from Ref. [42] with per-stabilized dispersion of graphene and few- mission from the American Chemical Society,layer graphene. Density gradient ultra- Copyright 2009.)centrifugation (DGU) yielded fractions ofcommon is that their surface tensions are close to 40 mJ ⋅ m−2 meeting that ofgraphite and graphene . Therefore, they should be the best candidates for exfoli-ating graphite. A hurdle in using NMP or DMF is their toxicity toward organs.Numerous attempts have been made to produce graphene of high concentration.However, up to now, the commonality among these methods is that few-layergraphene with flakes sizes <1 μm are yielded among nanometer-sized graphite.2.6.4Liquid Phase Exfoliation of Graphite – Shear MixingShear-forces, using a shear mixer are an alternative and scalable method toprepare few-layer graphene and graphene [44]. As shown in this example, shearmixing was used also in combination with surfactant stabilization. In addition,

2.6 Delamination of Graphene from Graphite 37HOOC O3S O3S OH O3S SO3 O3S OH O3S SO31-Pyrencarboxylic acid (PyCA) 1-Pyrensulfonic acid 6,8-Dihydroxy-1,3-pyrene- 1,3,6,8-Pyrene-tetra- sodium-salt (Py-SO3) disulfonic acid sodium salt sulfonic acid tetra- sodium salt (Py-(SO3)4) (Py-(OH)2(SO3)2) Cl Cl HOOC COOH NN HOOC COOH Diazaperproprenium dichloride (DAP) S Coronene tetracarboxylic acid (CTCA) O nm O N n O SS SS Quinquethiophene-terminated Poly[styrene-b-(2-vinylpyridine)] polyethylene glycol (5TN-PEG) (PS-b-P2VP)Figure 2.12 Selection of surfactants suitable to stabilize graphene and few-layer graphene in aqueous solution [39].

38 2 Graphene Synthesis exfoliated graphene and few-layer graphene dispersions were produced in NMP. Although most flakes are few-layer graphene and the lateral dimension of flakes is smaller than 1 μm it is one of the rare examples by which Raman spectroscopy results prove that graphene is yielded by a sharp 2D peak as depicted in Figure 2.13. Nevertheless, other Raman spectra clearly indicate few-layer graphene nature by the broad 2D peak. 2.6.5 Liquid Phase Exfoliation of Graphite Using Smart Surfactants Another non-covalent approach to graphene is facilitated using a smart surfac- tant. In this selected example, a bolaamphiphile is used to delaminate the layers of graphene from graphite [45, 46]. Flakes of natural graphite are used as the car- bon source, and graphene together with few-layer graphene were obtained using the water-soluble perylene as depicted in Figure 2.14. The π-system of the pery- lene can interact with the surface of graphene and the carboxylic acids attached to the perylene provide water-solubility. The flakes of graphene can be identified by Raman spectroscopy. Graphene with a flake size of about 1 μm and a moder- ate defect density of approximately 0.01% was found, as indicated by the D peak. The presence of a moderate amount of defects may be a prerequisite for the suc- cessful delamination. An SRM analysis was also performed, which revealed the polydisperse nature of the sample. 2.6.6 Electrochemical Exfoliation of Graphite The electrochemical exfoliation of graphite electrodes is a scalable method to pro- duce few-layer graphene and it was demonstrated that the density of defects in the generated material is rather low, by an ID/IG ratio of 0.4 [47]. The aim of the work is to produce solution-processable graphene to enable electronic applica- tions. Mainly, about 80% of flakes of G1–3 with a C/O ratio of 12.3 were produced by this electrochemical approach. The proposed mechanism of exfoliation that involves the electrochemical generation of reactive oxygen species that can oxi- dize graphite is shown in Figure 2.15. This oxidation leads to an expansion of layers and sulfate ions can penetrate the interlayer space. Finally, graphene and few-layer graphene that is suitable for further processing is obtained. AFM images reveal graphene with lateral dimensions of about 5 μm. It was demonstrated that large area and highly conductive films of overlapping flakes of graphene can be produced on various substrates and field-effect transistors were produced with a performance of several tens of centimeters squared per volt-second. However, these values are still at least two orders of magnitude lower compared to those of graphene. Further development of a controlled electrochemical approach may not only be scalable, but can also produce graphene as a major species.

2.6 Delamination of Graphene from Graphite 39 Number of objects 16 2D N high N high DG (NMP) (Nac) 12 8 Intensity C–C 123 C–H(b) Rotor Stator 4 (103 cm−1) C–NC–O(a) Rotor/stator (c) 0 5 10 15 20 25 284 288 292 0 Flake thickness, N (j) Binding energy (eV) (d) (i) Intensity N high (NG), ID/IG, C Є content 10 (NG) (NaC, AFM) (NMP) Ci low (NG) (NMP, Raman) D low G N low 1.0 C–C fraction (NMP) t high 200 nm 2600 2800 V low 0.1 ID/IG (NMP) (f) D high (l) V high D D′ 2D t low500 nm 500 nm 1000 2000 3000 Ci high Raman shift (cm–1) N high(e) (g) (h) (k)Figure 2.13 Preparation of graphene and few-layer graphene by shear mixing. Graphene can be identified by Raman spectra even if grapheneis a minor fraction among few-layer graphene. X-ray photoelectron spectroscopy can identify C–N bonds originating from NMP solvent. HR-TEMimages reveal areas of an intact honeycomb lattice. (Reprinted from Ref. [44] with permission from Macmillan Publishers Ltd: Nature Materials,Copyright 2014.)

40 2 Graphene Synthesis HOOC COOH COOH Graphite HOOC HOOC COOH HOOC HN O O O NH COOH HOOC O O O O NH N HOOC HN N O N N HN COOH H O HO HOOC O NH COOH O COOH COOH HOOC HOOC COOH Water-soluble peryleneIntensity (a.u.) 4 4 3 27.3 cm−1 2 1 3 21 1300 1400 1500 1600 2500 2600 2700 2800 Raman shift (cm−1)Figure 2.14 The water soluble bolaam- without the need of sonication or shearphiphile, comprising a perylene moiety forces. (From Ref. [2] with permission fromas well as carboxylic acids, was used to Wiley-VCH Verlag GmbH & Co, Copyrightdelaminate graphene from graphite in water 2014, and Ref. [45], Copyright 2009.)2.7Wet-Chemical Functionalization and Defunctionalization2.7.1Reductive Functionalization of GrapheneThe chemical reduction of graphene, by means of transferring electrons into theconduction band of graphene, is an activation step for the functionalization withelectrophiles [13]. Negatively charged graphene, called graphenide, is present indonor graphite intercalation compounds (GICs). The alkali metal in intercalatedgraphite acts as the electron donor [48, 49]. Stable GICs consist of subunits ofC6Li or C8K. In addition, C2Li is known but can only be formed under high pres-sure [50]. Donor-GICs can be dispersed in solvents such as dimethoxyethane toform few-layers and single layers of graphenide [13, 51]. These graphenides cansubsequently undergo addition reactions with electrophiles like alkylhalides orarylhalides, as illustrated in Figure 2.16. Following this functionalization route, itis possible to functionalize even individual layers of graphene by a C–C couplingreaction. It was demonstrated that graphene derivatives, functionalized by C–Ccoupling on the surface, are thermally stable in a wide range, as outlined below.

2.7 Wet-Chemical Functionalization and Defunctionalization 47preparation and isolation of GrO was described by Brodie in 1855. The productwas termed as graphitic acid. The oxidation of graphite was performed in nitricacid using potassium chlorate as oxidant. The procedure had to be repeated sev-eral times to yield a yellow dispersion of GrO. Chlorine dioxide can accumulateand spontaneous decomposition may occur by this procedure. Staudenmaier opti-mized the method and found a less dangerous and faster two-step protocol. First,the dispersion of graphite is treated in a mixture of sulfuric acid and nitric acidwith potassium chlorate as oxidant. Second, the partially oxidized product is fur-ther oxidized with potassium permanganate to GrO. In 1909 Charpy publisheda further improved synthesis of the desired yellow GrO. Graphite is dispersed insulfuric acid and oxidized by addition of portions of potassium permanganate. Itis described that the temperature must be kept below 45 ∘C not to rapidly over-oxidize the product. Higher temperatures lead to complete decomposition of thereaction mixture. Hummers and Offeman published the scalable synthesis of GrOby a similar method as described by Charpy in the 1950s [57]. This procedure iscalled Hummers’ method and was the most often used protocol for the synthesisof GO within recent years. The yield of the oxidation process generally dependson the type graphite and parameters, such as the flake size and crystallinity ofgraphite, the reaction time, temperature and amount of oxidant or the purity ofthe graphite. The successful preparation of GrO and GO, respectively, is generallyindicated by the transformation of black graphite to yellow GrO. Furthermore, theyield can be increased by pre-oxidizing graphite with persulfate in sulfuric acid.All methods have in common that GrO is yielded as brown to yellow solid afterremoving the solvent and after delamination of GrO to individual layers of GO. Moreover, all the described methods have in common that significant amountsof defects are introduced into the carbon lattice during the synthesis [62]. Thesedefects are permanent defects, due to CO2 formation. The amount of defectsdepends on the exact reaction protocol. Raman spectroscopy can be applied onsingle layers of reduced GO that have been deposited on a solid support, such asa Si-wafer with a grown SiO2 layer with 300 nm thickness. The reduction is donewithin about 15 min using a vapor of hydriodic acid and trifluoroacetic acid at80 ∘C. The density of defects can subsequently be determined by the analysis ofSRM data. The preparation of GO that bears defect densities below 0.1% can be achievedby oxidizing graphite in sulfuric acid with potassium permanganate and keepingthe reaction temperature below 5–10 ∘C. The work-up process must also beaccomplished avoiding local heat higher than 5–10 ∘C to prevent over-oxidation.GO with an almost intact carbon framework is yielded by this method and thedensity of defects within the carbon framework is minimized. Thus, the hon-eycomb lattice is largely intact, as it is in graphene. Therefore, this material canbe called oxo-functionalized graphene, which indicates that the graphene latticeis almost preserved and oxygen functional groups are added to the π-surface.Furthermore, functional groups at edges of flakes or defects play only a minorrole for the chemistry of this material, in contrast to defective GO. The low defectconcentration is important for using GO as a precursor generating graphene.

48 2 Graphene Synthesis Electrical properties are dependent on the chemical structure and graphene derived from oxo-functionalized graphene exhibits charge carrier mobility values in the range of 100–1000 cm2 V−1 s−1. In contrast, graphene derived from GO, also called reduced GO, exhibits a higher defect density and thus, charge carrier mobility values depend on hopping processes and have been determined to roughly 0.1–10 cm2 V−1 s−1 [62]. Structure of GO The structure of GO is highly variable. The absolute structure depends on the preparation conditions. Thus, no molecular equation can be given and no precise structure can be formulated. However, in general, with oxidizing graphite toward GO, oxo-functional groups are introduced. Ideally a hexagonal σ-framework of carbon atoms is the basis. This carbon framework is normally ruptured as outlined above due to over-oxidation. Therefore, the honeycomb lattice is not preserved and σ-hole defects are generated in amounts of about 3%. By controlling the oxi- dation procedure, the loss of carbon can be minimized and average densities of defects of about 0.01% have been demonstrated [62, 66, 74]. Oxygen functional groups decorate both sides of the basal plane in GO. How- ever, the amount and ratio of groups is arbitrary. About every second carbon atom is functionalized, as in part illustrated in Figure 2.21. It seems very likely that patches of sp2-hybridized carbon are part of the structure. The growth of the sp2- patches may increase with time and thus, the initial chemical structure can be interpreted as metastable [75]. However, covalent functional groups are mainly hydroxyl and epoxy groups. If sulfuric acid is used for the preparation of GO, cova- lently bound organosulfate groups will be generated. About 1 organosulfate on 30 carbon atoms has been reported [62]. Among other functional groups present on the basal plane are functional groups at edges of flakes and defect sites. The nature HO HO O O OHO OH OSO3H HO3SO OO OH O OH OH O O OFigure 2.21 Chemical sketch for the illustra- acids, lactol groups, and organosulfatetion of functional groups in GO that bears groups in arbitrary amounts. A defect withdefects. At both sides of the basal plane, a proposed structure that comprises onethere are hydroxyl groups, epoxy groups, and carbonyl group and a hemi-acetal is shownorganosulfate groups. At edges of defects as well. (From Ref. [2] with permission fromand edges of flakes there are possibly Wiley-VCH Verlag GmbH & Co, Copyrighthydroxyl groups, ketones, actetals, carboxylic 2014.)

2.7 Wet-Chemical Functionalization and Defunctionalization 49of these groups is not precisely known, but carboxylic acids, lactol, carbonyl andhydroxyl groups and also organosulfate groups are very likely bound. At pointdefects also acetal structures are plausible to be present. These structure motivesare in part considered in the models of Lerf and Klinowski, Gao et al., and Eigleret al. [2]. The development of these models is mainly based on NMR investigations,thermogravimetric analysis, infrared spectroscopy, and others. Another type ofheterogeneity is due to the lateral dimension of flakes that ranges between tens ofnanometers to hundreds of micrometers. In addition, high-resolution transmis-sion electron microscopy was used to evaluate the structure of GO and preservedregions of graphene like structures with lateral dimensions of about 1 nm wereidentified. About 80% of disordered regions are found, which were correlated tofunctionalized regions [66]. Furthermore, holes with diameters of about 1 nm werealso visualized. GO as Precursor for GrapheneGraphene can be prepared from GO by reduction. The quality of the obtainedgraphene is heterogeneous by means of the density of defects within the carbonframework. Structural defects in graphene are illustrated in Figure 2.23a. Vari-ous methods, including thermal or light driven disproportionation and chemicalreduction using various reductants, have been reported [76]. The reducing agentmay even be incorporated in the carbon lattice, for example for hydrazine. In addi-tion, metal hydrides, hydriodic acid, sulfur containing reducing agents, alcohols,ascorbic acid (AS), sugars, metals in acids, amino acids, plant extracts, and evenmicroorganism were used. Moreover, it was demonstrated that defects in GO limitthe final quality of graphene as well as the type of reducing agent [62]. Thermal disproportionation of GO is accompanied by a mass loss of about 40%up to 300 ∘C and CO2, CO and H2O are mainly formed [62]. The success of thechemical reduction of GO is often evaluated by the C/O ratio according to X-ray photoelectron spectroscopy (XPS). However, a quantitative evaluation is diffi-cult due to impurities and structural defects. Especially for low concentrations ofdefects results obtained by XPS analysis may be disturbed by impurities. Further-more, the C/O ratio cannot account for structural defects such as five-memberedrings beneath seven-membered rings. Such structures can also not be termed asgraphene and they reflect a defect that is detected by Raman spectroscopy. Persis-tent oxygen functional groups may remain at defect sites after reduction; however,quantification is barely possible. The most reliable evaluation of the quality of graphene produced from GO is byusing Raman spectroscopy. SRM is the method of choice for a reliable quantifica-tion of the quality [62]. Investigations indicate that the density of defects can beas low as 0.01% or up to several percent (Figure 2.23b). The chemical reduction of GO placed on a substrate using a potent reduc-tion agent can be highly efficient and termed as quantitative. As illustratedin Figure 2.22, hydriodic acid provides both electrons and protons. Iodide isoxidized and iodine is formed. The electrons can easily reduce GO; that is,oxo-functional groups, such as epoxy groups or hydroxyl groups are protonated.

50 2 Graphene Synthesis 2 HI I2 + 2 e + 2 HO OH ai-GO O OH 4 GHI/TFA SO 3H SO 3 GAS O O 2 GHAI/STFA H ID/IGHO O HO O H 1 3e – H2O 50 100 150 200 250 – H2SO4 G1 Γ2D (cm–1) H, e DD D HO G 2D 2D 2D 35 G G 70 – H2O 50 1500 2500 1500 2500 1500 2500(a) (b) Raman shift (cm–1) Figure 2.22 (a) Proposed general reduc- are efficiently reduced, even if only one side tion mechanism of GO with an almost of oxo-functionalized graphene is accessi- intact carbon framework, also called oxo- ble for the reducing agent. Red, Reduction functionalized graphene. Oxo-groups with HI/TFA from the top; blue, reduction attached to graphene are protonated and with ascorbic acid (AS) from the lower side; reductively cleaved on both sides of the and green, stepwise reduction from both basal plane. (b) Raman spectroscopic stud- sides [77]. ies reveal that both sides of the basal plane Water molecules and organosulfate sulfuric acid are formed and cleaved as good leaving groups. Finally, graphene is yielded. Even if the reduction agent can access only from the top and not from the bottom, because the flakes are placed on a substrate, the reduction proceeds at least as sufficient enough for the reducing agent to access both sides of the basal plane. As proof, AS was added to the subphase in the Langmuir–Blodgett trough to give the reducing agent access to the lower side of GO. Subsequently, the reduced flakes were transferred to a substrate and additionally reduced by HI/TFA: hydriodic acid / trifluoroacedic acid (HI/TFA) vapor that can access only from the top. SRM data reveal that HI/TFA reduction alone is highly sufficient, as depicted in Figure 2.22b. The wet-chemical synthesis of graphene is possible from GO with defects on the %-scale, almost intact GO with an average density of defects of about 0.03%, or oxo-functionalized graphene with an average density of defects of 0.05%. The processes for the controlled oxidation of graphite are illustrated in Figure 2.23b and c. Avoiding local heat during oxidation of graphite in sulfuric acid with potas- sium permanganate is crucial, as is the control of temperature during aqueous work-up. The individualized and almost intact GO can be placed on a surface for reduction to graphene (Figure 2.23c).

2.7 Wet-Chemical Functionalization and Defunctionalization 51 Intensity (a.u.) (About 3%) (About 0.03%) 1 nm 10 nm DG DG 2D 1500 2000 2500 3000 1500 2000 2500 3000(a) Raman shift (cm−1) Graphite oxide H2SO4/ O OH O OH K2S2O8 SO Graphite 1. KMnO4 SO O OH H2SO4 rt OH O SO Graphene < 10 °C OH(b) 2. H2O < 10 °C O OH O OH SO SO 3. H2O2 Graphite O < 10 °C OH O SO OH Delamination OH 2. Processing Graphite sulfate 3. Reduction 1. H2O -H2SO4 Graphene oxide 2. Sonication G1 OH + Reduction G2 OH or oxo-G1/ oxo-Gfew-layer Gfew-layer (c)Figure 2.23 (a) Illustration of the defect of oxo-functionalized graphene with andensity in graphene derived from GO almost intact carbon framework and relatedwith different defect densities; left: σ-hole graphene. (From Ref. [2] with permissiondefects within the carbon framework with from Wiley-VCH Verlag GmbH & Co, Copy-a defect density of several percentage val- right 2014.) (c) Schematic of the synthesis ofues (residual functional groups omitted); oxo-functionalized graphene from graphiteright: graphene from oxo-functionalized sulfate and related graphene. (From Ref.graphene with an almost intact carbon [72] with permission from Wiley-VCH Verlagframework and Raman spectra of derived GmbH & Co, Copyright 2015.)graphene. (b) Schematic of the synthesis An alternative graphene synthesis from oxo-functionalized graphene witha degree of functionalization of about 4% is also possible. The graphite isstirred in sulfuric acid with an oxidant, such as potassium persulfate, to formgraphite sulfate. In graphite sulfate the layers of graphene are positively chargedand during aqueous work-up hydroxyl groups are mainly added. The productcan in part be delaminated to oxo-functionalized graphene and few-layergraphene. After placement on a substrate, graphene can be prepared by reduction(Figure 2.23c). The quality of graphene derived from almost intact GO was determined by SRMand the data are shown in Figure 2.24a. The FWHM of the 2D peak is about50 cm−1 for single layers, which relates to a density of defects of about 3%. Thestatistical analysis by Raman spectroscopy of films of flakes of graphene derived

52 2 Graphene Synthesis Graphene Graphene dominated by defects 4.0 400 Γ2D = 33 cm−14.0 59.2 % 3.0 Optimized 2.0 300 ID/IG = 2.2 ± 0.6 25.0 % reduced Go-n3.5 17.4 % HI/TFA Reduced Go-n 9.5% HI/TFA3.0ID/IG 200 Γ2D = 37.7 ± 11.6 cm−1 ID/IG 1002.5 Counts 02.0 20 40 60 80 Γ2D (cm−1)1.5 1.01.0 0.5 40 50 70 100 150 200 250 0 30 50 100 200 250 30 Γ2D (cm−1) (b) Γ2D (cm−1)(a) Density of defects: 0.3% Density of defects: 0.05% Figure 2.24 SRM data for graphene from (a) Oxo-functionalized graphene derived from almost intact GO prepared by the method graphite sulfate, as illustrated in Figure 2.23d illustrated in Figure 2.23c, data relate to a and the average density of defects can be density of defects of about 0.3% [12]. (b) related to about 0.05% [72]. from oxo-functionalized graphene with a degree of functionalization of about 4% was also determined, and the FWHM of the 2D peak was determined to be about 33 cm−1 and the ID/IG ratio to about 2 in average (Figure 2.24b). These data suggest that the density of defects is roughly only 0.05%. Especially, a value of 33 cm−1 for the 2D peak can be related to graphene. Although progress has been made in recent years, methods must be further optimized to increase the yield and purity of graphene. Moreover, the density of defects should also be further reduced not to show any D peak in Raman spectra. Nevertheless, even if the GO-synthesis that provides a variable density of defects is very young, one can assume that the performance of graphene applications will increase using graphene with controlled amounts of defects. 2.8 Synthesis of Nanographene from Small Molecules It was demonstrated that graphene can be prepared either wet-chemically at high temperature from small molecules or directly from graphite. However, as dis- cussed earlier, the yield of graphene among few-layer graphene and the quality are statistically distributed. Precise control of the size of flakes is difficult, even if separation techniques improve [2, 78]. While the growth of graphene crystals on metal surfaces is improving, precise control of size and especially shape remains an unsolved problem [79]. The outstanding physical properties of graphene are due to the chemical structure of the honeycomb lattice. For example, graphene is a zero band-gap semiconductor [34]. A band-gap can be introduced by narrow- ing the width of a sheet of graphene down to few nanometers (Figure 2.25) [82].

2.8 Synthesis of Nanographene from Small Molecules 53 Nanographene GrapheneGraphene molecules PAHs n 40 nm1 nm Graphene Graphene nanoribbons quantum dots 5 nm 10 nm 100 nm 10 μm 1 nm 1 nm 2 nmFigure 2.25 Illustration of graphene derived of typically micrometer size crystals, at leaststructures ranging from molecules, such as >100 nm in both directions. (From Ref.benzene or hexabenzocorronene and even [80] with permission from Wiley-VCH Ver-larger units, such as C222 graphene like units. lag GmbH & Co, Copyright 2012.) The STMStructures are monodisperse with a diam- images of GQDs modified with permissioneter of up to 5 nm. Graphene nanoribbons from Ref. [81]. (Reprinted from Ref. [82] withof larger size and variable shape; graphene permission from Nature Publishing Group,quantum dots (GQDs) with lateral dimen- Copyright 2010.)sions between 10 and 100 nm and grapheneSuch structures exhibit a high aspect ratio by means of length and width, keep-ing the thickness at the atomic level. It was found that the band-gap relates tothe width of the graphene nanoribbons [83, 84]. Thus, graphene nanoribbons arevery interesting candidates for future applications. However, the precise synthe-sis of graphene of various shapes and sizes that can range from few nanometersto micrometer size remains elusive. In addition, control over the configurationof edges of graphene that can be either zigzag or armchair is highly desired butis still a challenging task. Therefore, chemists facilitate the so-called bottom-upapproach to gain control over these challenges [85]. Precise chemical reactions are used to build defined chemical structures.Examples of such carbon-rich molecules are depicted in Figure 2.25. The sizeof synthetically accessible molecules is on the nanometer-scale. The syntheticapproach for a molecule whose diameter approaches 5 nm is shown in Figure 2.26.The graphene-like core unit is C222 [86]. The precursor molecule is still solublein organic solvents due to its three dimensional structure. Such large moleculescan be synthesized by three steps only. The first step involves a Diels–Alderreaction of the triyne and two equivalents of tetraphenylcyclopentadienone at190 ∘C using diphenyl ether as solvent. Only the outer acetylene units are reactive

54 2 Graphene Synthesis 11 h, RR RR 190 °C + O O2 PH2O B I I+ B O O O [Co2(CO)8]/dioxane RR R R R 5 d, 100 °C RR R Suzuki polymerization Ph AICI3 R= R RR nR Cu(OTf)2 RR R R CS2 FeCI3 / MeNO2 CH2CI2 Ph R RR nR(a) (b)Figure 2.26 (a) Example of the synthesis of the C222 graphene-like unit by the wet-chemical bottom-up approach. The reaction scheme showsa Diels–Alder reaction followed by cyclotrimerization that yields the soluble precursor molecule that can be oxidized in CS2 to the depictedπ-conjugated hydrocarbon [86]. (b) Synthetic bottom-up approach to graphene nanoribbons. Soluble precursors functionalized by halides andboronic esters are C–C coupled to define the nanoribbon structure. Oxidation of the preoriented benzene rings yields π-conjugated nanorib-bons. (Modified from Ref. [87] with permission from the American Chemical Society, Copyright 2008.)

2.8 Synthesis of Nanographene from Small Molecules 55under these reaction conditions. The residual triple bond is still reactive towardcyclotrimerization and, therefore, yields the three dimensional soluble precursormolecule with 37 benzene units. Treating this precursor under oxidative condi-tions with a mixture of aluminum chloride and copper triflate in carbon disulfideaffords a black solid that can be identified as the target molecule with a C222graphene-like unit by mass spectrometry. Chemical strategies are a very promising way toward the controlled synthesisof graphene nanostructures, such as graphene nanoribbons [80, 87]. Graphenenanoribbons are synthetically accessible by the bottom-up approach. Smalland defined precursor molecules are synthesized to enable the Suzuki reaction(Figure 2.26b). Therefore, tetraphenyl-substituted and iodine-activated benzeneis coupled with a hexabenzene derivative bearing boronic esters. These solubleprecursors units define the graphene nanoribbon by repeated coupling to poly-mers. Finally, the polymer is cross-coupled by oxidation using Scholl-reactionconditions to yield the π-conjugated nanoribbon. The bottom-up approach is extremely versatile due to the link to organicchemistry. A set of differently shaped nanoribbons can be prepared, such asthe chevron-like nanoribbons, by using defined precursor molecules [82]. Asdepicted in Figure 2.27a, a chevron structure of nanoribbons is yielded using thesuitable brominated precursor molecule. Here, a hexaphenyl benzene derivativeis used, which bears a brominated biphenyl unit. The linear coupling reaction ofthis precursor defines the final length of the nanoribbon. The oxidation of thatpolymer leads to the π-conjugated nanoribbon. Unfortunately, the molecularweight of such nanoribbons does no longer allow a standard characterization,such as nuclear magnetic resonance spectroscopy, as usually used in organicchemistry. However, the nanoribbons can be visualized with atomic precision byscanning probe microscopy. In another example, a precursor molecule with twocoupling groups and three coupling groups are reacted and as predefined by thechemical structure, a junction is obtained as shown in Figure 2.27b [82]. Othergraphene-like structures, such as porous graphene, can also be synthesized.Cyclohexaphane derivatives can be used for such a polymerization approach.Finally, a porous 2D-polymer is obtained (Figure 2.27) [88]. Until now, there are manifold concepts for the bottom-up approach. However,only selected examples are discussed here to illustrate the basic concepts andstrategies of the current bottom-up approach. The greatest benefit is the atomicprecision of chemical structures and super-structures formed after polymer-ization. In some examples, metal surfaces are necessary to generate the finalπ-conjugated system. Moreover, the final molecules are not soluble and not trans-ferable from one substrate to another. Even if they could be transferred, defectsmust be considered. Another hurdle is due to the poor control of size, especiallyfor polymerized structures, such as graphene nanoribbons. Nevertheless, theprospects of the bottom-up approach are great and will certainly find applicationsin the future.

56 2 Graphene Synthesis I Br 250 °C 1.8AII Br 2 nm CHP 440 °CII 0 I Br (b) N BrBr Br Br I Br DTPA Br I Br 1 nm (c)(a)Figure 2.27 (a) Top: Chemical structure of a cyclohexaphane derivative as precursor for porous graphene (blue, illustrated on graygraphene structure). Bottom: Related high-resolution scanning probe microscope image. (From Ref. [80] with permission fromWiley-VCH Verlag GmbH & Co, Copyright 2012; modified from [88] with permission from the Royal Society of Chemistry. Copyright2008; images reproduced from Ref. [89] with permission from the Royal Society of Chemistry, Copyright 2011).) (b) Left: Synthesisof chevron-type graphene nanoribbons by the polymerization of bis-brominated tetraphenyl triphenylene at 250 ∘C to a prepoly-mer followed by dehydrogenation at 440 ∘C. Right: Scanning probe microscope image visualizing the chevron-type structure. (c)Bottom-up synthesis of a junction using bis-activated and tris-activated precursor molecules and scanning probe microscopy imageof the junction. (Reprinted from Ref. [82] with permission from the Nature Publishing Group, Copyright 2010.)

References 57References 1. Fitzer, E., Kochling, K.H., Boehm, H.P., 11. Blake, P., Hill, E.W., Castro Neto, A.H., and Marsh, H. (1995) Recommended Novoselov, K.S., Jiang, D., Yang, R., terminology for the description of Booth, T.J., and Geim, A.K. (2007) Mak- carbon as a solid (IUPAC Recommen- ing graphene visible. Appl. Phys. Lett., 91 dations 1995). Pure Appl. Chem., 67 (3), (6), 063124. 473 – 506. 12. Eigler, S., Grimm, S., 2. Eigler, S. and Hirsch, A. (2014) Chem- Enzelberger-Heim, M., Müller, P., and istry with graphene and graphene oxide- Hirsch, A. (2013) Graphene oxide: challenges for synthetic chemists. Angew. efficiency of reducing agents. Chem. Chem. Int. Ed., 53 (30), 7720–7738. Commun., 49 (67), 7391–7393. 3. He, K., Lee, G.D., Robertson, A.W., 13. Englert, J.M., Dotzer, C., Yang, G., Yoon, E., and Warner, J.H. (2014) Schmid, M., Papp, C., Gottfried, J.M., Hydrogen-free graphene edges. Nat. Steinrück, H.-P., Spiecker, E., Hauke, F., Commun., 5, 3040. and Hirsch, A. (2011) Covalent bulk functionalization of graphene. Nat. 4. MacDonald, A.H. and Bistritzer, R. Chem., 3 (4), 279–286. (2011) Materials science: graphene moire mystery solved? Nature, 474 (7352), 14. Badami, D.V. (1962) Graphitization of 453 – 454. α-silicon carbide. Nature, 193 (4815), 569 – 570. 5. Koehler, F.M. and Stark, W.J. (2013) Organic synthesis on graphene. Acc. 15. Virojanadara, C., Syvajarvi, M., Chem. Res., 46 (10), 2297–2306. Yakimova, R., Johansson, L.I., Zakharov, A.A., and Balasubramanian, T. (2008) 6. Ferrari, A.C. and Basko, D.M. (2013) Homogeneous large-area graphene layer Raman spectroscopy as a versatile tool growth on 6H-SiC(0001). Phys. Rev. B, for studying the properties of graphene. 78 (24), 245403. Nat. Nanotechnol., 8 (4), 235–246. 16. Emtsev, K.V., Speck, F., Seyller, T., 7. Englert, J.M., Vecera, P., Knirsch, K.C., Ley, L., and Riley, J.D. (2008) Interac- Schafer, R.A., Hauke, F., and Hirsch, A. tion, growth, and ordering of epitaxial (2013) Scanning-Raman-microscopy graphene on SiC{0001} surfaces: a com- for the statistical analysis of covalently parative photoelectron spectroscopy functionalized graphene. ACS Nano, 7 study. Phys. Rev. B, 77 (15), 155303. (6), 5472–5482. 17. Borovikov, V. and Zangwill, A. (2009) 8. Eigler, S., Hof, F., Enzelberger-Heim, M., Step-edge instability during epitaxial Grimm, S., Müller, P., and Hirsch, A. growth of graphene from SiC(0001). (2014) Statistical-Raman-microscopy and Phys. Rev. B, 80 (12), 121406R. atomic-force-microscopy on heteroge- neous graphene obtained after reduction 18. Emtsev, K.V., Bostwick, A., Horn, K., of graphene oxide. J. Phys. Chem. C, 118 Jobst, J., Kellogg, G.L., Ley, L., (14), 7698–7704. McChesney, J.L., Ohta, T., Reshanov, S.A., Rohrl, J., Rotenberg, E., Schmid, 9. Cançado, L.G., Jorio, A., Ferreira, A.K., Waldmann, D., Weber, H.B., and E.H.M., Stavale, F., Achete, C.A., Capaz, Seyller, T. (2009) Towards wafer-size R.B., Moutinho, M.V.O., Lombardo, A., graphene layers by atmospheric pressure Kulmala, T.S., and Ferrari, A.C. (2011) graphitization of silicon carbide. Nat. Quantifying defects in graphene via Mater., 8 (3), 203–207. Raman spectroscopy at different exci- tation energies. Nano Lett., 11 (8), 19. Riedl, C., Coletti, C., Iwasaki, T., 3190 – 3196. Zakharov, A.A., and Starke, U. (2009) Quasi-free-standing epitaxial graphene10. Lucchese, M.M., Stavale, F., Ferreira, on SiC obtained by hydrogen intercala- E.H.M., Vilani, C., Moutinho, M.V.O., tion. Phys. Rev. Lett., 103 (24), 246804. Capaz, R.B., Achete, C.A., and Jorio, A. (2010) Quantifying ion-induced 20. Sutter, P. (2009) Epitaxial graphene: how defects and Raman relaxation length silicon leaves the scene. Nat. Mater., 8 in graphene. Carbon, 48 (5), 1592–1597. (3), 171–172.

58 2 Graphene Synthesis21. Kazakova, O., Panchal, V., and Tan, C., Fallahazad, B., Ramanarayan, H.,Burnett, T. (2013) Epitaxial graphene Magnuson, C.W., Tutuc, E., Yakobson,and graphene-based devices studied by B.I., McCarty, K.F., Zhang, Y.W., Kim, P.,electrical scanning probe microscopy. Hone, J., Colombo, L., and Ruoff, R.S.Crystals, 3 (1), 191–233. (2013) The role of surface oxygen in the22. Gorantla, S., Bachmatiuk, A., Hwang, J., growth of large single-crystal grapheneAlsalman, H.A., Kwak, J.Y., Seyller, T., on copper. Science, 342 (6159), 720–723.Eckert, J., Spencer, M.G., and Rummeli, 31. Lee, J.H., Lee, E.K., Joo, W.J., Jang, Y.,M.H. (2014) A universal transfer route Kim, B.S., Lim, J.Y., Choi, S.H., Ahn,for graphene. Nanoscale, 6 (2), 889–896. S.J., Ahn, J.R., Park, M.H., Yang, C.W.,23. Riedl, C., Coletti, C., and Starke, U. Choi, B.L., Hwang, S.W., and Whang, D.(2010) Structural and electronic proper- (2014) Wafer-scale growth of single-ties of epitaxial graphene on SiC(0 0 0 1): crystal monolayer graphene on reusablea review of growth, characterization, hydrogen-terminated germanium. Sci-transfer doping and hydrogen interca- ence, 344 (6181), 286–289.lation. J. Phys. D: Appl. Phys., 43 (37), 32. Bae, S., Kim, H., Lee, Y., Xu, X., Park,374009. J.-S., Zheng, Y., Balakrishnan, J., Lei, T.,24. Avouris, P. and Dimitrakopoulos, C. Kim, H.R., Song, Y.I., Kim, Y.-J., Kim,(2012) Graphene: synthesis and applica- K.S., Özyilmaz, B., Ahn, J.-H., Hong,tions. Mater. Today, 15 (3), 86–97. B.H., and Iijima, S. (2010) Roll-to-25. Mattevi, C., Kima, H., and roll production of 30-inch grapheneChhowalla, M. (2011) A review of chem- films for transparent electrodes. Nat.ical vapour deposition of graphene Nanotechnol., 5 (8), 574–578.on copper. J. Mater. Chem., 21 (10), 33. Kamalakar, M.V., Groenveld, C.,3324 – 3334. Dankert, A., and Dash, S.P. (2015)26. Zhang, Y., Zhang, L., and Zhou, C. Long distance spin communication in(2013) Review of chemical vapor chemical vapour deposited graphene.deposition of graphene and related Nat. Commun., 6, 6766.applications. Acc. Chem. Res., 46 (10), 34. Novoselov, K.S., Geim, A.K., Morozov,2329 – 2339. S.V., Jiang, D., Zhang, Y., Bubonos,27. Yu, Q., Lian, J., Siriponglert, S., Li, H., S.V., Grigorieva, I.V., and Firsov, A.A.Chen, Y.P., and Pei, S.-S. (2008) (2004) Electric field effect in atomicallyGraphene segregated on Ni surfaces thin carbon films. Science, 306 (5696),and transferred to insulators. Appl. Phys. 666 – 669.Lett., 93 (11), 113103. 35. Yi, M. and Shen, Z. (2015) A review on28. Zhang, Y., Gomez, L., Ishikawa, F.N., mechanical exfoliation for the scalableMadaria, A., Ryu, K., Wang, C., production of graphene. J. Mater. Chem.Badmaev, A., and Zhou, C. (2010) A, 3 (22), 11700–11715.Comparison of graphene growth on 36. Nilsson, J., Castro Neto, A.H., Guinea, F.,single-crystalline and polycrystalline Ni and Peres, N.M.R. (2008) Electronicby chemical vapor deposition. J. Phys. properties of bilayer and multilayerChem. Lett., 1 (20), 3101–3107. graphene. Phys. Rev. B, 78 (4), 045405.29. Blume, R., Kidambi, P.R., Bayer, 37. Novoselov, K.S., McCann, E., Morozov,B.C., Weatherup, R.S., Wang, S.V., Fal’ko, V.I., Katsnelson, M.I.,Z.J., Weinberg, G., Willinger, Zeitler, U., Jiang, D., Schedin, F., andM.G., Greiner, M., Hofmann, S., Geim, A.K. (2006) Unconventional quan-Knop-Gericke, A., and Schlogl, R. (2014) tum Hall effect and Berry’s phase of 2πThe influence of intercalated oxygen on in bilayer graphene. Nat. Phys., 2 (3),the properties of graphene on polycrys- 177 – 180.talline Cu under various environmental 38. Nacken, T.J., Damm, C., Xing, H.,conditions. Phys. Chem. Chem. Phys., 16 Rüger, A., and Peukert, W.(47), 25989–26003. (2015) Determination of quan-30. Hao, Y., Bharathi, M.S., Wang, L., Liu, Y., titative structure–property andChen, H., Nie, S., Wang, X., Chou, H., structure–process relationships for

References 59 graphene production in water. Nano Feng, X., and Müllen, K. (2013) Elec- Res., 5 (71), 57328. trochemically exfoliated graphene as39. Ciesielski, A. and Samori, P. (2014) solution-processable, highly conductive Graphene via sonication assisted liquid- electrodes for organic electronics. ACS phase exfoliation. Chem. Soc. Rev., 43 Nano, 7 (4), 3598–3606. (1), 381–398. 48. Dresselhaus, M.S. and Dresselhaus, G.40. Damm, C., Nacken, T.J., and Peukert, W. (2002) Intercalation compounds of (2015) Quantitative evaluation of delam- graphite. Adv. Phys., 51 (1), 1–186. ination of graphite by wet media milling. 49. Enoki, T., Suzuki, M., and Endo, M. Carbon, 81 (1), 284–294. (2003) Graphite Intercalation Com-41. Khan, U., Porwal, H., O’Neill, A., pounds and Applications, Oxford Nawaz, K., May, P., and Coleman, J.N. University Press, Oxford. (2011) Solvent-exfoliated graphene at 50. Nalimova, V.A., Guérard, D., extremely high concentration. Langmuir, Lelaurain, M., and Fateev, O.V. (1995) 27 (15), 9077–9082. X-ray investigation of highly saturated42. Green, A.A. and Hersam, M.C. (2009) Li-graphite intercalation compound. Solution phase production of graphene Carbon, 33 (2), 177–181. with controlled thickness via density 51. Vallés, C., Drummond, C., Saadaoui, H., differentiation. Nano Lett., 9 (12), Furtado, C.A., He, M., Roubeau, O., 4031 – 4036. Ortolani, L., Monthioux, M., and43. Green, A.A. and Hersam, M.C. (2010) Pénicaud, A. (2008) Solutions of neg- Emerging methods for producing atively charged graphene sheets and monodisperse graphene dispersions. ribbons. J. Am. Chem. Soc., 130 (47), J. Phys. Chem. Lett., 1 (2), 544–549. 15802 – 15804.44. Paton, K.R., Varrla, E., Backes, C., 52. Vecera, P., Edelthalhammer, K., Smith, R.J., Khan, U., O’Neill, A., Hauke, F., and Hirsch, A. (2014) Reduc- Boland, C., Lotya, M., Istrate, O.M., tive arylation of graphene: insights into a King, P., Higgins, T., Barwich, S., reversible carbon allotrope functionaliza- May, P., Puczkarski, P., Ahmed, I., tion reaction. Phys. Status Solidi B, 251 Moebius, M., Pettersson, H., Long, E., (12), 2536–2540. Coelho, J., O’Brien, S.E., McGuire, 53. Englert, J.M., Knirsch, K.C., Dotzer, C., E.K., Sanchez, B.M., Duesberg, Butz, B., Hauke, F., Spiecker, E., and G.S., McEvoy, N., Pennycook, T.J., Hirsch, A. (2012) Functionalization of Downing, C., Crossley, A., Nicolosi, V., graphene by electrophilic alkylation of and Coleman, J.N. (2014) Scalable pro- reduced graphite. Chem. Commun., 48 duction of large quantities of defect-free (41), 5025–5027. few-layer graphene by shear exfoliation 54. Schäfer, R.A., Englert, J.M., Wehrfritz, P., in liquids. Nat. Mater., 13 (6), 624–630. Bauer, W., Hauke, F., Seyller, T., and45. Englert, J.M., Röhrl, J., Schmidt, Hirsch, A. (2013) On the way to C.D., Graupner, R., Hundhausen, M., graphane-pronounced fluorescence of Hauke, F., and Hirsch, A. (2009) polyhydrogenated graphene. Angew. Soluble graphene: generation of aque- Chem. Int. Ed., 52 (2), 754–757. ous graphene solutions aided by a 55. Pumera, M. and Wong, C.H. (2013) perylenebisimide-based bolaamphiphile. Graphane and hydrogenated graphene. Adv. Mater., 21 (42), 4265–4269. Chem. Soc. Rev., 42 (14), 5987–95.46. Kozhemyakina, N.V., Englert, J.M., 56. Ryu, S., Han, M.Y., Maultzsch, J., Heinz, Yang, G., Spiecker, E., Schmidt, C.D., T.F., Kim, P., Steigerwald, M.L., and Hauke, F., and Hirsch, A. (2010) Brus, L.E. (2008) Reversible basal plane Non-covalent chemistry of graphene: hydrogenation of graphene. Nano Lett., 8 electronic communication with den- (12), 4597–4602. dronized perylene bisimides. Adv. 57. Eigler, S. (2011) Transparent and electri- Mater., 22 (48), 5483–5487. cally conductive films from chemically47. Parvez, K., Li, R., Puniredd, S.R., derived graphene, in Physics and Appli- Hernandez, Y., Hinkel, F., Wang, S., cations of Graphene—Experiments

60 2 Graphene Synthesis(ed S. Mikhailov), InTech, New York, 69. Chung, C., Kim, Y.K., Shin, D., Ryoo,pp. 109–130. S.R., Hong, B.H., and Min, D.H. (2013)58. Eigler, S. (2014) Graphene oxide, in Biomedical applications of graphene andEncyclopedia of Polymeric Nano- graphene oxide. Acc. Chem. Res., 46 (10),materials (eds S. Kobayashi and K. 2211 – 2224.Müllen), Springer, Berlin, Heidelberg. 70. Feng, L., Wu, L., and Qu, X. (2013) Newdoi: 10.1007/978-3-642-36199-9_333-1 horizons for diagnostics and therapeutic59. Dimiev, A., Kosynkin, D.V., Alemany, applications of graphene and grapheneL.B., Chaguine, P., and Tour, J.M. (2012) oxide. Adv. Mater., 25 (2), 168–186.Pristine graphite oxide. J. Am. Chem. 71. Eigler, S., Dotzer, C., and Hirsch, A.Soc., 134 (5), 2815–2822. (2012) Visualization of defect densities60. Eigler, S., Enzelberger-Heim, M., in reduced graphene oxide. Carbon, 50Grimm, S., Hofmann, P., Kroener, W., (10), 3666–3673.Geworski, A., Dotzer, C., Rockert, M., 72. Eigler, S. (2015) Graphite sulphate—a precursor to graphene. Chem. Commun.,Xiao, J., Papp, C., Lytken, O., Steinrück, 51 (15), 3162–3165.H.P., Müller, P., and Hirsch, A. (2013) 73. Schafhaeutl, C. (1840) Ueber dieWet chemical synthesis of graphene. Verbindungen des Kohlenstoffes mitAdv. Mater., 25 (26), 3583–3587. Silicium, Eisen und andern Metallen,61. Dreyer, D.R., Park, S., Bielawski, C.W., welche die verschiedenen Gattungenand Ruoff, R.S. (2010) The chemistry of von Gusseisen, Stahl und Schmiedeeisengraphene oxide. Chem. Soc. Rev., 39 (1), bilden. J. Prakt. Chem., 21 (1), 129–157.228 – 240.62. Eigler, S. and Hirsch, A. (2014) Chem- 74. Georgakilas, V., Otyepka, M., Bourlinos, A.B., Chandra, V., Kim, N., Kemp, K.C.,istry with graphene and graphene Hobza, P., Zboril, R., and Kim, K.S. oxide—challenges for synthetic chemists. 75. (2012) Functionalization of graphene: Angew. Chem. Int. Ed., 53, 7720–7738. covalent and non-covalent approaches,63. Novoselov, K.S., Fal’ko, V.I., derivatives and applications. Chem. Rev., Colombo, L., Gellert, P.R., Schwab, 112 (11), 6156–6214. M.G., and Kim, K. (2012) A roadmap for Kim, S., Zhou, S., Hu, Y., Acik, M., graphene. Nature, 490 (7419), 192–200. Chabal, Y.J., Berger, C., de Heer, W.,64. Chang, H. and Wu, H. (2013) Graphene- Bongiorno, A., and Riedo, E. (2012)based nanomaterials: synthesis, prop- Room-temperature metastability of mul-erties, and optical and optoelectronic tilayer graphene oxide films. Nat. Mater.,applications. Adv. Funct. Mater., 23 (16), 11 (6), 544–549.1984 – 1997. 76. Chua, C.K. and Pumera, M. (2014)65. Dai, L. (2013) Functionalization of Chemical reduction of graphene oxide:graphene for efficient energy conversion a synthetic chemistry viewpoint. Chem.and storage. Acc. Chem. Res., 46 (1), Soc. Rev., 43 (1), 291–312.31 – 42. 77. Eigler, S. (2014) Mechanistic insights66. Chen, D., Feng, H., and Li, J. (2012) into the reduction of graphene oxideGraphene oxide: preparation, functional- addressing its surfaces. Phys. Chem.ization, and electrochemical applications. Chem. Phys., 16 (37), 19832–19835.Chem. Rev., 112 (11), 6027–6053. 78. Walter, J., Nacken, T.J., Damm, C.,67. Sun, X., Sun, H., Li, H., and Peng, H. Thajudeen, T., Eigler, S., and Peukert, W.(2013) Developing polymer com- (2015) Determination of the lateralposite materials: carbon nanotubes dimension of graphene oxide nanosheetsor graphene? Adv. Mater., 25 (37), using analytical ultracentrifugation.5153 – 5176. Small, 11 (7), 814–825.68. Lü, K., Zhao, G., and Wang, X. (2012) A 79. Luo, Z., Kim, S., Kawamoto, N., Rappe,brief review of graphene-based material A.M., and Johnson, A.T. (2011) Growthsynthesis and its application in environ- mechanism of hexagonal-shape graphenemental pollution management. Chin. Sci. flakes with zigzag edges. ACS Nano, 5Bull., 57 (11), 1223–1234. (11), 9154–9160.

References 6180. Chen, L., Hernandez, Y., Feng, X., and behind nanoscience. ACS Nano, 8 (7),Müllen, K. (2012) From nanographene 6531 – 6541.and graphene nanoribbons to graphene 86. Iyer, V.S., Wehmeier, M., Brand, J.D.,sheets: chemical synthesis. Angew. Chem. Keegstra, M.A., and Müllen, K. (1997)Int. Ed., 51 (31), 7640–7654. From hexa-peri-hexabenzocoronene to81. Lu, J., Yeo, P.S., Gan, C.K., Wu, P., and “superacenes”. Angew. Chem. Int. Ed., 36Loh, K.P. (2011) Transforming C60 (15), 1604–1607.molecules into graphene quantum dots. 87. Yang, X., Dou, X., Rouhanipour, A.,Nat. Nanotechnol., 6 (4), 247–252. Zhi, L., Rader, H.J., and Müllen, K.82. Cai, J., Ruffieux, P., Jaafar, R., Bieri, M., (2008) Two-dimensional grapheneBraun, T., Blankenburg, S., Muoth, M., nanoribbons. J. Am. Chem. Soc., 130Seitsonen, A.P., Saleh, M., Feng, X., (13), 4216–4217.Müllen, K., and Fasel, R. (2010) Atom- 88. Bieri, M., Treier, M., Cai, J.,ically precise bottom-up fabrication Aït-Mansour, K., Ruffieux, P.,of graphene nanoribbons. Nature, 466 Gröning, O., Gröning, P., Kastler, M.,(7305), 470–472. Rieger, R., Feng, X., Müllen, K., and83. Son, Y.-W., Cohen, M.L., and Louie, S.G. Fasel, R. (2009) Porous graphenes: two-(2006) Energy gaps in graphene nanorib- dimensional polymer synthesis withbons. Phys. Rev. Lett., 97 (21), 216803. atomic precision. Chem. Commun., 45,84. Han, M.Y., Özyilmaz, B., Zhang, Y., and 6919 – 6921.Kim, P. (2007) Energy band-gap engi- 89. Bieri, M., Blankenburg, S., Kivala, M.,neering of graphene nanoribbons. Phys. Pignedoli, C.A., Ruffieux, P., Müllen, K.,Rev. Lett., 98 (20), 206805. and Fasel, R. (2011) Surface-85. Müllen, K. (2014) Evolution of graphene supported 2D heterotriangulenemolecules: structural and func- polymers. Chem. Commun., 47 (37),tional complexity as driving forces 10239 – 10241.

633Graphene CompositesSuman Chhetri, Tapas Kuila, and Naresh Chandra Murmu3.1IntroductionNanomaterials form the core of modern science and technology and, as compactand light-weight materials, are in high demand in fields as diverse as aerospaceto automotive, electronics to energy, environment to biomedical, and so on [1].However, the design and development of efficient and high-performance materi-als is challenging as it requires understanding of the property, morphology, anddimension of existing materials as well as of the synthesis of new materials fromit, with improved properties. Therefore, in order to cope up with the demand ofmodern day life, it is imperative to fabricate novel materials having possible appli-cations. Thus, tailoring of a new material is very crucial to enhance the intrinsicproperties of existing materials and to make it applicable to a wide spectrum offields without compromising on environmental sensitivities. Of late, combining various types of materials to achieve synergy in their prop-erties has become the common practise in the realm of material science and tech-nology. Polymeric materials have always been used as one component because oftheir easy availability in various scaffolds, low cost, easy fabrication, and broadareas of applicability. The integration of the second phase called filler or addi-tives and the polymeric matrix results in the formation of a new material knownas polymer composites [2]. If the filler dimension is in the range of a nanoscale,then the resultant material formed is called polymer nanocomposites [3]. Theabove described filler can be of zero-dimensional fullerenes, 1D carbon nanotube(CNT), 2D layered silicate or graphene, and 3D graphite [4–7]. The field polymernanocomposites have been considered as promising materials in the paradigm ofnanoscience and technology. The earliest research on nanocomposites incorporating layered silicate as fillerdates back to 1950 [8]; however, it could not create much hype in the scientific andindustrial realm until Toyota Motor Corporation came up with their harbingerwork in this area where they reported the enhancement in the mechanical prop-erty of Nylon-6 using montmorillonite as fillers [9]. This work stimulated plethoraof research in the area of polymer composites by incorporating nano-dimensionGraphene Technology: From Laboratory to Fabrication,First Edition. Edited by Soroush Nazarpour and Stephen R. Waite.© 2016 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2016 by Wiley-VCH Verlag GmbH & Co. KGaA.

64 3 Graphene Composites filler of different nature, geometry and size, to achieve high performance and multifunctional material. Conventional polymer composites usually employed discrete fillers such as glass fibre, metal particle, wood sheets, and so on [10]. Only modest improvement in various properties was observed in conventional polymer composites due to lower surface area of filler particles, mismatch in the properties of the individual components, incompatibility, and poor interfacial interaction/adhesion and difference in coefficients in thermal expansion [11]. For many years, nanoadditives such as clay or LDHs (synthetic clays), expanded graphite (EG), carbon nanofibre (CNF), CNT, carbon black, metal nanoparticles, and so on have remained candidates for vibrant fields of research due to their high surface area, good compatibility with matrix, and ability to tailor their surface to control the interaction [12–17]. The developed materials showed enhanced physicochemical properties and proved to be superior to conventional polymer composites [18–22]. The underlying principle behind the preparation of polymer nanocomposites using the above mentioned fillers is to couple the intrinsic potential of multiple components and to enhance the performance of the resulting materials. Further, small loadings of nanofillers are more effective in enhancing the performance of composites as compared to nanostructured filler, where its higher content makes the material brittle and heavy [23]. Owing to their layered structure and low cost, clay nanominerals (montmorillonite, MMT) are very effective in increasing the mechanical strength of the composites, but their low thermal electrical conduc- tivity and stiff nature make it a less favorite for composite fabrication [24, 25]. Though, metal nanoparticles proved effective filler and it impart conductivity to the matrix otherwise insulator but its poor compatibility with the matrix make it less viable. Conductive carbonaceous nanomaterials such as EG, CNT, CNF, and carbon black are considered as favorable alternatives as filler material [26–29]. Among them, CNT nanofiller dominated the area as a promising candidate and is still seen as a potential filler because of its superb mechanical and thermal prop- erties, low density, and high aspect ratio [10]. However, poor dispersion in and interaction with polymer matrix and high cost of production inhibited its candi- dature in the development of polymer composites [30]. The discovery of graphene in 2004 showed a new horizon in the path toward the development of advanced multifunctional composite materials [31, 32]. Graphene, one-atomic-thick sheet of sp [2] bonded carbon arranged in a two- dimensional lattice has drawn lot of interest from fields such as physics, chemistry, materials science, and medical science owing to its intriguing and exciting prop- erties such as high surface area; large aspect ratio; outstanding mechanical, electrical, thermal properties; flexibility; high transparency; and optoelectronic properties to name a few [33–36]. Its high surface-to-volume ratio and light weight makes it superior to CNT in the manufacture of light weight material for aerospace and automotive applications. Thermal and electrical properties of graphene are higher than those of CNT albeit their mechanical properties are comparable. Further, graphite is the precursor of graphene and its derivatives; so, cost of production of graphene is low as compared to other carbon-based fillers.

3.2 Preparation and Properties of Graphene 65Graphene as well as its derivatives has the potential to modify the properties andeven has the ability to induce new properties to the matrix. Graphene/polymercomposites show improved mechanical, thermal, barrier, and flame retardantproperties [32, 37–40]. Hence, ever since its inception, graphene has been seen asa versatile candidate for the fabrication of polymer nanocomposites. However, dif-ficulties lie in transforming or imparting the unique properties of graphene to thepolymer. To get the utmost benefit of a filler in terms of properties, homogeneousdispersion of the filler in the matrix is most crucial with good interfacial interac-tion and adhesion between the phases [41]. Degree of enhancement in the prop-erties of the resultant composites depends upon the extent of interfacial bondingbetween the host matrix and the filler. Individual graphene sheets show poor com-patibility with the polymer, and therefore, it is crucial to exploit its extraordinaryproperties to make graphene compatible with and dispersible in the host matrix.Functionalization of graphene not only makes graphene compatible with thematrix but also increases its affinity toward the matrix and adhesion/interactionbecomes strong [42]. Both covalent and non-covalent surface modification of GOhave been investigated in detail using organic and inorganic surface-modifyingagents. This chapter tries to address a wide range of thermoplastic and thermoset poly-mers, biodegradable polymers, and so on Graphene nanoplatelets (GnPs), chem-ically functionalized graphene (CFG), GO, functionalized graphene oxide (FGO),thermally reduced graphene oxide (TrGO) and pristine graphene-incorporatedpolymer composites have been highlighted in detail. The improved physiochemi-cal properties of the graphene/polymer composites including their possible areasof applications have also been covered.3.2Preparation and Properties of GrapheneThe emergence of graphene in 2004 mesmerized the scientific community becauseof its unique properties that could be exploited to generate new class of materials.Although graphene was first isolated by mechanical exfoliation of graphite [31],this method could not generate the amount of graphene required for various possi-ble applications. The micromechanical exfoliation technique employed repeatedlypeeling off pyrolytic graphite using scotch tape. The other synthetic methods thatcan supply graphene required for various applications are chemical vapor depo-sition (CVD) (also known as “bottom-up”), epitaxial growth, plasma enhancedCVD, unzipping of CNT, solution-based reduction of GO, and so on [31, 43–48].CVD is one of the promising methods to synthesize mono- or few-layer graphenesheets using metal foil at over 1000 oC [49]. These preparative methods producehighly crystalline graphene sheets that are suitable for studying the fundamentalproperties of graphene but are not viable for composite preparation as it demandslarge amount of fillers. Moreover, CVD grown graphene is not compatible with theorganic polymer matrix. Thus, for composite fabrication purpose, preparation of

66 3 Graphene Composites graphene from GO is most appropriate as its precursor graphite is abundantly available and can be engineered for scaled-up production. The route followed is oxidation of graphite to graphite oxide, which is exfoliated to GO layers and subsequently converted to RGO (reduced GO) sheets. Although, this is also not without flaws, reduction might not preserve all the intrinsic properties of pristine graphene. Modified Hummer’s method is the most common method to prepare GO from graphite [50]. Most of the procedures to prepare graphene/polymer composites and graphene-like materials have used GO as the starting material. The oxidation process introduces various oxygen functionalities such as epoxide, hydroxyl, diol, ketone on the basal plane and carboxyl, and carbonyl on the edges of the surface and electronic conjugation of graphite gets disturbed/disconnected [51–53]. To generate RGO, typically exfoliation of graphite oxide to GO and its reduction by chemical, thermal, electrochemical, and so on are carried out [54]. The materials so formed are predicated to structurally resemble pristine graphene and said to have conserved its unique properties. Graphene, an allotrope of carbon has the thickness of 0.35–1 nm [55]. The Young’s modulus and Poisson’s ratio for graphene are 1.02 TPa and 0.149, respec- tively [56]. Electrical and thermal conductivity of graphene is 6 × 105 S m−1 and 5.1 × 103 W m K−1, respectively [36,57]. The theoretical intrinsic carrier mobility (200,000 cm2 V−1) and specific surface area (2600 m2 g−1) of graphene are also very high as compared to the traditional reinforcing fillers [32, 58]. High aspect ratios and good mechanical flexibility make graphene an interesting material among the carbon allotrope members. Graphene being an aromatic moiety, it undergoes electrophonic reaction other than nucleophilic and cyclo-addition click reaction and so on [59, 60]. A reaction on the surfaces of graphene destroys the graphitic planar structure, disconnecting sp2 network that eventually hampers electrical conductivity. 3.3 Functionalization of Graphene The reinforcement ability of filler to matrix basically depends upon the extent of its dispersion and its effectiveness in transferring load at the interface. Property enhancement of composites only becomes pragmatic when the filler is uniformly distributed and is able to develop some affinity toward the matrix. The nature of the interface plays a crucial role in the performance of the composites. However, pristine graphene does not disperse in organic polymer and agglomerates due to Vander Waals force of attraction and π–π interaction thereby affecting the desired properties. Moreover, the hydrophobic nature of graphene is incompat- ible with the polymer matrix. Therefore, prior to incorporation into the matrix, the surface of graphene should be functionalized with surface modifying agents depending on the type of polymer matrix. Study shows two types of surface modification: (i) covalent and (ii) non-covalent [59]. In the first kind, small or large organic moieties usually grafted to the surface of GO through covalent bond

3.3 Functionalization of Graphene 67and latter occurs through π–π interaction between the graphene surface and thebenzene ring carrying organic species [60, 61]. In non-covalent functionalization,no bond formation takes place between surfactant and GO, so the interfaceis not much effective in load transformation as covalent functionalization.Although, covalent functionalization is a better technique to tune the interface,which eventually influences the load transfer efficiency it hampers the electricalproperties.3.3.1Covalent ModificationCovalent modification can be carried out through nucleophilic substitution, elec-trophilic substitution, condensation, and cyclo-addition reaction. The functionalgroups such as porphyrins, poly(norepinephrine), oleylamine, and so on havebeen grafted onto the surface of GO [62–64]. Primary amines have been used tosimultaneously functionalize and reduce GO [65]. In nucleophilic substitutionreaction, the reactive epoxy groups of GO react with organic species bearinglone pair of electrons. This method is easy to process as it takes place at roomtemperature (RT) and produces scalable functionalized GO (fGO) required forthe fabrication of graphene/polymer composites. Li et al. obtained simultane-ously reduced and functionalized GO using octadecyl amine and incorporatedinto the polystyrene (PS) matrix. The fGO/PS composite showed enhancedmechanical and electrical properties at very low fGO loadings [66]. Kuila et al.prepared octadecyl amine functionalized graphene and fabricated ethylene vinylacetate (EVA) co-polymer composites with enhanced mechanical properties andthermal stability [67]. Jang et al. modified the surface of GO using three types ofalkyl amine namely, octylamine (OA), dodecylamine (DA), and hexadecylamine(HDA) to increase dispersibility in low-polar solvents and incorporated into PS[68]. It is seen that the properties of the composites were significantly affectedby alkyl chain length. The thermal stability and mechanical properties of thecomposites with longer alkyl chain amine–modified GO was higher as comparedto the composites with short chain alkyl amine functionalized GO. Kim et al.reported ternary roles of ethylenediamine to functionalize, reduce, and stitchingof GO and, so, formed conductive RGO was incorporated into the linear lowdensity polyethylene (LLDPE) matrix [65]. The composites showed enhancedmechanical properties suggesting the effective reinforcement of LLDPE withstitched RGO. One of the widely practiced and effective approaches other than nucleophilicsubstitution to achieve FG is amidation reaction, where edge –COOH groupsof GO attached with the –NH2 or cyanate groups of the modifiers form amidelinkage. Stankovich et al. used organic isocyanates to functionalize GO whichformed a stable dispersion in polar aprotic solvents [69]. Figure 3.1 showsthe proposed reaction mechanism between the organic isocyanates and GOshowing the formation of amide and carbamate ester linkage. Many –NH2 ter-minated molecules have been attached to GO surfaces intended for various end

68 3 Graphene Composites O OH O O OH OH HO O OH O O OH O OH RNCO RNCO R R O CO2HN NH R OH O O O O O N RN O O O H O O O H H HN–R ON ON R R O OFigure 3.1 Proposed reactions during the oxide sheets to form carbamate and amideisocyanate treatment of GO, where organic functionalities, respectively. (Reproducedisocyanates react with the hydroxyl (left oval) from Ref. [69] with permission from Elsevier.)and carboxyl groups (right oval) of grapheneapplications. Amine terminated polyethylene glycol (PEG) reacted with GOplatelets through an amide bond and showed potential applications in drugdelivery [70]. Biocompatible and water dispersible polysaccharides such aschitosan has been grafted onto the surface of GO by amidation reaction, which isuseful in gene and drug delivery system [71]. Esterification is a widely used method to tune the surface GO. The edge–COOH groups of GO react with CH2OH terminated molecules. Salavagioneet al. covalently functionalized graphene with poly(vinyl alcohol) (PVA) andpolyvinyl chloride (PVC) through esterification reaction [72, 73]. The esterifica-tion reaction of PVC with GO was carried out in the presence of DCC/DMAPfollowed by NaBH4 reduction. PVA grafting was carried out by two methods:first by using DCC/DMAP and second by using SOCl2 activated GO followed byreduction. Zhang et al. used fullerene to functionalize the surface of graphene[74]. The attachment of fullerene to graphene surface restrains graphene sheetpiling and helped in exfoliation. The GO can be covalently functionalized with silane through hydrolytic con-densation. Wang et al. functionalized graphene nanosheets (f-GNSs) by grafting3-aminopropyl triethoxysilane [75]. The silane-functionalized graphene was thenincorporated into the epoxy matrix, which showed significant improvementin thermal and mechanical properties. Addition of 1 wt% of functionalizedgraphene nanosheets (f-GNSs) increased the thermal degradation temperatureby ∼20 oC and the tensile strength by 45% as compared to neat epoxy. Figure 3.2

3.3 Functionalization of Graphene 69 CH3CH2O NH2 H2N H2N NH2 CH3CH2O Si OO Si O Si O HOOC Si COOH CH3CH2O(ATPS) O OOO O + H2O COOH Ethanol OHOOC COOH OH OH OH O Si O OH NH2 COOH O O O O HOOC OH Si Si O OH HOOC H2N (f-GNS) HO NH2 (GNS) Add Add Curing epoxy curing agentSonication of f-GNS High amplitude Stirring in ice bath Degassing under in the solvent sonication vaccumFigure 3.2 Schematic representation of functionalization and composite fabrication. (Reproduced from Ref. [75] with permission from Else-vier.)

70 3 Graphene Composites shows the schematic for the functionalization and fabrication of f-GNSs/epoxy composites. The sp2 carbons of graphene undergo free radical addition reaction and react with dienophiles. The reactive free radicals of diazonium salt would attack the sp2 bond of graphene and form a covalent bond [61]. During this reaction, sp2 hybridized carbon transforms into sp3, and the aromatic system gets disturbed, which resulted in a remarkable decrease in electrical conductivity. Fang et al. showed that the hydroxyl aryl groups can be grafted on the surface of graphene via diazo-coupling reaction and at the same time they can act as initiators for the polymerization of styrene via the atomic transfer radical polymerization (ATRP) method [76]. It has been found that the diazonium salt of benzoyl peroxide can also be used for the surface modification of graphene via free radical addition reaction. Kan et al. fabricated graphene grafted PS–polyacrylamide (PS–PAM) copolymer of styrene and acyl amide using benzoyl peroxide as initiator [77]. Addition of dienophiles to graphene is also a common method in the addition reaction category of graphene besides free radicals. Azomethine ylide is the most common dienophiles used in the functionalization of carbonaceous nanostruc- ture [61]. It reacts through 1,3 dipolar cyclo addition reaction and can offer a range of organic derivatives, which shows utility in a wide area of applications. The surface of graphene can be functionalized by polymers using surface- initiated ATRP and this technique generates surface controlled graphene that is required for various applications. In the ATRP technique first monomers and the initiators are adsorbed to the hydroxyl or carboxylic acid groups of GO followed by the polymerization of the monomers to polymer chains. Lee et al. carried out polymerization of styrene, methyl methacrylate (MMA), and t-butyl acrylate (tBA) via ATRP on the GO surface [78]. The polymer-functionalized graphene formed stable dispersion in DMF, chloroform, toluene, and dichloromethane. Layek et al. grafted PMMA on the surface of graphene by ATRP technique to prepare FG [79]. The PMMA-functionalized graphene (MG) showed good dispersion in a poly(vinylidene fluoride) (PVDF) matrix, which resulted in an increase of 124% storage modulus, stress at break 157%, and Young’s modulus 321% for 5% MG and 21 oC increase in Tg. 3.3.2 Non-Covalent Modification The non-covalent surface modification with surfactants, macromolecules, ionic liquids, and aromatic moieties are of paramount interest because these tech- niques tune the surface of graphene keeping its intrinsic properties intact. The interaction that involved is π–π, hydrogen bonding and hydrophobic attraction [60]. Non-covalent technique has been widely exploited to achieve the stable dispersion of graphene in various solvents. Largely, π–π stacking interaction and hydrogen bonding are most common as it concurrently stabilizes the graphene colloidal suspension and maintains the integrity of graphene. However, no bond formation takes place in non-covalent mode of functionalization; transfer of

3.6 Properties of Graphene/Polymer Composites 773.6Properties of Graphene/Polymer CompositesGraphene is the strongest known material with many astonishing properties thatcan be useful in many applications. It has the ability to significantly improve vari-ous properties of the composites at very low loadings. Graphene and its derivativeshave been extensively studied in relation to polymer composites, energy storage,and biology, and flame- retardant properties to name a few. With the develop-ment of advanced processing techniques, graphene is getting more attention as apromising material for future generation technology.3.6.1Mechanical PropertiesGraphene has exceptional mechanical properties, namely, high Young’s modu-lus, high aspect ratio, large tensile strength, and flexibility, which make it suitablefor reinforcing the physicochemical properties of neat polymer. Although pris-tine graphene is difficult to disperse in polymeric matrices due to its hydropho-bic nature, suitably functionalized graphene is very effective in reinforcing themechanical properties. The degree of enhancement in the mechanical propertiesof the composites depends upon two critical factors: (i) the dispersion of filler inthe composite, which determines the specific interfacial area and (ii) the inter-facial strength, which administrates the stress transfer across the interface. Boththe factors are very crucial but interfacial strength plays a major role that ulti-mately decides the mechanical properties. Interface can be designed by modifyingthe surface of graphene, which also helps to disperse it in the organic solvent andprevents restacking of layers. Thus, homogenous dispersion, proper exfoliation ofgraphene materials, and a strong interface are crucial to improve the mechanicalproperties. However, fabrication of composites with enhanced mechanical properties iscumbersome. Fine dispersion of nanofillers into the matrix, large interfacial area,and strong interface decides the overall mechanical properties of the composites.If any kind of interaction such as covalent, Vander Waals interaction, π–π inter-action, hydrogen bonding, and so on exists across the interface, then compositeswith enhanced mechanical properties can be anticipated. Such types of interac-tions play a pivotal role in strengthening the interface, as strong interface facilitatesthe stress transfers from nanofiller to the host matrix, thereby, making compositesmore robust. Graphene and its derivatives have been considered as a success-ful filler in improving composite’s mechanical properties; however, there are afew studies that report improvements in mechanical parameters such as tensilestrength, elastic modulus, elongation, and toughness. It is noted that the additionof filler may be helpful for the improvement of a certain property but may cause thedeterioration of other properties. Therefore, careful optimization of fillers (GNPs,rGO), dispersion, and interfacial strength is paramount to fabricate multifunc-tional high-performance composites. Wang et al. fabricated graphene-reinforced

78 3 Graphene Composites PVA composites film by deploying GO and graphene into an aqueous solution of PVA [113].The enhancement in tensile strength and elongation at break was 212% and 34%, respectively with ∼0.5 wt% of graphene content. Figure 3.7 shows the stress–strain curve of GO/PVA and graphene/PVA composites. Efficient load transfer takes place across the interface of materials only when the dispersion of fillers in polymer and the interaction between the two compo- nents are optimum. Liang et al. accomplished a simple and eco-friendly tech- nique to fabricate graphene/PVA composites [81]. The hydrogen bond formed between graphene and PVA helped in stress transfer across the interface. About 76% increase in tensile strength and a 62% improvement of Young’s modulus by the addition of only 0.7 wt% of GO was recorded. Cano et al. covalently func- tionalized GO with six different molecular weights of PVA through esterification of carboxylic groups of GO with hydroxyl groups of PVA [114]. The PVA func- tionalized GO (f-PVA-GO) was prepared through vacuum filtration and showed enhanced mechanical properties than the composites filled with GO. The f-PVA- GO/PVA composite showed ∼60% improvement in Young’s moduli and ∼200% improvement in tensile strength relative to GO filled PVA composites. Figure 3.8 shows the SEM images of the fracture surfaces of vacuum filtered paper-like mate- rials of pure GO and f-PVA-GO. Vadukumpully et al. prepared surfactant-wrapped graphene nanoflakes and PVA ultrathin composite films by a simple solution blending, drop casting, and annealing route [115]. The reinforcing ability of graphene improved the properties of the composites by ∼58% in Young’s modulus and 130% in tensile strength. Zhao et al. prepared fully exfoliated GNS/PVA composites through a facile solution mixing [116]. The composite with 1.8 vol% of graphene loadings showed ∼150% improvement in tensile strength and ∼10 times increase in Young’s modulus as compared to neat PVA. In another study, poly(dopamine)- modified GO/PVA (“dGO/PVA”) was prepared through solution mixing [117]. The composites showed improvement in tensile modulus, ultimate tensile 60 70 50 Tensile stress (MPa) 60 40 Tensile stress (MPa) 30 4 wt% 50 0 wt% 3.5 wt% 40 0.5 wt% 20 3 wt% 30 1 wt% 10 2.5 wt% 20 1.5 wt% 2 wt% 10 2 wt% 0 1.5 wt% 2.5 wt% 0 1 wt% 0 3 wt% 0.5 wt%(a) 0 wt% 50 100 150 200 250 300 0 30 60 90 120 150 180 210 240 Strain (%) Strain (%) (b) Figure 3.7 Typical stress–strain curves of (a) GO/PVA composites with varying GO loadings and (b) graphene/PVA composites with varying graphene loadings. (Reproduced from Ref. [113] with permission from John Wiley and Sons.

3.6 Properties of Graphene/Polymer Composites 79GO f-(PVA)GO (MW = 145 kg mol–1) 1 μm 10 μm 1 μm 10 μm(a) (b)Figure 3.8 SEM images of (a) GO and (b) f-PVA-GO composites. The inset shows high resolu-tion image. (Reproduced from Ref. [114] with permission from Elsevier.)strength, and strain-to-failure by 39%, 100%, and 89%, respectively with 0.5 wt%dGO loadings [117]. It may be due to the combined hydrogen bonding interactionamong GO, poly(dopamine), and PVA. Istrate et al. prepared graphene/PET bymelt compounding and reported enhanced mechanical properties at 0.07 wt% ofgraphene loadings [118]. About 23% and 42% improvement in Young’s modulusand tensile strength were recorded for the composites with 0.08 wt% fillerloadings. The improvement in mechanical properties of the composites at verylow loadings is industrially viable and may pave the way for graphene-basedcomposites from laboratory to commercial production. In another study, Shimet al. prepared functionalized GO (f-GO) with alkylbromide of different alkylchain length (hexyl, octyl, and dodecyl) by SN2 reaction and used as filler toprepare PET composites [119]. The f-GO/PET composite showed improvementin tensile strength, Young’s modulus, and elongation at break with the additionof f-GO. Figure 3.9 shows the tensile, Young’s modulus and elongation-at-breakcurve of GO/PET and fGO/PET composites against filler loadings. Liao et al. functionalized GO by octa(aminophenyl)silsesquioxane (OAPS) andincorporated into a PI matrix to study the reinforcement ability of GO/OAPS inthe matrix by in situ polymerization [120]. The authors revealed that the resultsshowed ultra-strong mechanical properties and low dielectric constant of theresultant composites. It may be ascribed to the strong interfacial interactiongenerated between OAPS and PI through covalent bonding. About 11.2-foldincrease in tensile strength and a 10.4-fold improvement in tensile modulusagainst neat PI were observed, which are very encouraging to explore andfabricate mechanically robust light weight composite with low dielectric prop-erties for various applications. Figure 3.10a,b shows the TEM image of GO andOAPS-GO at different magnifications and stress–strain curve of OAPS-GOfilm at different loadings of OAPS. Wang et al. prepared octadecylamine (ODA)functionalized GO (ODA-GA)/PI composites [121]. The composite with 0.3 wt%of ODA-GA loadings exhibited ∼6.4-times increase in tensile modulus and 240%enhancement in tensile strength. Further, the composite showed decreases in

80 3 Graphene Composites70 PET/fGO1 900 PET/fGO1 140 PET/fGO1 PET/GO PET/GO PET/GO60 800 120Tensile strength (MPa) Young’s modulus (MPa) Elongation at break (%)50 700 10040 600 8030 500 6020 400 40 0.0 0.5 1.0 1.5 2.0 2.5 3.0 wt% of filler 0.0 0.5 1.0 1.5 2.0 2.5 3.0 20 wt% of filler 0.0 0.5 1.0 1.5 2.0 2.5 3.0 wt% of filler Figure 3.9 Typical stress–strain, Young’s modulus and elongation-at-break cure of PET com- posites with GO and f-GO. (Reproduced from Ref. [119] with permission from the American Chemical Society.)(A) (B) 1000 4 1 Neat PI films 800 5 2 0.5 wt% OAPS-GO/PI films500 nm 50 nm 600 3 1.0 wt% OAPS-GO/PI films(C) (D) Tensile stress (MPa) 400 4 5 3.0 wt% OAPS-GO/PI films 3 5.0 wt% OAPS-GO/PI films 2 50 nm 20 nm 200(a) 1 0 01234567 (b) Strain (%) Figure 3.10 (a) TEM images of GO, OAPS-GO at low and magnification. (b) Typical stress–strain curve of OAPS-GO/PI film at various loadings of OAPS. (Reproduced from Ref. [120] with permission from the American Chemical Society.) dielectric constant with the increase in ODA-GA content. The unprecedented rise in mechanical properties may be due to the fine dispersion of polymer grafted GO and its bonding interaction with PI. Liao et al. prepared graphene/PI composites that exhibited ∼7.4 times increase in Young’s modulus and 240% improvement in tensile strength with only 0.3 wt% graphene loadings [122]. Cai et al. prepared GO/PU composites and investigated the mechanical, thermal, and scratch resistance properties [123]. The improvement in mechanical properties can be ascribed to the good interface formed due to the covalent

3.6 Properties of Graphene/Polymer Composites 81interaction between GO and PU. The composite showed ∼7 times improvementin modulus as compared to neat matrix and ∼50% toughness was achieved with1% of GO loading. These findings indicate that GO is capable of reinforcingthe properties of the composites and opens up many avenues to ponder uponthe fabrication of engineering composites for probable applications. Xianget al. fabricated hexadecyl-functionalized low-defect graphene nanoribbons(HD-GNRs)/TPU composite film by solution casting and investigated theirmechanical and barrier properties [124]. The finely distributed HD-GNRs withinthe PU matrix resulted in phase separation of the TPU as evidenced from theFT-IR spectra. The integration of HD-GNRs with the TPU matrix enhancedmechanical properties and is presented in Figure 3.11. Wang et al. PreparedGNS/PU composites and it showed ∼239% increase in tensile as compared toneat PU [125]. Rafiq et al. reported the preparation of FG filled nylon-11 and nylon-12 compos-ites by direct melt compounding method with the assistance of premixing [126].The authors studied the effects of FG on the ultimate tensile strength, elongationat break, impact strength, toughness, and permeation resistance of FG/nylon-11and FG/nylon-12 composite. It is seen that the reinforcing ability of FG was verygood at lower loadings. The FG/nylon 12 composites with 0.6 wt% of FG loadingsexhibited improvement in ultimate tensile strength, elongation at break, fracturetoughness, and impact failure energy by 35%, 200%, 75%, and 85%, respectivelyas compared to neat nylon 12. On the contrary, the FG/nylon 11 composites dis-played impressive improvement of 250% in impact strength at the loading of 1 wt%FG but only slight improvement was recorded in tensile properties and fracturetoughness. Roy et al. prepared polyethylenimine (PEI) functionalized GO (GO-PEI)/nylon composites by melt mixing technique. The improvement in tensilestrength was 37% and 54% for 0.25 and 0.35 wt% of GO-PEI loadings, respectively.50 0 14 0.05 0 0.2 0.05 12 0.540 0.2Stress (MPa) 1 Tensile modulus (MPa)0.5 2 10 330 1 5 2 8 320 510 60 4 6 0 12 3 4567 0 1 2 34 5(a) Strain (100%) (b) Filler concentration (wt%)Figure 3.11 (a) Stress–strain curves of TPU and TPU/HD-GNRs composite films. (b) Summaryof tensile moduli of different samples. (Reproduced from Ref. [124] with permission from theAmerican Chemical Society.)

82 3 Graphene Composites The improvement in Young’s modulus of the composite was found to be 65% and 74% as compared to that in neat nylon [127]. The multiple reactive amine func- tional groups of PEI helped in dispersion and in the improvement in mechanical properties of the nylon matrix. The inner surface of CNT is not accessible to make contact with polymer matrix, thus, affecting the reinforcing ability of CNT. Rafiee et al. unzipped multiwalled CNT to graphene nanoribbons and incorporated into the epoxy matrix to study the mechanical properties of the composites [112]. The improvement in Young’s modulus and ultimate tensile strength at 0.3 wt% loading was found to be 30% and 22%, respectively, as compared to neat epoxy matrix. It may be attributed to the increased adhesion between the components after unzipping of cylindrical inter- facial zone into a layered 2-D structure facilitating better load transfer at the inter- face. The effect of reinforcing ability of graphene platelets and single- and multi- walled CNT on the mechanical properties of epoxy nanocomposites has been investigated systematically [128]. Young’s modulus, tensile strength, and mode I fracture toughness of the graphene/epoxy composites were 31%, 40%, and 53% higher than those of CNT/epoxy composites. It may be attributed to the higher platelet orientation of graphene as compared to that of CNT. Wajid et al. employed two different processing techniques, namely, solution mixing and freeze drying for the incorporation of polyvinylpyrrolidone (PVP) stabilized pristine graphene into the epoxy matrix [129]. Although, both the techniques are promising in terms of dispersion of stabilized graphene in the epoxy matrix, freeze drying/mixing was relatively more versatile in terms of graphene dispersion. It is seen that PVP not only stabilizes the graphene but also helps to create a strong interface with epoxy matrix through its polar lactam rings and load transfer across the matrix. The composite exhibited ∼38% improvement in tensile strength and ∼37% in Young’s modulus at the loading of 0.46 wt%. Recently, efforts have been made to assemble graphene sheets in 3D scaffold and find out its applications in different fields such as the automobile, aircraft, and marine industries, as well as in the manufacture of fire retardant materials. Its superb performance as filler material in the fabrication of composites has started gaining attention due to its unique 3D structure and excellent properties. The techniques that are being used to prepare 3D graphene nanostructure are the template-directing method, cross-linking method, in situ reduction assembly method, and so on [130–132]. Ni et al. prepared 3D graphene skeleton (3DGS) using poly(amidoamine) (PAMAM) dendrimer as reductant and cross-linking agent by self-assembly and in situ reduction method, and incorporated into the epoxy matrix [133]. In order to obtain fine dispersion and to arrange the graphene sheet in the matrix they adopted resin transfer molding method rather than direct mixing. The unique arrangement of graphene sheets and better dispersion resulted in ∼120.9% and 148.3% increase in tensile and compressive strength, respectively. These results signify that if graphene sheets are arranged architecturally in a 3D structure, then it will endow better reinforcement and noble high-performance material can be anticipated. Figure 3.12 shows the variation of tensile, flexural, and compressive strength against filler loading.

3.6 Properties of Graphene/Polymer Composites 83Tensile strength (MPa) 160 160 epoxy Tensile strength (MPa)120 120 0.10 wt% EG/epoxy 0.20 wt% EG/epoxy 80 0.20 wt% 3DGS/epoxy 40 80 0 40 (a) 0.10 wt% epoxy0.20 wt% EG/epoxyEG/epoxy 3DGS/epoxy 0 300 0.0 0.5 1.0 1.5 2.0 2.5 3.0 200 Strain(%) 100 0.20 wt% (b)Compressive strength (MPa) 0 Flexural strength (MPa) 5 Flexural modulus (GPa) 160 4 120 3 80 2 40 0.10 wt% epoxy 0.20 wt% EG/epoxyEG/epoxy 3DGS/epoxy 0.10 wt% epoxy0.20 wt% EG/epoxyEG/epoxy 3DGS/epoxy 0.20 wt% 0.20 wt% (c) (d)Figure 3.12 Mechanical properties of com- and modulus. (Reproduced from Ref. [134]posites with different filler loadings: (a) ten- with permission from the American Chemicalsile strength, (b) tensile stress–strain curves, Society.)(c) compressive strength, (d) flexural strength Jiang et al. modified GO by APTES functionalized silica nanoparticles (ATGO)and incorporated into the epoxy matrix [135]. They studied the effect of fillerloadings on the tensile strength and impact strength both at RT and cryogenictemperature (CT). It is seen that the tensile strength and tensile modulus of theATGO (1 wt%)/epoxy at CT were 29.2% and 22.0% higher than that of neat epoxy.The impact strength at RT was much higher than that at CT. It may be attributedto the presence of multifunctional groups and hence, good interaction betweenmatrix and ATGO. Jiang et al. also modified the surface of silica nanoparticlesby APTES and then attached with GO using isocyanate group-terminated flex-ible chains to synthesize modified filler (SATPGO) for the fabrication of epoxycomposites [134]. The impact strength of the SATPGO filled (0.5 wt%) epoxy com-posite was found to be 154% and 92% higher than that of neat epoxy at RT and

84 3 Graphene Composites CT, respectively. Tang et al. prepared RGO filled epoxy composites and investi- gated the influence of dispersion on the mechanical properties of the composites [136]. They reported improvement in KIC values by 24% and 52% at 0.2 wt% RGO loadings. It indicates that the state and extent of dispersion greatly influence the end properties of composites. In a recent study, Jia et al. prepared interconnected porous structure graphene foam (GF)/epoxy composites and reported ∼53% and 38% enhancements in flexural modulus and strength at ∼0.2 wt% of GF [137]. The search for sustainable biodegradable polymers and the quest for mechani- cally robust and thermally enduring biodegradable polymer composites have stim- ulated many research activities. Gao et al. prepared poly GO/(propylene carbon- ate) (PPC) composites [138]. They reported 1000% increase in tensile strength, 1800% increase in Young’s modulus, and 10 oC improvement in Tg. This encourag- ing result has shown new prospect of mechanically strong biodegradable polymers alternatives to petroleum-based polymers. Layek et al. prepared GO/sodium car- boxymethylcellulose (NaCME) composite and reported ∼188% and 154% increase in tensile strength and Young’s modulus as compared to neat NaCME at only 1 wt% loading [139]. Zhang et al. reported the preparation of 3D graphene aerogel (GA)/polydi- methylsiloxane (PDMS) composites (GAPC) by incorporating GA into PDMS by direct infiltration [140]. The well-aligned interconnected 3D GA frame- works/PDMS composites exhibited large mechanical deformability (compressive strain = 80% and tensile strain = 90%) along with stable piezo-resistance effect, and excellent electric Joule heating performance. With such outstanding prop- erties, they may find applications in the field of stretchable electronic devices, ultra-large strain sensors, thermal interface materials, and so on. 3.6.2 Thermal Properties Polymer composites with conductive fillers have emerged as one of the efficient heat dissipater materials. Despite interfacial resistance between the filler and poly- mer matrix, graphene has been extensively used to impart thermal conductivity to the host matrix due to its high thermal conductivity and flat surface area. Thermal expansion is also a very crucial property of graphene-based polymer compos- ites. Due to negative CTE of graphene, high specific surface area and mechanical properties may reduce the CTE of the composites when it is deployed in poly- mers. The composites with high thermal conductivity and low CTE find potential applications as thermal interfacial material to reduce the thermal resistance and to dissipate heat in various electronic components. Li et al. prepared vertically aligned and densely packed multilayer graphene (MLG)/epoxy composite (AG/E) and reported maximum thermal conductivity of ∼33.54 W m K−1 [141]. The ther- mal conductivity measured in perpendicular and parallel directions was found to be 16.75 and 33.54 W m K−1, respectively. Further, the thermal conductivity of AG/epoxy composite exhibited a positive temperature response while increasing the temperature from 40 to 90 ∘C. Figure 3.13 represents the thermal conductivity

3.6 Properties of Graphene/Polymer Composites 85 Parallel Perpendicular Thermal conductivity (W m–1 K–1) 40 Parallel direction of AG/E direction direction 30 Perpendicular direction of AG/E DG/E X1, K11 X3, K33 20 10 0(a) X2, K22 40 50 60 70 80 90 (b) Temperature (°C)Thermal conductivity (W m–1 K–1) 40 First cycle 40 Thermal conductivity (W m–1 K–1) Second cycle 30 20 30 Third cycle 10 20 0 Parallel direction of AG/E 0 Perpendicular direction of AG/E 10 (d) 0 40 50 60 70 80 90 10 20 30 40(c) Temperature (°C) Angle (θ)Figure 3.13 (a) Schematic diagram of par- Thermal conductivity as a function of tem-allel and perpendicular tensile strength, as perature for AG/E for three heating cycles. (d)well as electrical and thermal conductivi- The effect of the alignment extent on the-ties of AG/E in parallel and perpendicular oretical thermal conductivity in AG/epoxy.directions. (b) Temperature dependence of (Reproduced from Ref. [139] with permissionthermal conductivities in AG/E and DG/E. (c) from the American Chemical Society.)of the MLG/epoxy composites along parallel and perpendicular directions. Thecommercially available GnPs/epoxy composites were prepared by applying highcompression force to closure the gaps between the adjacent GnPs of large lat-eral dimensions. The thermal conductivity of the composites was found to be veryhigh (12.4 W m K−1) as compared to neat epoxy (0.2 W m K−1) [142]. The additionof electrically insulating boron-nitride into the composites further improved thethermal conductivity and suppressed the electrical conductivity. Thermally con-ducting but electrically insulting composites so formed may find huge applicationsin the field of electronic capsulation. Song et al. prepared noncovalently modified non-oxidized graphene flakes(GFs) by 1-pyrenebutyric acid (PBA) and the functionalized GFs (f-GFs) wereincorporated into the epoxy matrix [143]. The thermal conductivity of thecomposites was found to be ∼1.53 W m K−1 at 10 wt% of f-GF loadings. In

86 3 Graphene Composites addition to thermal conductivity, f-GFs also enhanced the mechanical properties of the epoxy composites and the modulus value was found to be 1.03 GPa at 1 wt% loading. Wang et al. incorporated GO into the epoxy matrix and investigated CTE behavior [144]. It has been found that the CTE values of the composites with 5 wt% GO loadings were decreased by 32% as compared to those of neat epoxy. Thermal conductivity of the composites was also increased fourfold as compared to that of neat epoxy. All these findings indicate that graphene/epoxy composites can be considered as promising candidates in the field of thermal management. Gallego et al. studied the effect of graphene platelets on the UV cured epoxy and reported the unprecedented 40 ∘C rise in Tg of the composites along with improved stiffness [145]. TEM results as shown in Figure 3.14 demonstrate the dispersion state of graphene platelets as well as the interaction between the graphene and matrix, which is reflected in the improvement of thermal and visco- elastic properties. Jia et al. prepared interconnected porous structure GF/epoxy and reported ∼31 oC increase in Tg [137]. Araby et al. prepared electrically and thermally conductive GnPs/styrene butadiene rubber elastomeric composites by melt compounding and the reinforcing ability of GnPs was compared with that of CNTs [146]. It has been found that the percolation threshold in thermal conductivity was at 16.5 vol% GnPs loading. The highest thermal conductivity of the composites at 41.6 vol% GnP loading was ∼240% higher as compared to that of neat epoxy. The significant improvement of thermal conductivity is encouraging for scaled-up production of elastomeric composites. Design and control of interface is very crucial to achieve high-performance composites materials. Fang et al. tuned the interface between single-layer graphene nanosheets (SLGNs) and PS by controlling the grafting density and chain length of polymer through combination of diazonium addition and ATRP 0.2 μm 20 nm(a) (b)Figure 3.14 TEM images of graphene-epoxy resin composite (1.5 wt% of filler loadings) at(a) low and (b) high magnification. (Reproduced from Ref. [145] with permission from theAmerican Chemical Society.)

3.6 Properties of Graphene/Polymer Composites 87[145]. The Tg for the composites with high grafting density and low molecularweight polymer-grafting was found to be ∼18 and 9 oC higher as comparedto that of neat PS. Further, the thermal conductivity of the SLGN (2 wt%)/PScomposites was ∼2.6 times higher than that of neat PS. Patole et al. functionalizedgraphene sheets by PS nanoparticle following water-based in situ micro-emulsionpolymerization and prepared PS composite film [147]. The composite exhibited∼16 ∘C rise in degradation temperature and temperature at the maximumreactive velocity obtained from differential thermal analysis peaks was increasedby 15 oC at 20 wt%. The Tg was found to be increased by ∼17 oC due to the goodinterfacial bonding that restricted the segmental motion of the polymer chains.Salavagione et al. prepared RGO/PVA composites by reducing GO in the pres-ence of PVA matrix and coagulating the obtained material with 2-propanol [148].The hydrogen-bonding interaction between PVA and RGO resulted in loweringthe melting temperature (Tm) of the composites by ∼35 oC and the crystallinitydecreased to a half for the composites with 5 wt% of RGO loadings as compared tothat of neat PVA. The Tg of the composite was increased by ∼21 oC. The maximumdegradation temperature of the composites was increased by more than 100 oC. Wang et al. prepared GNS/PU composites by in situ polymerization [124]. Thecomposite exhibited ∼40 oC improvement in thermal stability with the addition of2.0 wt% of GNS. The FG/PU composite exhibited ∼50 oC improvement in thermaldegradation temperature with 4 wt% of FG loadings [123]. These results revealedthat graphene can reduce the diffusion of heat in the matrix and can act as a barrierto prevent the emission of thermally degraded small gaseous molecules. Myriadof studies have been carried out with graphene as filler material to fabricate poly-mer composites anticipating excellent material properties. However, apparentlythe intrinsic properties of graphene has not fully embedded in the composite. Thelimitation may lie in the processing and designing of the composite materials. Ifthe graphene surface is meticulously designed in order to improve the interfacewith the polymer matrix, end materials with better properties can be expected.Liao et al. designed a flexible and covalently bonded interphase using two differ-ent molecular weight linear poly(oxyalkylene) amines with GO (D400-GO andD2000-GO) and incorporated into the PI matrix by in situ polymerization [122].The Tg of the composite was increased by ∼23.96 ∘C suggesting the restricted seg-mental motion of the polymer chains. Figure 3.15 shows the TMA curves of neatPI film and D400-GO/PI films. Kuila et al. prepared graphene/PMMA composites by in situ emulsion poly-merization followed by in situ reduction of GO and subsequent film casting in aglass petri dish [149]. The fine dispersion of graphene sheets in the PMMA matrixshowed enhanced thermal stability of 26 oC as compared to that of neat PMMA. Yun et al. studied the reinforcing effects of alkylated graphene oxide (AGO) ona nonpolar PP matrix and investigated the thermal and mechanical properties rel-ative to 1D CNTs [150]. They reported ∼33 ∘C increase in thermal degradationtemperatures with the addition of only 1 wt% AGO into the PP matrix. Further,∼0.1 wt% of AGO increased the Young’s modulus of the PP by more than 70%.The reinforcing ability of AGO relative to that of CNTs is mainly due to its 2-D

88 3 Graphene Composites(a) (b) Figure 3.15 TMA curves of (a) neat PI film and D400-GO/PI films and (b) neat PI film and D2000-GO/PI films. (Reproduced from Ref. [122] with permission from the American Chem- ical Society.)100 10080 80Weight loss (%) Weight loss (%)60 6040 PP 40 PP/AGO 01 20 PP20 PP/AGO 03 PP/ACNT 05 PP/AGO 05 0 PP/AGO 05 PP/AGO 07 0 PP/AGO 10 330 360 390 420 450 480 330 360 390 420 450 480(a) Temperature (°C) (b) Temperature (°C)Figure 3.16 TGA profile of (a) alkylated graphene oxide/polypropylene composites and (b)different carbon fillers/polypropylene composites. (Reproduced from Ref. [148] with permis-sion from Elsevier.)structure and fine dispersion in the matrix, which may help to build strong inter-facial strength. Figure 3.16 shows the TGA curve of AGO/PP composites and itscomparison with CNTs. The above findings substantiate the fact that grapheneand its derivatives are very useful to fabricate thermally conductive compositesmaterials and also confirm that graphene sheets can act as a barrier to inhibitthe propagation of heat in a polymer matrix, which results in improved thermalstability of polymer composites material.3.6.3Electrical PropertiesThe high aspect ratio and excellent electrical conductivity of graphene make itattractive as a filler to impart conductivity to the insulating polymer material.

3.6 Properties of Graphene/Polymer Composites 89Percolation threshold, ϕc, (the value at which conductive network of filler is estab-lished) of the graphene/polymer composites is anticipated to be very low due toits large intrinsic electrical conductivity and high surface area, thereby, enhancingthe electrical properties of the composites at lower loading of graphene. Largenumbers of polymers such as PMMA, PS, PU, epoxy PDMS, PVA, and so onhave been used as matrix to incorporate graphene to fabricate conductive poly-mer composites. Conductive polymer composites find applications in electromag-netic shielding (EMI), antistatic coating and conductive paints. Rather than to usegraphene prepared by CVD, thermally or chemically reduced graphene are used toimpart conductivity to the composites mostly due to the processing simplicity andcost effectiveness. Stankovich et al. reported exfoliated phenyl isocyanate-treatedGO/PS composite by solution mixing followed by in situ reduction and reportedconductivity of 0.1 S cm−1 at ϕc of 0.1 volume fractions [151]. This unprecedentedenhancement in electrical conductivity may be due to the large aspect ratio ofCFG and its uniform distribution into the host matrix. Figure 3.17 shows the elec-trical conductivity of PS composites against filler loadings. Pham et al. fabricatedchemically converted graphene (CCG)/PS composite by solution mixing followedby compression molding [152]. The authors revealed low percolation threshold of0.19 vol% and high electrical conductivity of 72.18 S m−1 at loading of 2.45 vol%.The high electrical conductivity of CCG contributed to make conductive compos-ite. Fine dispersion of CCG also made composites more stiff, which was reflectedin increasing storage modulus value. 101 10–2Conductivity, σc (S m–1)IVVI 0 10–5 Log σc –2 10–8 V I –410–11 V I10–14 –6 –3 –2 –4 Log (ϕ–ϕc) 0.0 0.1 0.5 1.0 1.5 2.0 2.5 Filler volume fraction, ϕ (vol%)Figure 3.17 Electrical conductivity of isocyanate treated graphene/PS composites. Insetfigure represents percolation threshold and set up for measurement. (Reproduced from Ref.[149] with permission from Nature Publication.)

90 3 Graphene Composites Li et al. simultaneously functionalized and RGO with ODA and fabricated ODA-functionalized GO (GO-ODA)/PS composite [66]. The in situ reduction was confirmed by the enhanced electrical conductivity of GO-ODA and it was further increased during the compression-molding of the composites at 210 oC. The sharp transition in electrical conductivity (percolation threshold) for the composites was achieved at 0.45 vol% of GO-ODA loadings. The electrical conductivity of the composites at 0.46 and 0.92 vol% was found to be 9.2 × 10−4 and 4.6 × 10−1 S m−1, respectively. Qi et al. compared the efficiency of graphene over CNT in improving electrical conductivity by fabricating PS composites. The graphene/PS composites showed 2–4 orders of magnitude higher electrical conductivity than CNT/PS composites [153]. The electrical conductivity of graphene/PS composites increased from ∼6.7 × 10−12 to 349 S m−1 with the graphene loadings ranging from ∼0.11 to 1.1 vol%. Further, the percolation threshold in electrical conductivity of the graphene/PS composites decreased approximately four- to fivefold while employing the exclusion (or selective localization) principle. Figure 3.18 shows the electrical conductivity of the graphene/PS and CNT/PS composites against filler content. Wu et al. constructed 3D, compactly interconnected graphene network by self-assembling of PS and EVA polymer matrix [155]. The compact, continuous 3D architectures of graphene in polymer matrix facilitated electron flow seam- lessly and exhibited electrical conductivity of 1083.3 and 260 S m−1 at 4.8 and 2.2 vol% graphene loadings, respectively. It is seen that the EVA composites not only exhibited high electrical conductivity but also showed enhanced flexibility. Self-assembly process could be proven to be advantageous as a compact structure can be designed with less contact resistance. In addition, this process is eco- friendly due to processing in an aqueous media. Figure 3.19 shows the electrical conductivity of PS composites and foam-like morphology of the 3D graphene 101DC electrical conductivity (S m–1) 10–3 10–7 10–11 PS/CNT PS/Graphene PS/Graphene/PLA 10–15 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Filler content (vol%)Figure 3.18 Variation of electrical conductivity of PS composite against filler loadings.(Reproduced from Ref. [154] with permission from the American Chemical Society.)

3.6 Properties of Graphene/Polymer Composites 91 103 CNT, solvent blending 101 GS, solvent blending 10–1 10–3 GS, self-assembly and hot-press (a)Conductivity (S m–1) 2.2 vol% 4.8 vol% (b) 500 nm (c) 100 μm (d) 2 μm (e) 500 nmFigure 3.19 (a) Electrical conductivity cross-sectional SEM images of the remainingcomparison of the graphene and carbon graphene skeleton after annealing the test-nanotube–polystyrene composites fabricated ing samples at high temperature under nitro-by solution mixing and electrostatic self- gen atmosphere. (Reproduced from Ref. [151]assembly methods. (b) Top view and (c–e) with permission from John Wiley and Sons.)after annealing at high temperature. Hsiao et al. prepared GO/WPU compositesby LbL alternate deposition of cationic surfactant (didodecyldimethylammoniumbromide, DDAB)-adsorbed positive charge GO and negative charge carryingGO suspension on the negatively charged WPU fibers followed by HI reductionto obtain RGO [156]. The RGO formed a network in the matrix and resulted inenhanced electrical conductivity of ∼16.8 S m−1 for the composite. Zhang et al. prepared graphene/PET composites by melt compounding tech-nique with very low percolation threshold in electrical conductivity at 0.47 vol%[157]. The electrical conductivity of the composite was found to be 2.11 S m−1at 3.0 vol% of graphene loadings. The high electrical conductivity that grapheneimparted is due to its high aspect ratio, intrinsic electrical conductivity, anddispersion in the matrix. Vadukumpully et al. prepared surfactant-wrappedgraphene nanoflakes/PVC ultrathin composite films. The percolation thresholdin electrical conductivity of the composite was 0.6 vol% and showed a maximumelectrical conductivity of 5.8 S m−1 at 6.47 vol% of graphene loading [115].

92 3 Graphene Composites Shen et al. compared the efficiency of GO reduction of glucose and polyvinyl- pyrrolidone (PVP) and found glucose to be more effective than PVP [158]. The glucose-reduced GO (rGO-g)/PLA composites exhibited electrical conductivity of 2.2 S m−1 with 1.25 vol%. Furthermore, it is seen that glucose is a more efficient reducing agent as compared to that of the chemically RGO. Vuluga et al. prepared RGO/PMMA composites through biphasic process by radical addition pathway [159].The electrical conductivity of the composites was 3.4 S m−1, which could be beneficial for EMI shielding purpose. Recently, Jia et al. prepared an interconnected porous structure GF/epoxy com- posites by impregnation of epoxy resin into the 3D GF [137]. The composites showed the electrical conductivity of 3 S m−1 at the loading of 0.2 wt% GF. The significant increase in electrical conductivity may be due to the large surface area of GF and fast movement of charge carrier through the 3D network. The 3D GF is not only effective in enhancing electrical conductivity but also in influencing the mechanical and thermal properties significantly. Chen et al. synthesized 3D GF by template-directed CVD and fabricated GF/poly(dimethyl siloxane) composites [160]. The flexible interconnected network structure of GF acted as a fast trans- port channel for charge carrier and the composites showed electrical conductivity of ∼1000 S m−1 even with loading of 0.5 wt%. The high electrical conductivity with low ϕc can be attributed to the unique interconnected network structure of GFs through which electrons can pass without interruption. Recently, Zhang et al. fab- ricated 3D GA–PDMS composites (GAPC). The interconnected 3D framework composites structure exhibited electrical conductivity of 114.7 S m−1 [140]. Zheng et al. performed one-step in situ reduction and polymerization pro- cess to fabricate polyamide RGO/PA-six composites using insulating GO and ε-caprolactam monomer [154]. Homogeneous dispersion of RGO in the PA six matrices resulted in electrical conductivity of ∼0.028 S m−1 only at loading of ∼1.64 vol% of GO. Yoonessi et al. prepared GNS/bisphenol A polycarbonate composite by both emulsion mixing and solution blending technique was subsequently compression molded at 287 ∘C [161]. The composites prepared by emulsion mixing and solution blending showed dc electrical percolation threshold of ∼0.14 and 0.38 .vol% and the conductivity of 0.512 (emulsion mixing) and 0.226 S m−1 (solution blending) at 2.2 vol%. The improvement in electrical conductivity can be ascribed to the high aspect ratio, nanostructure directed assembly of GNS, and inherent conductivity of the GNS. Recently, Vasileiou et al [162]. incorporated TRGO into the maleated LLDPE and its amino-pyridine derivatives by melt compounding [163]. The aromatic moi- eties on the backbone of polymer chains interacted with graphene surface through non-covalent interaction and formed electrically conducting graphene sheets. The percolation threshold in electrical conductivity for the non-covalently modified graphene was very low (0.5–0.9 vol%). Hyunwoo et al. coated ultrahigh molec- ular weight PE with GO sheets and then reduced using hydrazine subsequently resulting in the formation of segregated composites [164]. The two-step prepa- ration technique might help to prevent aggregation of RGO sheets; the compos- ites exhibited high electrical conductivity at a very low percolation threshold of

3.6 Properties of Graphene/Polymer Composites 930.028 vol%. The process could be effective to fabricate electrically conductive com-posites with low percolation threshold. Kim et al. applied three most widely usedprocedures, that is, melt intercalation, solvent mixing, and in situ polymerizationto fabricate thermoplastic PU composite using TRG and isocyanate treated GOas fillers [165]. Effective exfoliation of thin sheets made composite conductive at∼0.5 wt% of TRG loadings. It has been found that the solution mixing was betteras compared to melt mixing or in situ polymerization in terms of distribution offillers in the polymer matrix.3.6.4Dynamic Mechanical PropertiesThe dynamic mechanical properties such as storage modulus and loss angletangent (tan ) values depend on the interfacial bonding strength between thefiller and matrix. Loss tangent depends on the restricted movement of polymersegments and is helpful for determining the Tg value. Roy et al. functionalizedgraphene with PEI and fabricated GO-PEI/nylon composites [127]. The PEIinteracts with GO through hydrogen bonding as it contains several aminefunctional groups resulting in the homogeneous dispersion of PEI-GO into thenylon matrix. The storage modulus of the composite (0.25 wt%) was ∼63% higheras compared to neat polymer and showed ∼15 ∘C increment in thermal stabilitywith ∼0.35 wt% loadings. Mittal et al. prepared TRGO filled HDPE, LLDPE, PP, PS, and PC compositesby melt mixing technique [166]. The effects of filler content and type of polymeron the mechanical, thermal, and rheological properties have been investigatedin detail. Among the studied composites, LLDPE composites showed 100%improvement in tensile modulus and 100% storage modulus with 7 wt% of TRGOloadings. These findings offered the explanation that not only polarity matchbetween the matrix and filler but also the nature of polymer and processingconditions is important for improvement in physicochemical properties. Chang-sheng Xiang et al. reported GNS/PU composites by in situ polymerization [124].Di-isocyanates were grafted onto the GNS followed by in situ polymerization.The composites exhibited improved tensile strength, storage modulus, electricalconductivity, and thermal stability due to the direct bond formation between GNSand PU, fine dispersion, and interfacial interaction. The storage modulus and thetensile strength were found to be increased by ∼202% and ∼239%, respectively,with the incorporation of 2.0 wt% of GNS. The effect of RGO and surface-modifiedRGO on the mechanical properties of PI composite was investigated in detail[167]. The surface-modified RGO filled composites showed 25–30% higher stor-age modulus than that of RGO filled composite. The improvement in mechanicalproperties can be attributed to the fine dispersion and the stiff benzene ring of PImoieties. Figure 3.20 shows the variation of storage modulus for RGO and surfacetreated RGO filled PI composites at different temperature. It shows that thestorage modulus increased with increasing RGO or surface treated RGO content.

94 3 Graphene Composites 2500 50 °C 150 °C 1700 2000 1600E′ (MPa)1500 1500 150 °C E′ (MPa) 1400 1000 1300 1200 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0.00 0.25 0.50 0.75 1.00 1.25(a) Graphene wt% (b) Graphene wt% Figure 3.20 (a) Dynamic tensile moduli of the polyimide nanocomposites containing graphene polyimide nanocomposites at 50 (b) and 150 ∘C (2) as a function of graphene graphene (9) and rigid imidized graphene (0) content. (b) Dynamic storage moduli of at 150 oC. (Reproduced from Ref. [166] with permission from American Chemical Society.) Layek et al. prepared GO/sodium carboxymethyl cellulose (NaCME) compos- ite by solution casting method [130]. Hydrogen bond between the hydrophilic group of GO and NaCME resulted molecular level dispersion of GO in NaCME, which is reflected in the 174% increase in storage modulus at the loading of 1 wt% GO along with the increase 188% and 154% in tensile stress and Young’s modulus respectively. Figure 3.21 shows the schematic for the hydrogen bonding interac- tion between GO and NaCME. In another report of biodegradable polymer, poly (propylene carbonate) (PCC) an increase of 1600 MPa was recorded with the addi- tion of 1 wt% GO along with 10 oC rise in Tg [138]. Tang et al. dispersed RGO sheets into the epoxy matrix with and without ball mill mixing [41]. The effect of dispersion state of graphene on the various properties of the composites have been investigated in detail. It is seen that the well-dispersed RGO significantly improved the Tg, tensile and flexural strength of the composites as compared to the poorly dispersed RGO. The increase in Tg was ∼11 oC with the addition of only 0.2 wt% well-dispersed RGO to epoxy. Figure 3.22 shows the dynamical mechanical properties of well-dispersed and poorly dispersed RGO/epoxy composites and it is apparent that storage modulus improvement is comparatively high in the case of highly dispersed RGO. There- fore, it is seen that the fine dispersion of graphene nanoparticles is very crucial to property improvements. 3.7 Application of Graphene Based Polymer Composites Graphene/polymer composites have been studied for a myriad of applications such as engineering materials, gas barrier, sensor, EMI shielding, fire retardant

3.7 Application of Graphene Based Polymer Composites 95 O O OO O O CH2C O– CH2C O– CH2C O– CH2C O– CH2C O– CH2C O– O O O O O OOO HO OH O HO O HOHO O HO OH O HO OH O HO OH HO O OH O O O O O O O O O OO O O HO OH HO HO O HO OH O HO O OH O OH OO O OH O OH O O CH2C O– CH2C O– CH2C O– CH2C O– CH2C O– CH2C O– OO O O OOHO OH HO HO OH C OH CO C OH O O OH CO O O OC OC OH O OH O HO HO OH OH O O O O O O CH2C O– CH2C O– CH2C O– CH2C O– CH2C O– CH2C O– O O O O O O HO OHOO O OO HO O HO OH O HO OH O HO OH O HO OH O O HO OHHO O O HO O OO HO O OO HO O O HO O O O OH O O O HO CH2C O– O OH OH OH OH O O O CH2C O– CH2C O– CH2C O– CH2C O– CH2C O– OO O O OOFigure 3.21 Hydrogen bonding interaction between GO and NaCME. (Reproduced from Ref.[130] with permission from John Wiley and Sons.)materials, turbine blades, conductive composite, in automotive, aircraft, andmarine industries, and so on [25, 96]. Advanced polymer composites materialsare becoming very crucial components in the design of new windmill blades,aircraft components, and other various applications, which require light-weight,high-strength materials. Graphene-based polymer composites are promisingmaterials in the pursuit of developing structural components with a highstrength-to-weight ratio. The amount of filler loading, dispersion of graphenein the matrix, and interaction between the components are very crucial param-eters to determine the prospect of graphene-based composites for real-worldapplications.3.7.1Gas BarrierThe large surface area and the non-permeable flat sheet-like structure of graphenehas allowed its applications in the areas of gas barrier applications. If graphene orits derivatives are finely dispersed in polymer matrix, the “tortuous path” so gen-erated by graphene would inhibit the molecular diffusion of gaseous moleculesto remarkably reduce permeability. It is imperative to design thin film polymericmaterials that can prevent diffusion of moisture, which is desirable in the field offood packing, electronics, and fuel cells. Yang et al. reported thin film structure

96 3 Graphene CompositesStorage modulus (MPa) 3000 Poorly dispersed Storage modulus (MPa) 3000 Highly dispersed 2500 2500 2000 2000 1500 1500 1000 0 wt% 1000 0 wt% 0.05 wt% 0.05 wt% 500 0.10 wt% 500 0.10 wt% 0.20 wt% 0.20 wt% 0 0 40 60 80 100 120 140 160 180 40 60 80 100 120 140 160 180(a) Temperature (°C) (b) Temperature (°C) 1.4 Poorly dispersed 1.4 Highly dispersed 1.2 1.2 1.0 0 wt% 1.0 0 wt%Tan δ 0.8 0.05 wt% Tan δ 0.8 0.05 wt% 0.10 wt% 0.10 wt% 0.6 0.20 wt% 0.6 0.20 wt% 0.4 0.4 0.2 0.2 0.0 0.0 40 60 80 100 120 140 160 180 40 60 80 100 120 140 160 180(c) Temperature (°C) (d) Temperature (°C) Figure 3.22 Dynamic mechanical properties of cured epoxy composites containing: (a) and (c) poorly dispersed RGO; (b) and (d) highly dispersed RGO. (Reproduced from Ref. [41] with permission from Elsevier.) of GO/PEI composite prepared by LbL assembly of alternate GO and PEI layers to investigate the oxygen barrier properties of these composites [168]. The signif- icant reduction of oxygen as well as carbon dioxide permeation properties of the composites has the potential to prevent hydrogen permeation. The remarkable performance of the composites with regard to barrier properties can be ascribed to the “tortuous path” that the packed nano brick structure composites created prevented the diffusion of gases. Liu et al. prepared RGO/PEI composite film, where PEI performed ternary roles of a reductant, modifier, and polymer host [169]. The brick and mortar type structure so formed by vacuum filtration of PEI- RGO showed gas barrier properties and the properties increased with increasing PEI content whereas conductivity got decreased. The gas barrier properties can be attributed to the finely wrapping of graphene sheet by polymer chains. The FG/nylon composites has shown very efficient water vapor and gas barrier prop- erties at very small loading of FG. The composites showed decreased water vapor and oxygen permeability by 49% and 47%, respectively, with 0.3 wt% of FG load- ings [126]. It can be ascribed to the intrinsic flexibility of graphene, 2D structure, high aspect ratio, and fine dispersion of FG. Layek et al. showed that the oxygen

3.7 Application of Graphene Based Polymer Composites 97permeability coefficient of the SPG/PVDF composites was significantly lowerthan that of the neat PVDF [85]. Based on the above results, graphene/polymercomposite can be used as food packaging materials. Specific examples of appli-cations include packaging for cheese, cereals, edible oils, confectionery, fruitjuice, dairy products, processed meats, beer, carbonated drinks, and so on. Theuse of graphene/polymer composite as packaging materials would be expectedto significantly improve the shelf life of many types of food. The transportationand storage of hydrogen gas are highly dangerous due to its continuous diffusionthrough the wall of the container. Therefore, graphene filled polymer compositescan be used as coating materials to prevent the diffusion of hydrogen gas.3.7.2SensorPrecisely designed thin films (sensing layer) of graphene/polymer composites havedrawn significant interest in the paradigm of sensing materials. Conducting poly-mer/graphene composites are important as sensing materials due to its effective-ness as senors, low cost of production, and high electrical response. The excellentelectrical conductivity of graphene is the main factor that makes it suitable for usein sensing technology. Wu et al. fabricated graphene/PANI composites by chem-ical oxidative polymerization for the detection of NH3 [170]. It is seen that thesensitivity of graphene/PANI composites was higher than that of neat PANI withthe detection limit of 1 ppm. Yang et al. prepared CVD graphene/PMMA com-posites for the development of flexible sensors on paper substrate that can detectNO2 at the concentration of 200 ppm under different strains [171]. Bai et al. testedthe sensing ability of graphene/polypyrrole hydrogels against NH3 and found goodsensing ability toward NH3 [172]. It has been found that the dG-O/PVA compositefilm can be used as moisture sensors [117].3.7.3EMI ShieldingGraphene-based thermoplastic/thermoset composite has attracted significantattention as EMI shielding materials due to the low percolation threshold inthe mechanical and electrical properties of the graphene filled composites. It isvery essential to protect interference prone equipment from malfunctioning.EMI not only hampers electronic devices but is also critical to human health.Thus, metal and metallic composites were the choice for shielding purposesdue to their high electrical conductivity. However, difficulty in processing,heavy weight, and its chemical instability has urged to look beyond it. Therefore,graphene filled polymer composites has started getting attention as EMI shieldingmaterials due to its light weight, large specific surface area and good electricalconductivity. Yan et al. fabricated RGO/PS composites via high pressure solid-phase com-pression molding for EMI shielding applications [173]. Multi-facet segregated

98 3 Graphene Composites architecture of RGO in PS matrix provides many interfaces that facilitated efficient shielding of EMI. The shielding effectiveness of the composite was found to be 45.1 dB at a loading level of only 3.47 vol%. Further, the compressive strength and compressive modulus of the composite was 94% and 40% higher as compared to those of neat PS. Figure 3.23 shows the variation of electrical conductivity, EMI shielding effectiveness, and mechanical properties. The electrical conductivity of f-G/PVDF composite was 10−4 S m−1 for 0.5 wt% of f-G loadings. Further, the EMI shielding effectiveness of the composite was found to be ∼20 dB in X-band (8–12 GHz) region and 18 dB in broadband (1–8 GHz) region for 5 wt% of f-G loadings [174]. The EMI shielding efficiency of 21 dB over the frequency range of 8.2–12.4 GHz (X-band) was recorded for the 0.52 vol% of RGO filled epoxy composite [106]. Hsiao et al. showed the EMI SH effec- tiveness of ∼34 dB over the frequency range of 8.2–12.4 GHz for the RGO/PU composite [156].101 50 3.47 vol%Conductivity (S/m)10–2 2 φc = 0.09 vol% EMI SE (dB) 40 1.95 vol%10–5 0 t = 4.69±0.06 Log conductivity 30 1.36 vol% 10–8 –210–11 –4 20 0.87 vol% –6 9 10 11 12 1010–14 –8 Frequency (GHz) –3.5 –3.0 –2.5 –2.0 –1.5 0 Log (φ–φc) 810–17 (b) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5(a) rGO volume fraction (vol%)100 High pressure Compressive strength (MPa)120.0 108.4 MPa 2.75 GPa 3.00 Compressive modulus (GPa) 100.0 1.96 GPa 2.50Stress (MPa) 80 2.00 1.50 60 80.0 60.0 55.9 MPa Conventional pressure 40 20 40.0 0 20.0 1.00 012345678 (d) I II I II(c) Strain (%) Figure 3.23 (a) Electrical conductivity ver- compressive strength and modulus of the sus rGO loading for the s-rGO/PS composites; s-rGO/PS composites. I and II indicate com- (b) EMI SE as a function of frequency in X- posites molded under conventional and high band range for the composites; (c) typical pressure, respectively. (Reproduced from Ref. stress–strain curves of the s-rGO/PS com- [173] with permission from John Wiley and posites with 1.95 vol% rGO molded under Sons.) high and conventional pressure; and (d)

3.7 Application of Graphene Based Polymer Composites 993.7.4Flammability ReductionPolymers are prone to burn easily, so, it has become crucial to make them fireresistive as the horizon for polymer composite applications is increasing day byday. One of the adopted strategies to improve the fire safety of polymers is to incor-porate fire retardant additives. Being two-dimensional with high specific surfacearea and high thermal stability, graphene has the potential to be a fire retardant. Ifthe interfacial interaction between graphene and the polymer matrix is good thenit can act as barrier material that can prevent the release of heat and degrade smallgaseous molecules efficiently. Wang et al. used octa-aminophenyl polyhedral oligomericsilsesquioxanes(OapPOSS) for the simultaneous functionalization and reduction of GO. TheOapPOSS-rGO was incorporated into the epoxy matrix to study the thermalstability and flammability of the composites [175]. The peak heat release rate,total heat release, and CO production rate of OapPOSS-rGO/epoxy compos-ites was found to be 49%, 37%, and 58%, respectively. Furthermore, the onsetdegradation temperature was improved by 43 ∘C. The significant improvement insuch properties can be ascribed to the fact that OapPOSS-rGO forms insulatingcharred layer that may act as a barrier and prevent the underlying polymer fromburning. Bao et al. functionalized GO with char-catalyzing agents hexachlorocy-clotriphosphazene (HCCP) and hydroxyethyl acrylate (HEA) for the developmentof fire retardant graphene/PS composites [176]. The physical barrier effect ofFGO, its char formation ability, and polymer-FGO adhesion contributed to theimprovement of fire safety of the composites. The composites exhibited improvedpeak CO release rate (66% decreases), peak CO2 release rate (54% decrease), peakheat release rate (53% decrease), thermal degradation rate (30% decrease), andtotal heat release (38% decrease).3.7.5Automotive and AircraftsGraphene-based polymer composites are promising materials that could beused to make lighter, more fuel-efficient aircraft and automobile parts. With thefirst breakthrough in 1991 by Toyota Motor Co. significant research activitiesand development have occurred focusing polymer composites as promisingcandidates in the field of automotive sector. Polymer composites are promis-ing materials that exhibit a plethora of properties such as superb mechanicalproperties, high modulus and dimensional stability, excellent thermal properties,enhanced scratch and mar resistance, flame retardant properties, high impactresistance, and barrier properties, owing to which they can be perceived as analternative to metals in automotive and other applications [177]. The factorsthat stimulated the used of polymer composites as automotive components arereduction in vehicle’s weight that improves engine efficiency, reduction in CO2emission and better performance in terms of safety and comfort. As graphene has

100 3 Graphene Composites excellent properties and is the stiffest material if dispersed seamlessly in polymer matrix, composites so formed can be used to develop strong and lightweight automotive components. The University of Sunderland is leading a project with five research partners from Italy, Spain, and Germany through Graphene Flagship Competitive Call to develop novel graphene-based polymer composites material and their application in the development of lighter, greener, and improved crash- worthiness vehicle. The research done on GNP and CNT reinforced technology by UK based Haydale and The School of Engineering at Cardiff University have reported a 13% increase in compression strength and a 50% increase in compres- sion after impact performance, which is crucial in high-performance composites such as aircraft wings [178]. Another research group dispersed GNP at a load- ing of 15 wt% in material matrix, which resulted in electrical conductivity of 1000 S m−1 and which may find use in aircraft brake and automotive friction plate components [179]. 3.7.6 Turbine Blades Lately, wind power has attracted attention as one of the emerging sources of energy [180]. Those countries having widespread coastlines and high onshore wind activity are intensively pursuing new ways in the direction of harnessing wind energy and converting it to useful energy as this is pollution free, requires minimum raw materials and low cost of power generation [181]. Development of components such as wind blades in wind power technologies becomes practical only when the cost of production of material is minimum and material is strong/lightweight that consumes less energy to function. It has been found that graphene is more superior to CNTs in reinforcing the composites’ properties and may produce lightweight stiffer components than CNTs. For example, graphene-incorporated epoxy composites can be used to produce windmill blades or aircraft machineries. This will make the pursuit of lightweight, strong, and longer windmill blades a necessity, which will increase the generation of electricity from blades. 3.7.7 Others Graphene/polymer composites can be used to construct various submarine struc- tures such as bow domes on combatant submarines, periscope fairings on nuclear submarines, pressure hulls, and specialized sonar transducers. The composite materials can also be used in the light weight speed boat, life vest, swimming pool kits, and so on. The other possible applications of graphene/polymer composite materials include the manufacture of children entertainment equipments, highly durable industrial and domestic plastic products, scratch/abrasion resistant materials, anti corrosion paints, and so on.

References 107graphene sheet nanocomposites pre- oxide-polyelectrolyte nano membranes.pared by an in situ method. Macromol. ACS Nano, 4, 4667–4676.Chem. Phys., 210, 1247–1254. 109. Layek, R.K., Samanta, S., and Nandi,100. Kuila, T., Bose, S., Hong, C.E., Uddin, A.K. (2012) The physical properties ofM.E., Khanra, P., Kim, N.H., and Lee, sulfonated graphene/poly(vinyl alcohol)J.H. (2011) Preparation of function- composites. Carbon, 50, 815–827.alized graphene/linear low density 110. Saini, P., Kuila, T., Saha, S., andpolyethylene composites by a solu- Murmu, N.C. (2014) Graphene andtion mixing method. Carbon, 49, its nanocomposites for gas sensing1033 – 1051. applications, in Advanced Carbon101. Kalaitzidou, K., Fukushima, H., and Materials and Technology, ScrivenerDrzal, L.T. (2007) A new compounding Publishing LLC, pp. 501–534.method for exfoliated graphite- 111. Kuila, T., Bose, S., Mishra, A.K.,polypropylene nanocomposites with Khanra, P., Kim, N.H., and Lee, J.H.enhanced flexural properties and lower (2012) Effect of functionalized graphenepercolation threshold. Compos. Sci. on the physical properties of linear lowTechnol., 67, 2045–2051. density polyethylene nanocomposites.102. Kim, I.H. and Jeong, Y.G. (2010) Poly- Polym. Test., 31, 31–38.lactide/exfoliated graphite nanocom- 112. Rafiee, M.A., Lu, W., Thomas, A.V.,posites with enhanced thermal stability, Zandiatashbar, A., Rafiee, J., Tour, J.M.,mechanical modulus, and electrical and Koratkar, N.A. (2010) Grapheneconductivity. J Polym. Sci. B Polym. nanoribbon composites. ACS Nano, 4,Phys., 48, 850–858. 7415 – 7420.103. Zhang, H.B., Zheng, W.G., Yan, Q., 113. Wang, J., Wang, X., Xu, C., Zhang, M.,Yang, Y., Wang, J.W., Lu, Z.H., Yu, and Shang, X. (2011) Preparation ofG.Y., and Ji, Z.Z. (2010) Electrically graphene/poly(vinyl alcohol) nanocom-conductive polyethylene terephtha- posites with enhanced mechanicallate/graphene nanocomposites prepared properties and water resistance. melt compounding. Polymer, 51, Int., 60, 816–822.1191 – 1196. 114. Cano, M., Khan, U., Sainsbury, T., Neill,104. Song, P., Cao, Z., Cai, Y., Zhao, L., A.O., Wang, Z., McGovern, I.T., Maser,Fang, Z., and Fu, S. (2011) Fabrica- W.K., Benito, A.M., and Coleman,tion of exfoliated graphene-based J.N. (2013) Improving the mechanicalpolypropylene nanocomposites with properties of graphene oxide basedenhanced mechanical and thermal materials by covalent attachment ofproperties. Polymer, 52, 4001–4010. polymer chains. Carbon, 52, 363–371.105. Lin, Y., Jin, J., and Song, M. (2011) 115. Vadukumpully, S., Paul, J., Mahanta,Preparation and characterisation N., and Valiyaveettil, S. (2011) Flex-of covalent polymer functionalized ible conductive graphene/poly(vinylgraphene oxide. J. Mater. Chem., 21, chloride) composite thin films with3455 – 3461. high mechanical strength and thermal106. Yang, M., Hou, Y., and Kotov, N.A. stability. Carbon, 49, 198–205.(2012) Graphene-based multilayers: 116. Zhao, X., Zhang, Q., and Chen, D.critical evaluation of materials assembly (2010) Enhanced mechanical propertiestechniques. Nano Today, 7, 430–447. of graphene-based poly (vinyl alco-107. Zhao, X., Zhang, Q.H., Hao, Y.P., Li, hol) composites. Macromolecules, 43,Y.Z., Fang, Y., and Chen, D.J. (2010) 2357 – 2363.Enhanced mechanical properties 117. Hwang, S., Kang, D., Ruoff, R.S., Shin,of graphene based poly(vinyl alco- H.S., and Park, Y. (2014) Poly(vinylhol) composites. Macromolecules, 43, alcohol) reinforced and toughened with9411 – 9416. poly(dopamine)-treated graphene oxide,108. Kulkarni, D.D., Choi, I., Singamaneni, and its use for humidity sensing. ACSS.S., and Tsukruk, V.V. (2010) Graphene Nano, 8, 6739–6747.

108 3 Graphene Composites118. Istrate, O.M., Paton, K.R., Khan, U., 125. Wang, X., Hu, Y., Song, L., Yang, H.,O’Neill, A., Bell, A.P., and Coleman, Xing, W., and Lu, H. (2011) In situJ.N. (2014) Reinforcement in melt- polymerization of graphene nanosheetsprocessed polymer-graphene compos- and polyurethane with enhancedites at extremely low graphene loading mechanical and thermal properties.level. Carbon, 78, 243–249. J. Mater. Chem., 21, 4222–4227.119. Shim, S.H., Kim, K.T., Lee, J.U., and 126. Jin, J., Rafiq, R., Gill, Y.Q., and Song, M.Jo, W.H. (2012) Facile method to (2013) Preparation and characterizationfunctionalize graphene oxide and its of high performance of graphene/nylonapplication to poly(ethylene terephtha- nanocomposites. Eur. Polym. J., 49,late)/graphene composite. ACS Appl. 2617 – 2626.Mater. Interfaces, 4, 4184–4191. 127. Roy, S., Tang, X., Das, T., Zhang, L.,120. Liao, W.H., Yang, S.Y., Hsiao, S.T., Li, Y., Ting, S., Hu, X., and Yue, C.Y.Wang, Y.S., Li, S.M., Ma, C.C.M., (2015) Enhanced molecular level dis-Tien, H.W., and Zeng, S.J. (2014) persion and interface bonding at lowEffect of octa(aminophenyl) polyhedral loading of modified graphene oxideoligomeric silsesquioxane functionalized to fabricate super nylon 12 compos-graphene oxide on the mechanical and ites. ACS Appl. Mater. Interfaces, 7,dielectric properties of polyimide com- 3142 – 3151.posites. ACS Appl. Mater. Interfaces, 6, 128. Rafiee, M.A., Rafiee, J., Wang, Z.,15802 – 15812. Song, H., Yu, Z.Z., and Koratkar, N.121. Wang, J., Yang, S., Huang, Y., Tien, (2009) Enhanced mechanical propertiesH., Chin, W., and Ma, C.M. (2011) of nanocomposites at low graphenePreparation and properties of graphene content. ACS Nano, 3, 3884–3890.oxide/polyimide composite films with 129. Wajid, A.S., Tanvir, A.H.S., Das, S.,low dielectric constant and ultrahigh Irin, F., Jankowski, A.F., and Green,strength via in situ polymerization. J. M. (2013) High-performance pris-Mater. Chem., 21, 13569–13575. tine graphene/epoxy composites with122. Liao, W., Yang, S., Wang, J., Tien, enhanced mechanical and electricalH., Hsiao, S., Wang, Y., Li, S., Ma, properties. Macromol. Mater. Eng., 298,C.M., and Wu, Y. (2013) Effect 339 – 347.of molecular chain length on the 130. Ye, S., Feng, J., and Wu, P. (2013)mechanical and thermal properties Deposition of three-dimensionalof amine-functionalized graphene graphene aerogel on nickel foam asoxide/polyimide composite films pre- a binder-free supercapacitor elec-pared by in situ polymerization. ACS trode. ACS Appl. Mater. Interfaces, 5,Appl. Mater. Interfaces, 5, 869–877. 7122 – 7129.123. Cai, D., Jin, J., Yusoh, K., Rafiq, R., and 131. Tang, B., Hu, G., Gao, H., and Shi, Z.Song, M. (2012) High performance (2013) Three-dimensional graphenepolyurethane/functionalized graphene network assisted high performance dyenanocomposites with improved sensitized solar cells. J. Power Sources,mechanical and thermal properties. 234, 60–68.Compos. Sci. Technol., 72, 702–707. 132. Sun, H., Xu, Z., and Gao, C. (2013)124. Xiang, C., Cox, P.J., Kukovecz, A., Multifunctional, ultra-flyweight, syn-Genorio, B., Yan, D.P.H.Z., Peng, Z., ergistically assembled carbon aerogels.Hwang, C., Ruan, G., Samuel, E.L.G., Adv. Mater., 25, 2554–2560.Sudeep, P.M., Konya, Z., Vajtai, X., 133. Ni, Y., Chen, L., Teng, K., Shi, J., Qian,Ajayan, P.M., and Tour, J.M. (2013) X., Xu, Z., Tian, X., Hu, C., and Ma,Functionalized low defect graphene M. (2015) Superior mechanical prop-nanoribbons and polyurethane compos- erties of epoxy composites reinforcedite film for improved gas barrier and by 3D interconnected graphene skele-mechanical performances. ACS Nano, ton. ACS Appl. Mater. Interfaces, 7,7, 10380–10386. 11583 – 11591.

References 109134. Jiang, T., Kuila, T., Kim, N.H., and Lee, synergy effects. Chem. Mater., 27,J.H. (2014) Effects of surface-modified 2100 – 2106.silica nanoparticles attached graphene 143. Song, S.H., Park, K.H., Kim, B.H.,oxide using isocyanate terminated Choi, Y.W., Jun, G.H., Lee, D.J., Kong,flexible polymer chains on the mechan- B.S., Paik, K.W., and Jeon, S. (2013)ical properties of epoxy composites. J. Enhanced thermal conductivity ofMater. Chem. A, 2, 10557–10567. epoxy-graphene composites by using135. Jiang, T., Kuila, T., Kim, N.H., Ku, non-oxidized graphene flakes with non-B.C., and Lee, J.H. (2013) Enhanced covalent functionalization. Adv. Mater.,mechanical properties of silanized 25, 732–737.silica nanoparticle attached graphene 144. Wang, S., Tambraparni, M., Qiu, J.,oxide/epoxy composites. Compos. Sci. Tipton, J., and Dean, D. (2009) ThermalTechnol., 79, 115–125. expansion of graphene composites.136. Tang, L.C., Wan, Y.J., Yan, D., Pei, Macromolecules, 42, 5251–5255.Y.B., Zhao, L., Li, Y.B., Wu, L.B., Jiang, 145. Fang, M., Wang, K., Lu, H., Yanga,J.X., and Lai, G.Q. (2013) The effect of Y., and Nutt, S. (2010) Single-layergraphene dispersion on the mechan- graphene nanosheets with controlledical properties of graphene/epoxy grafting of polymer chains. J. Mater.composites. Carbon, 60, 16–27. Chem., 20, 1982–1992.137. Jia, J., Sun, X., Lin, X., Shen, X., Mai, 146. Araby, S., Zhang, L., Kuan, H., Dai,Y.W., and Kim, J.K. (2014) Exceptional J., Majewski, P., and Maa, J. (2013)electrical conductivity and fracture A novel approach to electrically andresistance of 3D interconnected thermally conductive elastomers usinggraphene foam/epoxy composites. graphene. Polymer, 54, 3663–3670.ACS Nano, 8, 5774–5783. 147. Patole, A.S., Patole, S.P., Kang, H.,138. Gao, J., Chen, F., Wang, K., Deng, H., Yoo, J., Kim, T., and Ahn, J. (2010) AZhang, Q., Bai, H., and Fu, Q. (2011) A facile approach to the fabrication ofpromising alternative to conventional graphene/polystyrene nanocompositespolyethylene with poly (propylene by in situ micro emulsion polymer-carbonate) reinforced by graphene ization. J. Colloid Interface Sci., 350,oxide nanosheets. J. Mater. Chem., 21, 530 – 537.17627 – 17630. 148. Salavagione, H.J., Martınez, G., and139. Layek, R.K., Kundu, A., and Nandi, Gomez, M.A. (2009) Synthesis of polyA.K. (2013) High-performance (vinyl alcohol)/reduced graphite oxidenanocomposites of sodium car- nanocomposites with improved ther-boxymethyl cellulose and graphene mal and electrical properties. J. Mater.oxide. Macromol. Mater. Eng., 298, Chem., 19, 5027–5032.1166 – 1175. 149. Kuila, T., Bose, S., Khanra, P., Kim,140. Zhang, Q., Xu, X., Li, H., Xiong, G., N.H., Rhee, K.Y., and Lee, J.H. (2011)Fisher, T.S., and Hu, H. (2015) Mechan- Characterization and properties of inically robust honeycomb graphene situ emulsion polymerized poly(methylaerogel multifunctional polymer com- methacrylate)/graphene nanocompos-posites. Carbon, 93, 659–670. ites. Compos. Part A, 42, 1856–1861.141. Li, Q., Guo, Y., Li, W., Qiu, S., Zhu, 150. Yun, Y.S., Bae, Y.H., Kim, D.H., Lee,C., Wei, X., Chen, M., Liu, C., Liao, J.Y., Chin, I., and Jin, H. (2011) Rein-S., Gong, Y., Mishra, A.K., and Liu, forcing effects of adding alkylatedL. (2014) Ultrahigh thermal conduc- graphene oxide to polypropylene.tivity of assembled aligned multilayer Carbon, 49, 3553–3559.graphene/epoxy composite. Chem. 151. Stankovich, S., Dikin, D.A., Dommett,Mater., 26, 4459–4465. G.H.B., Kohlhaas, K.M., Zimney, E.J.,142. Shtein, M., Nadiv, R., Buzaglo, M., Stach, E.A., Pine, R.D., Nguyen, S.B.T.,Kahil, K., and Regev, O. (2015) Ther- and Ruoff, R.S. (2006) Graphene-basedmally conductive graphene-polymer composite materials. Nature, 442,composites: size, percolation, and 282 – 286.

110 3 Graphene Composites152. Pham, V.H., Cuong, T.V., Dang, T.T., Detrembleur, C. (2011) Straight-Hur, S.H., Kim, B.K.E.J., Shin, E.W., and forward synthesis of conductiveChung, J.S. (2011) Superior conduc- graphene/polymer nanocompositestive polystyrene-chemically converted from graphite oxide. Chem. Commun.,graphene nanocomposite. J. Mater. 47, 2544–2546.Chem., 21, 11312–11316. 160. Chen, Z., Ren, W., Gao, L., Liu, B.,153. Qi, X.Y., Yan, D., Jiang, Z., Cao, Y.K., Pei, S., and Cheng, H.M. (2011) Three- Yu, Z.Z., Yavari, F., and Koratkar, N. dimensional flexible and conductive (2011) Enhanced electrical conductivity interconnected graphene networks in polystyrene composites at ultra-low grown by chemical vapour deposition. graphene content. ACS Appl. Mater. Nat. Mater., 10, 424–428. Interfaces, 3, 3130–3133. 161. Yoonessi, M. and Gaier, J.R. (2010) Highly conductive multifunctional154. Zheng, D., Tang, G., Zhang, H., Yu, graphene polycarbonate nanocompos- Z., Yavari, F., Koratkar, N., Lim, S., ites. ACS Nano, 4, 7211–7220. and Lee, M. (2012) In situ thermal 162. Vasileiou, A.A., Kontopoulou, M., and reduction of graphene oxide for high Docoslis, A. (2014) A Noncovalent electrical conductivity and low per- Compatibilization Approach to Improve colation threshold in polyamide 6 the Filler Dispersion and Properties nanocomposites. Compos. Sci. Technol., of Polyethylene/Graphene Compos- 72, 284–289. ites. ACS Appl. Mater. Interfaces, 6, 1916 – 1925.155. Wu, C., Huang, X., Wang, G., Lv, L., 163. Qi, X.Y., Yan, D., Jiang, Z., Cao, Y.K., Chen, G., Li, G., and Jiang, P. (2013) Yu, Z.Z., Yavari, F., and Koratkar, N. Highly conductive nanocomposites (2011) Enhanced electrical conductivity with three-dimensional, compactly in polystyrene composites at ultra-low interconnected graphene networks via graphene content. ACS Appl. Mater. a self-assembly process. Adv. Funct. Interfaces, 3, 3130–3133. Mater., 23, 506–513. 164. Hyunwoo, H., Zhang, G., Xiao, L., Wang, H., Zhang, Q., and Zhao, Z.156. Hsiao, S., Ma, C.M., Liao, W., Wang, Y., (2012) Preparation and electrical Li, S., Huang, Y., Yang, R., and Liang, conductivity of graphene/ultrahigh W. (2014) Lightweight and flexible molecular weight polyethylene com- reduced graphene oxide/water-borne posites with a segregated structure. polyurethane composites with high Carbon, 50, 4596–4599. electrical conductivity and excellent 165. Kim, H., Miura, Y., and Macosko, electromagnetic interference shield- C.W. (2010) Graphene/polyurethane ing performance. ACS Appl. Mater. nanocomposites for improved gas bar- Interfaces, 6, 10667–10678. rier and electrical conductivity. Chem. Mater., 22, 3441–3450.157. Zhang, H.B., Zheng, W.G., Yan, Q., 166. Mittal, V. and Chaudhry, A.U. (2015) Yang, Y., Wang, J.W., Lu, Z.H., Ji, Polymer–graphene nanocomposites: G.Y., and Yu, Z.Z. (2011) Electrically effect of polymer matrix and filler conductive polyethylene terephtha- amount on properties. Macromol. late/graphene nanocomposites prepared Mater. Eng., 300, 510–521. by melt compounding. Polymer, 51, 167. Yoonessi, M., Shi, Y., Scheiman, D.A., 1191 – 1196. Colon, M.L., Tigelaar, D.M., Weiss,158. Shen, Y., Jing, T., Ren, W., Zhang, J., Jiang, Z.G., Yu, Z.Z., and Dasari, A. (2012) Chemical and thermalreduction of graphene oxide and its R.A., and Meador, M.A. (2012)electrically conductive polylactic acid Graphene polyimide nanocompos-nanocomposites. Comp. Sci. Technol., ites; thermal, mechanical, an high-72, 1430–1435. temperature shape memory effects.159. Vuluga, D., Thomassin, J.M., ACS Nano, 6, 7644–7655.Molenberg, I., Huynen, I., Gilbert, 168. Yang, Y., Bolling, L., Priolo, M.A., andB., Jerome, C., Alexandrea, M., and Grunlan, J.C. (2013) Super gas barrier

References 111 and selectivity of graphene oxide- 175. Wang, X., Song, L., Yang, H., Xing, polymer multilayer thin films. Adv. W., Kandol, B., and Hu, Y. (2012) Mater., 25, 503–508. Simultaneous reduction and surface169. Liu, H., Kuila, T., Kim, N.H., Kud, functionalization of graphene oxide B., and Lee, J.H. (2013) In situ syn- with POSS for reducing fire hazards in thesis of the reduced graphene epoxy composites. J. Mater. Chem., 22, oxide–polyethyleneimine composite 22037. and its gas barrier properties. J. Mater. Chem. A, 1, 3739–3746. 176. Bao, C., Guo, Y., Yuan, B., and170. Wu, Z., Chen, X., Zhu, S., Zhou, Z., Lei Song, Y.H. (2012) Functionalized Yao, Y., Quan, W., and Liu, B. (2013) graphene oxide for fire safety appli- Enhanced sensitivity of ammonia cations of polymers: a combination sensor using graphene/polyaniline of condensed phase flame retardant nanocomposite. Sens. Actuators B: strategies. J. Mater. Chem., 22, 23057. Chem., 178, 485–493.171. Yang, G., Lee, C., Kim, J., Ren, F., and 177. V. Patel, Y. Mahajan, Polymer Pearton, S.J. (2013) Flexible graphene- nanocomposites drive opportunities based chemical sensors on paper in the automotive sector. substrates. Phys. Chem. Chem. Phys., 15, 1798. 178. Bai, H., Sheng, K., Zhang, P., Li, composite-materials-study-can-boos C., and Shi, G. (2011) Graphene (accessed 29 February 2016). oxide/conducting polymer compos- ite hydrogels. J. Mater. Chem., 21, 179. Song L., Guo, J., and Zhamu, A. 18653 – 18658. (2009) Highly conductive nano scaled173. Yan, D., Pang, H., Li, B., Vajtai, R., graphene plate nanocomposites. US Xu, L., Ren, P., Wang, J., and Li, Z. Patent 7566410. (2015) Structured reduced graphene oxide/polymer composites for ultra- 180. Grujicic, M., Arakere, G., and efficient electromagnetic interference Pandurangan, B. (2010) Multidis- shielding. Adv. Funct. Mater., 25, ciplinary design optimization for 559 – 566. glass-fiber epoxy-matrix composite174. Eswaraiah, V., Sankaranarayanan, V., 5MW horizontal-axis wind turbine and Ramaprabh, S. (2011) Functional- blades. J. Mater. Eng. Perform., 19, ized graphene/PVDF foam composites 1116 – 1127. for EMI shielding. Macromol. Mater. Eng., 296, 894–898. 181. Al-Qabandi, O., Silva, A.D., Al-Enezi, S., and Bassyouni, M. (2014) Synthesis, fabrication and mechanical charac- terization of reinforced epoxy and polypropylene composites for wind turbine blades. J. Reinf. Plast. Compos., 33, 2287–2299.

1134Graphene in Lithium-ion BatteriesCyrus Zamani4.1IntroductionEnergy is one of the main concerns of our era. There are several reasons for this,among them, the increased population of the planet and, therefore, the need formore energy, the high level of pollution, and global warming. On one hand, theworldwide energy demand is growing continuously and, on the other hand, envi-ronmental issues related to the current energy resources and technologies raise abig concern about the future of human beings. To cope with the existing problemsand design a better future for the inhabitants of this planet, researchers are work-ing actively in close collaboration with governments and policy makers. Figuresand statistics show that still about 80% of the global energy demand is coveredby fossil fuels and the same trend is expected until 2040 (Figure 4.1) althoughthe demand will rise to one-third (from 500 EJ in 2008 to about 750 EJ by 2040)[1]. However, the conventional fuels are expected to run out in the near future(Table 4.1) [2] and, therefore, there is a great interest on using other sources ofenergy to answer the growing energy demand although the cumulative $7.4 tril-lion by 2040 looks to be still insufficient. Looking at the global investment in the field of renewable energies, its impor-tance is disclosed. According to UNEP [3], in 2014, $270.2 billion was investedon renewable energies and fuels, about 17% higher than 2013. Developingcountries such as Chile, Mexico, Kenya, South Africa, and Turkey are amongthose who spent in the order of billion dollars. Jordan, Uruguay, Panama, ThePhilippines, and Myanmar also invested on renewable energies (in the range of$500 million to $1 billion). In total, about 48% of GW capacities installed in 2014were renewables increasing the share of electricity from renewable sources to9.1%. Of these, many installations are on-site and generate off-grid electricity, afact that highlights the need to have storage devices for the produced electricity.High-performance batteries are therefore required for such systems distributedworldwide. Advanced lithium-ion batteries are considered to be the bestcandidates for such applications. Thanks to nanomaterials and nanotechnology,Graphene Technology: From Laboratory to Fabrication,First Edition. Edited by Soroush Nazarpour and Stephen R. Waite.© 2016 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2016 by Wiley-VCH Verlag GmbH & Co. KGaA.

114 4 Graphene in Lithium-ion Batteries 40 36 33Consumption (%) 18 18 23 8 8 1 1 10 7 8 0 29 7 23 27 1990 2013 2040Petroleum & other liq. Coal Nuclear Liq. biofuels Renewables Natural gasFigure 4.1 Evolution of energy consumption by resource type [1].Table 4.1 The availability of fuels [2]. Reserves Resources (years) (years)Oil (conventional) 43 67Oil (conventional 62 157+ Non-conventional)Gas (conventional) 64 149Gas (conventional 64 756+ Non-conventional)Hard coal 207 1425Soft coal 198 1264Uranium 42 527these devices are gaining more and more importance in today’s electronics. Thischapter deals with the role of graphene materials in lithium-ion batteries.4.2Renewable EnergiesEnergy sources can be categorized in different ways. The most widely acceptedone is, perhaps, according to their potentiality in being regenerated or not. Inthis aspect, energies are classified as non-renewables and renewables. The former

4.3 Batteries, What are They? 115mainly consists of fossil fuels (i.e., oil, gas, and coke) while the latter includesenergy from sources such as sunlight, wind, biomass, and so on. Renewable ener-gies are therefore named as clean or green energy as they do not have any negativeimpact on the environment and living things. There exist numerous types of renewable energy sources. Basically, any sourceof energy that can be renewed in a very short period of time (in contrast to fossilsources that need thousands of years to form), falls in this category. Solar, tidal,biomass, wind, geothermal, and hydroenergy are common examples althoughthere exist other sources that are less common (e.g., energy derived from algae,sugar, garbage, human fat, etc.,). With all their advantages, the global share ofrenewable energies is about 15% only, since these are not competitive in cost.No doubt, these emerging technologies come with their business risks, but theyare accompanied with opportunities too. The fact is that if well established, therenewable energy industry will re-design the future of our planet and that is whythe technology associated with renewable energies is considered as one of thegreat game players in tomorrow’s technology sector. Whatever be the source ofenergy, there is always a need to store it for further use. This is more highlightedwhen we take into account the scattered nature of renewable energy productionsites since, for example, wind turbines are in need of storing the produced energyfrom wind and this task is done by batteries implemented inside the machine orin the vicinity of the wind farm.4.3Batteries, What are They?Batteries are devices used by mankind for energy storage. The energy is storedin the form of chemical energy, which is later converted to electricity to puta machine into operation. The oldest battery known is the Parthian battery(Figure 4.2) and dates back to the Sassanid period (about 250 BC) of the PersianEmpire. Given this, batteries have a long history though the first mass-producedbattery was released in 1802. Since then, different types of batteries (i.e., lead-acid,NiCd, lithium-ion, and NiMH) have been introduced (see Table 4.2 for a historicalbackground on batteries and their evolution).Neagative electrode Positive electrode (copper) (iron) Battery case ElectrolyteFigure 4.2 Parthian battery.

116 4 Graphene in Lithium-ion Batteries Table 4.2 Evolution of batteries.Year/period Battery type Inventor250 BC Parthian battery Sassanid Empire1660 Electrical machine Otto von Guericke1744 Leyden jar Ewald Georg von Kliest1745 – 1827 Electric pistol Alessandro Volta1791 Galvanic electricity Luigi Galvani1800 Electrolysis Sir HumphryDavi1802 First electric battery (mass production) William Cruickshank1859 First rechargeable battery (lead acid) Gaston Planté1899 First nickel–cadmium battery Waldmar Junger1912 First Li-ion battery G.N. Lewis1947 Sealed NiCd cell Georg Neumann1970s Non-rechargeable Li-ion battery1990s Nickel-metal hydrides (NiMH) Sony Corporation1991 Commercial rechargeable Li-ion battery4.4Lithium-ion BatteriesLithium-ion batteries are a family of batteries that operate based on the interca-lation of electrode structure with lithium ions. Today’s batteries are rechargeablewith lithium ions moving from the cathode (positive electrode) to the anode (nega-tive electrode) during charging, as shown in Figure 4.3. The flow of ions is reversedduring discharging (i.e., utilization) of the battery. ee Li+Copper collectorAluminum collector Li+ Li+ Li+Figure 4.3 Principles of lithium-ion battery operation.

4.5 Anodes, Cathodes, and Electrolytes 117Anode Separator CathodeFigure 4.4 Schematic structure of a lithium-ion battery. In comparison to other battery types, lithium-ion batteries offer higher energydensity, longer cycle life, low self-discharge rate, and no memory effect. Theseadvantages qualify lithium-ion batteries as the best choice for numerous appli-cations including portable electronics. It is believed that lithium-ion batteries areenvironmentally friendly since no toxic metal is used in their structure and thismakes them the best option for electric vehicles although some concern on theCO2-emission of large battery packs is raised [4]. The main disadvantage is thatthese batteries are in need of special safety attention. Short-circuiting, overcharg-ing, and overheating can arise which may lead to battery explosion. This makes avery careful design necessary. All batteries have three parts in common, that is, anode, cathode, and elec-trolyte. Commercially available batteries are more sophisticated in structure.Basically, two common constructions are available: cylindrical and prismatic.Active materials (for cathode and anode) are deposited on aluminum and copperfilms. A four-layer spiral structure comprising the anode, cathode, and separatorsis put inside a package. In addition, there exist a variety of safety protection mech-anisms including controller IC (to prevent overcharging), control switches (forcharge/discharge interruption), temperature fuse (to prevent abnormal heating),and gas release control valve that is used to prevent battery explosion in case theinternal gas pressure exceeds a critical value. The whole assembly is packed usinga material (usually resin), which makes disassembling rather difficult (Figure 4.4).4.5Anodes, Cathodes, and ElectrolytesCommercially available lithium-ion batteries can be different in their electrodematerials as well as electrolytes. Table 4.3 lists the common materials used asanode, cathode, and electrolyte. However, difficulties in fabrication of lithium-ion batteries with satisfactory cycling properties have made researchers to add

118 4 Graphene in Lithium-ion Batteries Table 4.3 Common materials used in lithium-ion batteries.Anode Graphite, hard carbon, LTOa)Cathode LCO, LFP, LMO, NCA, NMCb)Electrolyte Alkyl carbonates (EC, DMC, EMC, DEC), LiPF6c)a) LTO: Li4Ti5O12.b) LCO, LiCoO2; LFP, LiFePO4; LMO, LiMn2O4; NCA, LiNiCoAlO2; NMC, LiNiMnCoO2.c) EC, ethylene carbonate (EC); DMC, dimethyl carbonate; EMC, ethyl methyl carbonate; DEC, diethyl carbonate.alloying compounds to the electrodes and create nanostructures for improvedperformance [5–7].The electrolyte usually comprises lithium salt and transports lithium ionsbetween the cathode and anode. An example of the electrolyte materials is LiPF6in a solution to which other additives are also added for higher stability. Theelectrolytes are optimized for the specific anode and cathode materials used inbattery structure.For positive electrode, a number of materials are already in use. They usuallyhave nickel, manganese, or cobalt in their crystal structure. Today, the maingroups available in the market are lithium cobalt oxide (LCO), nickel manganesecobalt (NMC), lithium manganese oxide (LMO), and lithium-iron phosphate(LFP). The latter is gaining more importance due to its higher safety and lowdischarge rate.In the case of negative electrode, so far, carbon has always been in use althoughrecently, other materials such as silicon [8] and lithium titanate (Li4Ti5O12) havealso been announced [9].For batteries with LCO (cathode), and graphite (anode), the chemical reactionswould be: LiCoO2 Charge Li1−xCoO2 + xLi+ + xe− −−−−−−→ C + xLi+ + xe− Charge CLix −−−−−−→ Charge LiCoO2 + C −−−−−−→ Li1−xCoO2 + CLix4.6Carbon MaterialsGraphite is the most common allotrope of carbon: an electrical conductor that hasthe highest stability among all carbon allotropes. The other well-known memberof the family is diamond, a very hard material with light dispersion properties,which introduces the material as a good candidate for mechanical applicationsand in the manufacture of jewelry. In addition to graphite and diamond, there exist

4.7 Graphite 119several other allotropes such as Lonsdaleite, nanotubes, fullerenes, graphene, andamorphous carbon.4.7GraphiteGraphite (Figure 4.5) – the most stable form of carbon – is known for its planarstructure and relatively pure carbon composition. The material shows a combi-nation of physical and chemical features. Graphite is electrically (and thermally)conductive. It is a corrosion-resistant material and so it is known as a chemicallypassive material (resistant to chemical attack). All these characteristics along withits low density, 2.266 g cm−3 (even lighter than aluminum) and high strength andstiffness make the material so special that in a wide variety of applications – fromtraditional furnace linings to advanced electrode materials for batteries – graphiteis considered as one of the best candidates. In renewable energies and devices,fuel cells also benefit from graphite features in their bi-polar planes. In fact, thematerial is revolutionizing the world. The increasing consumer demand (bothin traditional as well as advanced applications) forced global powers to classifygraphite as a critical strategic material [10]. In lithium-ion batteries, both natural and synthetic graphite (obtained frommines and coke, respectively) are used. Natural graphite needs purification forbeing considered as an anode material for lithium-ion batteries. Currently, naturalgraphite is preferred for such applications since the material offers higher purity(99.9 vs. 99.0 in synthetic product) and higher degree of crystallinity (improvedthermal and electrical conductivity) at a lower cost. As anode material, the mor-phology of graphite should be spherical (or potato type), Figure 4.6 [10].Figure 4.5 Structure of graphite.

120 4 Graphene in Lithium-ion Batteries Figure 4.6 Battery grade graphite [11]. 4.8 Graphene Graphene is one of the greatest discoveries of materials scientists [12]. Discov- ered by A. Geim and K. Novoselov of the University of Manchester in 2004, it is a two-dimensional carbon allotrope with a hexagonal lattice (Figure 4.7), which can be prepared through a number of physical and/or chemical methods [13–15]. Graphene is the thinnest compound known so far comprising a single layer of carbon atoms that bond together (with sp2 hybridization) forming a tight honey- comb structure [16]. The bond length is 0.142 nm and the interlayer spacing (when stacked to form graphite) is 0.335 nm. Physically, graphene is the strongest mate- rial [17] with the highest room temperature and thermal and electrical conduc- tivity (Table 4.4). Needless to say, these properties depend on the synthesis route and the structure of the material. For example, mechanically exfoliated products exhibit room temperature charge carrier mobility >105 cm2 V−1 s−1 while that of chemically exfoliated material (from graphene oxide) is about 1 cm2 V−1 s−1 [18]. Moreover, the material is anisotropic with highest in-plane conductivity while in Figure 4.7 Graphene structure.

Table 4.4 Selected properties of graphene. 4.9 Graphene in Lithium-Ion Batteries 121Property ValueDimensional Thickness 3.35 A (varies depending on how layersMechanical are stacked) Specific surface area 2630 m2 g−1Electronic Stiffness 150 × 106 PSiThermal Young’s modulus 1100 GPa Fracture strength 125 GPa Hardness Highest among all carbon allotropes Tenacity Flexible Electron transport (RT) Zero-gap semiconductor, semi-metal >15 000 cm2 V−1 s−1 (nearly identical Electrical conductivity electron & hole transport) 2000 S cm−1 Thermal conductivity (RT) (4.84 ± 0.44) × 103 to (5.30 ± 0.48) × 103Optical Transmittance W m−1 K–1 97.7%the direction normal to the sheets, these properties drop to much lower values[19, 20]. On the other hand, with its basic hexagonal structure, graphene can actas the building block of other carbonaceous materials such as carbon nanotubesand fullerenes [21].4.9Graphene in Lithium-Ion BatteriesGraphene is a carbonaceous material with fantastic physical and chemical prop-erties: highest electrical and thermal conductivity among all materials knownto man, mechanical flexibility, thermal and chemical stability, extremely largesurface area, and presence of proper sites for chemical adherence of materials.No other material with such a combination of properties can be found. All thesefeatures have attracted the interest of researchers for employing graphene inelectrochemical devices such as lithium-ion batteries and fuel cells [13, 22, 23].In lithium-ion batteries, for instance, graphene is a promising material sincethese properties make the insertion and extraction of Li+ ions faster, easier,and reversible. In comparison to the rest of the members of the carbon family,graphene (in the form of graphene oxide and reduced graphene oxide (RGO)) hasthe advantage of possessing functional groups on the surface and edges. Whenmaking composites of graphene and metal oxides, for instance, the distributionof oxide particles as well as their size and shape are affected by these functionalgroups.

122 4 Graphene in Lithium-ion Batteries 4.10 Graphene in Anodes As a member of the carbon family, researchers started to explore the possibility of using graphene to improve the durability and charge capacity of lithium-ion battery anodes. In the case of graphite – which has been traditionally used as anode – the theoretical capacity is limited to 372 mA h g−1, which is far below the needs of new devices (this capacity corresponds to intercalation of one lithium ion to six carbon atoms, LiC6) [24]. On an industrial scale, this deficiency is more pronounced since the microstructural features of carbon and graphite, and their crystallinity and morphology affect the functionality of the anode. The storage capacity of commercial devices is about 310 mAhg−1 [25]. To go beyond this limit, different scenarios have been designed (Figure 4.8). Modification of graphite by mild oxidation of the material induces micro porosities inside the material, thereby causing a better intercalation. On the other hand, a dense oxide layer is formed, which resists against electrode decompo- sition. The method, however, lacks reproducibility and uniformity in the final products, which is considered as a disadvantage [26]. The second strategy is based on creating composite structures using metals or metal oxides in a graphite matrix. Silver, nickel, and metal oxides such as SnO2 are among the materials researched for this purpose. The idea behind this has been benefiting from the properties of the second phase (e.g., higher conductiv- ity of silver and faster transport) to overcome the limitations of graphite. Making composite anodes is reported to enhance the electrode properties in several ways Process Anode Structure Anodemodification material modification material modification replacementOxidizing Metals and metal Anode coating CNTs & buckygraphite oxide composites paperNanostructuring Carbon Morphology Silicon allotropes enhancementFigure 4.8 Different scenarios designed to improve anode performance.

4.10 Graphene in Anodes 123including higher electrical conductivity, preventing the anode from decomposi-tion, and creating hosts for lithium. The disadvantage, however, is the irreversiblecapacity noted in the first cycle [26]. Coating of anode with a polymer is anotherapproach through which the anode material is coated with a polymer to improveits electrochemical performance. Both electronically conductive as well as ioni-cally conductive polymers have been tried [27]. Despite all efforts made on the anode aspect of the lithium-ion batteries, stillnegative electrode is considered as a bottleneck restricting the device perfor-mance. Here, employing other nanostructured carbon allotropes would be a goodstrategy. While conventional carbons in the range of 0.1–10μm do not meet thedemands, carbon nanotubes with diameters less than 100 nm could be a solution.The main advantage of carbon nanotubes is their ballistic electronic transportthat ensures superior electrical conductivity [28, 29]. The major drawback ofCNTs is limited intercalation since the only available sites are nanotube ends [30].Moreover, nanostructuring can facilitate the charge/discharge processes since,according to Fick’s second law of diffusion, a shorter diffusion distance results inshorter diffusion times [6]. While in graphite anodes, the intercalation is limited to the formation of LiC6[(0002) planes], graphene sheets expose both surfaces to lithium ions, thus pro-viding the possibility of Li2C6 stoichiometry formation [24] and increasing thetheoretical capacity to 740 mA h g−1. Moreover, graphene offers additional sites(defects and imperfections) for ion storage, which therefore, ensure higher capac-ity for future batteries. Currently, various forms of graphene-based materials areexamined for their usage in lithium-ion battery anodes. The major categories arelisted in Figure 4.9 being graphene [22, 31], RGO [32], doped graphene [33, 34],and defected graphene sheets (DGNs) [35]. In comparison to graphene, RGO is always accompanied with a large amountof topological defects (mainly dislocations) and lack of crystallinity [36].Defected Dopedgraphene graphene Reduced COOH COOH COOH OH B GO D O OO OGraphene HO COOH O OO O COOH O OH COOHFigure 4.9 Graphene-based materials in lithium-ion battery anodes.

124 4 Graphene in Lithium-ion Batteries N+ N N N N Pyridinic N −OPyridinic+N–O− H H Quaternary N Pyrrolic NFigure 4.10 Bonding in nitrogen-doped graphene. (Reprinted from Ref. [33] with permissionfrom the American Chemical Society, Copyright 2012.)Therefore, it is expected to make use of such a structure in obtaining higherstorage capacities. Heteroatom doping of graphene sheets with elements suchas nitrogen, boron, phosphorus, and sulfur is also reported [21, 34]. Doping isdone to improve graphene properties for devices such as lithium-ion batteriesand supercapacitors. High-power and high-energy batteries featuring enhancedreversible charge/discharge are fabricated through doping. Wu et al. presentedanode materials with reversible capacities above 1040 mA h g−1 [21]. Severalnitrogen doping routes including CVD, thermal treatment, solvothermal, arc-discharge, plasma, and N2H4 treatment are reported as a result of which, threebonding configurations are obtained: pyridinic N, pyrrolic N, and graphitic N(Figure 4.10) [33]. Another way to create defects in graphene sheets is patterningtheir surface through processes such as lithography, electro plasma etching, andcatalysis. Creating these defects provides the opportunity for the Li+ ions topenetrate into the structure, which is equivalent to higher capacities. Following acobalt-catalyzed gasification strategy, Hu et al. obtained DGNs with a reversiblespecific capacity of about 1009 mA h g−1 [35]. All the above-mentioned graphene-based materials have shown great promisein enhancing lithium-ion battery operation. However, there remains a big chal-lenge: the high tendency of graphene sheet to agglomeration and re-stacking dueto van der Waals forces followed by a drop in electrical properties. To solve thisserious issue, graphene composites are under research [37–45]. The behavior ofgraphene composites is a combination of the properties of both graphene and theother phase. In this respect, composites comprising graphene and metal oxideshave shown great promise (Figure 4.11). The capacity of metal oxides is more thantwo times larger than that of graphite material and the reaction. For this, threemechanisms (conversion reaction, Li-alloy reaction, and Li insertion/extractionFigure 4.11 Graphene/MO composite formation. (Reproduced with permission fromRef. [21].)

4.10 Graphene in Anodes 125 Anchored Model Nano-sized oxides anchoring on graphene for LIBs Wrapped Model Graphene-wrapped metal oxide particles Encapsulated Model Graphene encapsulates metal oxide particles Sandwich like Model MO/Graphene/MO sandwiches with graphene as template Layered Model Aligned layers of graphene/MO Mixed Model 3D networks of graphene and metal oxidesFigure 4.12 Structural mechanisms for compositing graphene with metal oxides for lithiumbattery applications. (Reproduced with permission from Ref. [23].)reaction) are proposed and numerous metal oxides are tested. Examples are NiO,SnO2, TiO2, CuO, Cu2O, MnO, Fe2O3, and Fe3O4 for which, six structural mech-anisms are proposed (Figure 4.12). In all mechanisms, functional groups on thegraphene surface and its edges result in improved chemical functionality and bet-ter compatibility to oxide particles. Based on their morphological features, onthe other hand, metal oxide particles offer a high capacity. With these compositestructures the following features are expected to be improved [21]:• Capacity/capacitance• Cyclability• Rate capability• Energy/power density In addition to metal oxides, graphene/non-metal composites are also attractinggreat interest for being used as anode materials for lithium-ion batteries. Inthis respect, silicon (Si) anodes are introduced as one of the promising optionsfor future batteries. Thanks to its potential in forming a stable alloy containing4.4 lithium per silicon, the theoretical capacity of silicon is more than 10 timesthat of graphite being 4200 mA h g−1. As anodes in lithium-ion batteries, siliconoffers a variety of advantages over their rivals: high theoretical capacity, low dis-charge rate, no memory effect, environmentally friendly, and natural abundance.Moreover, silicon anodes show a voltage plateau at 370 mV against Li/Li+, which

126 4 Graphene in Lithium-ion Batteries means stable operating voltage [6]. Despite all these properties that introduce silicon anodes as outstanding assets for lithium-ion energy storage, the intrinsic drawbacks of the material put obstacles in the way of its utilization in batteries. One major problem of silicon is its large expansion/contraction during lithia- tion/delithiation [46]. The expansion of silicon up to 400% of its original volume results in anode failure due to cracking. Cracking and loss of contact between anode and its substrate (usually copper) as well as formation of unstable interfaces with solid electrolyte affect the cyclabilty and rate performance of the anode, thus reducing the device functionality in a drastic manner. Therefore, researchers have tried to overcome this deficiency through creating nanostructured silicon anodes, which show a better response to volume fluctuations. For obtaining a better cyclability and performance, silicon composites have been synthesized [8]. In this way, different materials are explored such as Si/CNTs [47, 48, 49], Si/C [50, 51], Si/graphene [52–55], Si/polymer [56–58], and so on. Although these materials reveal significant improvement in anode functionality, there still remain issues to be solved: production cost, the complexity of the processes employed, and the use of dangerous materials during synthesis. In this scenario, silicon/RGO composites and Si/RGO/C composites with their three-dimensional structure offer improved electrical conductivity and higher stability [59]. Such structures combine the three-dimensional high theoretical capacity of silicon with flexibility and conductivity of graphene and will constitute the future batteries for portable and flexible electronic devices. 4.11 Graphene in Cathodes Parallel to anode materials, new cathode materials for lithium-ion batteries have also been under development [9]. Basically, the same strategies (followed in anode enhancement) can be found here. In comparison to anodes, cathodes in commer- cially available batteries are more complex in structure. These are lithium-based compounds such as lithium cobaltite (LiCoO2, layered structure), lithium nickel cobalt aluminum oxide (LiNi0.8Co0.15Al0.05O2 or NCA, layered structure), lithium nickel manganese cobalt oxide (LiNi1/3Co1/3Mn1/3O2 or NMC, layered structure), lithium manganese oxide (LiMn2O4 or LMO, spinel structure), and lithium ferro phosphate (LiFePO4 or LFP, olivine structure) [60, 61]. Each material comes with its advantages and shortages. For instance, LMO and LFP are low-cost materials but their capacity is also low. On the other hand, NMC and NCA offer a better capacity but they are expensive due to the high cost of Ni and Co. The capacity of all these electrodes falls below 200 mA h g−1, which is lower than the capacity of the negative electrode. Therefore, works on lithium-ion battery cathodes are mainly aimed at improving the electrical properties of the cathode. There is also some concern about the structural instabilities and loss of capacity and cyclability. Therefore, composites of different materials with graphene-based products have been synthesized and electrically characterized

4.12 Graphene in Other Types of Lithium Batteries 127[62, 63]. Using graphene, capacities beyond the theoretical capacity of existingcathodes are obtained.4.12Graphene in Other Types of Lithium BatteriesLithium batteries come in different types including lithium-ion, lithium-polymer,lithium-sulfur, and lithium-air (the new member of the family) all of which workwith lithium ion as charge carrier. Sodium-ion batteries (NIBs), on the otherhand, use sodium ions as charge carriers. The advantage of NIBs over lithium-ionis that sodium salts are cheap and abundant. In addition, the current trend in theconsumption of lithium in lithium-ion batteries has created a great demand forthis material, which is, unfortunately, limited in abundance. However, comparedto lithium-ion batteries, NIBs offer lower charge capacity (115 mA h g−1 at 3.6 V).Due to the fact that sodium and lithium are a lot similar in a number of properties,it is expected that these batteries act as alternatives to lithium-ion batteries [64].However, the larger size of sodium ion (1.02 Å compared to 0.76 Å of lithiumions) has a limitation on the application of these batteries since it affects thecycling process negatively [65]. Similar to lithium-ion batteries, NIBs make useof carbon as the anode material although composite materials such as SnSexand compounds such as sodium titanate have also been introduced [66, 67]. Thecathode side, on the other hand, usually is based on sodium iron phosphate oreven lithium iron phosphate. Obviously, the performance of the device depends highly on the materials andtheir structure. Thus, as the first and main strategy in cathode enhancement, coat-ing cathode with conductive carbon layer and its effect on battery operation hasbeen the focus of many researchers [68, 69]. In NIBs, special attention has beengiven to positive electrodes so far and a number of materials are proposed includ-ing sulphides, phosphates, fluorides, oxides, and sulphates [70]. Recently, a growing interest has been observed on using carbonaceous mate-rials including graphene to fabricate high-capacity electrode materials for thesebatteries [70–77]. Whether as additive or the mainframe, graphene-based mate-rials have proved to be efficient in improving battery performance. However, thisfield needs more investigation in order to obtain high-capacity devices that theworld of electronics demands.SummaryApplication of graphene-based materials in advanced lithium-ion batteries wasreviewed. A systematic approach was followed starting from the history ofbatteries to currently available types emphasizing on lithium-ion batteries as thestorage devices for portable electronics as well as future hybrid and electric cars.Current challenges in developing batteries were noted and the role of graphene

128 4 Graphene in Lithium-ion Batteries in improving the device performance was highlighted. The tremendous work on graphene and its derivatives in enhancing the lithium-ion battery operation and the huge interest in following the same strategy in sodium-ion devices – the new member of the family – reveals the importance of this material, which is believed to be a star in the sky of materials science and engineering.References1. Annual Energy Outlook 2015 with Pro- of electrospun Li4Ti5O12 for lithium-jections to 2040 (AEO2015) (2015) s.l.: ion batteries. Electrochim. Acta, 55,U.S. Energy Information Administration, 5813 – 5818.April. 10. About Graphite, http://2. Energy Study 2014, Reserves, Resources Availability of Energy Resources 01-24-57 (accessed 22 February 2016).(2014) BGR. Hannover: s.n.. 11. Crowe, P. Hybrid Cars, http://www3. Frankfurt School-UNEP Centre (2015) Trends in Renewable Energy ion-batteries-could-be-north-america-Investment 2015, Frankfurt School sourced/ (accessed 22 February 2016).of Finance & Management gGmbH, 12. Sur, U.K. (2012) Graphene: a rising starFrankfurt. on the horizon of materials science. Int.4. Hsing, D., Kang, P., Chen, M., and J. Electrochem., 2012, 12. Article IDOgunseitan, O.A. (2013) Potential envi- 237689.ronmental and human health impacts 13. Avouris, P. and Dimitrakopoulos, C.of rechargeable lithium batteries in elec- (2012) Graphene synthesis and applica-tronic waste. Environ. Sci. Technol., 47, tions. Mater. Today, 15, 86–97.5495 – 5503. 14. Park, S. and Ruoff, R.S. (2009) Chemical5. Song, M.K., Park, S., Alamgir, F.M., Cho, methods for the production of graphenes.J., and Liu, M. (2011) Nanostructured Nat. Nanotechnol., 4, 217–224.electrodes for lithium-ion and lithium-air 15. Choi, W., Lahiri, I., Seelaboyina, R., andbatteries: the latest developments, chal- Kang, Y.S. (2010) Synthesis of graphenelenges, and perspectives. Mater. Sci. Eng. and its applications: a review. Crit. Rev.R, 72, 203–252. Solid State Mater. Sci., 35, 52–71.6. Armstrong, M.J., O’Dwyer, C., Macklin, 16. Allen, M.J., Tung, V.C., and Kaner, R.B.W.J., and Holmes, J.D. (2014) Evaluating (2010) Honeycomb carbon: a review ofthe performance of nanostructured mate- graphene. Chem. Rev., 110, 132–145.rials as lithium-ion battery electrodes. 17. Ovid’ko, I.A. (2013) Mechanical prop-Nano Res., 7, 1–62. erties of graphene. Rev. Adv. Mater. Sci.,7. Etacheri, V., Marom, R., Elazari, R., 34, 1–11.Salitra, G., and Aurbach, D. (2011) Chal- 18. Novoselov, K.S., Fal’ko, V.I., Colombo,lenges in the development of advanced L., Gellert, P.R., Schwab, M.G., andLi-ion batteries: a review. Energy Envi- Kim, K. (2012) A roadmap for graphene.ron. Sci., 4, 3243–3262. Nature, 490, 192–200.8. Liu, L., Lyu, J., Li, T., and Zhao, T. 19. Castro Neto, A.H., Guinea, F., Peres,(2016) Well-constructed silicon-based N.M.R., Novoselov, K.S., and Geim,materials as high-performance lithium- A.K. (2009) The electronic properties ofion battery anodes. Nanoscale, 8, graphene. Rev. Mod. Phys., 81, 109–161.701 – 722. 20. Novoselov, K.S., Morozov, S.V.,9. Zhu, N., Liu, W., Xue, M., Xie, Z., Zhao, Mohinddin, T.M.G., Ponomarenko,D., Zhang, M., Chen, J., and Cao, T. L.A., Elias, D.C., Yang, R., Barbolina, I.I.,(2010) Graphene as a conductive addi- Blake, P., Booth, T.J., Jiang, D., Giesbers,tive to enhance the high-rate capabilities J., Hill, E.W., and Geim, A.K. (2007)

References 129Electronic properties of graphene. Phys. 31. Wu, S., Xu, R., Lu, M., Ge, R., Iocozzia,Stat. Sol. (B), 244, 4106–4111. J., Han, C., Jiang, B., and Lin, Z. (2015)21. Wu, Z.-S., Ren, W., Xu, L., Li, F., and Graphene-containing nanomaterials forCheng, H.-M. (2011) Doped graphene lithium-ion batteries. Adv. Energy Mater.,sheets as anode materials with superhigh 1500400 (40 pages).rate and large capacity for lithium ion 32. Prosini, P.P., Carewska, M., Maroni,batteries. ACS Nano, 5, 5463–5471. F., Tossici, R., and Nobili, F. (2015) A22. Wang, H., Yuan, X., Zeng, G., Wu, Y., lithium-ion battery based on LiFePO4Liu, Y., Jiang, Q., and Gu, S. (2015) Three and silicon/reduced graphene oxide dimensional graphene based materials: nanocomposite. Solid State Ionics, 283, synthesis and applications from energy 145 – 151. storage and conversion to electrochemical 33. Wang, H., Maiyalagan, T., and Wang, sensor and environmental remediation. X. (2012) Review on recent progress Adv. Colloid Interface Sci., 221, 41–59. in nitrogen-doped graphene: synthe-23. Raccichini, R., Varzi, A., Passerini, sis, characterization, and its potential S., and Scrosati, B. (2015) The role of applications. ACS Catal., 2, 781–794. graphene for electrochemical energy 34. Quan, B., Yu, S.-H., Chung, D.Y., Jin, storage. Nat. Mater., 14, 271–279. A., Park, J.H., Sung, Y.-E., and Piao, Y.24. Zhu, Y., Murali, S., Cai, W., Li, X., Suk, (2014) Single source precursor-based J.W., Potts, J.R., and Ruoff, R.S. (2010) solvothermal synthesis of heteroatom- Graphene and graphene oxide: synthesis, doped graphene and its energy storage properties, and applications. Adv. Mater., and conversion applications. Nat. Sci. 22, 1–19. Rep., 4, Endo, M., Kim, C., Nishimura, K., srep05639. Fujino, T., and Miyashita, K. (2000) 35. Hu, Y., He, D., Wang, Y., Fu, M., An, X., Recent development of carbon mate- and Zhao, X. (2016) Defect-introduced rials for Li ion batteries. Carbon, 38, graphene sheets with hole structure as 183 – 197. lithium-ion battery anode. Mater. Lett.,26. Wu, Y.P., Rahm, E., and Holze, R. (2003) 164, 278–281. Carbon anode materials for lithium 36. Gómez-Navarro, C., Meyer, J.C., ion batteries. J. Power Sources, 114, Sundaram, R.S., Chuvilin, A., Simon, 228 – 236. K., Burghard, M., Kern, K., and27. Luo, L., Zhao, P., Yang, H., Liu, B., Kaiser, U. (2010) Atomic structure of Zhang, J.-G., Cui, Y., Yu, G., Zhang, S., reduced graphene oxide. Nano Lett., 10, and Wang, C.-M. (2015) Surface coating 1144 – 1148. constraint induced self-discharging of sil- 37. Fang, H., Zhao, L., Yue, W., Wang, icon nanoparticles as anodes for lithium Y., Jiang, Y., and Zhang, Y. (2015) ion batteries. Nano Lett., 15, 7016–7022. Facile and large-scale preparation of28. Morris, R.S., Dixon, B.G., Gennett, T., sandwich-structured graphene-metal Raffaelle, R., and Heben, M.J. (2004) oxide composites as anode materials for High-energy rechargeable Li-ion battery Li-ion batteries. Electrochim. Acta, 186, based on carbon nanotube technology. 397 – 403. J.Power Sources, 138, 277–280. 38. Zhang, L., Wang, Z., Wang, H., Yang, K.,29. Bellucci, S. (ed) (2011) Physical Proper- Wang, L., Li, X., Zhang, Y., and Dong, H. ties of Ceramic and Carbon Nanoscale (2016) Preparation and electrochemical Structures, Lecture Notes in Nanoscale performances of CoO/3D graphene com- Science and Technology 11, Springer-Verlag, Berlin. posite as anode for lithium-ion batteries.30. Song, B., Yang, J., Zhao, J., and Fang, J. Alloys Compd., 656, 278–283.H. (2011) Intercalation and diffusion 39. Deng, Y., Fang, C., and Chen, G. (2016)of lithium ions in a carbon nanotube The developments of SnO2/graphenebundle by ab initio molecular dynam- nanocomposites as anode materials forics simulations. Energy Environ. Sci., 4, high performance lithium ion batteries: a1379 – 1384. review. J. Power Sources, 304, 81–101.

130 4 Graphene in Lithium-ion Batteries40. Xiang, J., Dong, D., Wen, F., Zhao, J., 50. Badi, N., Erra, A.R., Hernandez, F.C.R.,Zhang, X., Wang, L., and Liu, Z. (2016) Okonkwo, A.O., and Hobosyan, M.Microwave synthesized self-standing (2014) Low-cost carbon-silicon nanocom-electrode of MoS2 nanosheets assembled posite anodes for lithium ion batteries.on graphene foam for high-performance Nanoscale Res. Lett., 9, 360.Li-Ion and Na-Ion batteries. J. Alloys 51. Zhang, Y., Zhao, Z.G., Zhang, X.G.,Compd., 660, 11–16. Zhang, H.-L., Li, F., Liu, C., and Cheng,41. Siwin´ska-Stefan´ska, K. and Kurc, B. H.-M. (2008) Pyrolytic carbon-coated(2015) Preparation and application of a silicon/carbon nanotube composites:titanium dioxide/graphene oxide anode promising application for Li-ion batteries.material for lithiumeion batteries. J. Int. J. Nanomanuf., 2, 4–15.Power Sources, 299, 286–292. 52. Son, I.H., Park, J.H., Kwon, S., Park, S.,42. Zhang, C., Zhang, P., Dai, J., Zhang, Rümmeli, M.H., Bachmatiuk, A., Song,H., Xie, A., and Shen, Y. (2016) Facile H.J., Ku, J., Choi, J.W., Choi, J.-M., Doo,synthesis and electrochemical properties S.-G., and Chang, H. (2015) Siliconof MoO2/reduced graphene oxide hybrid carbide-free graphene growth on siliconfor efficient anode of lithium-ion battery. for lithium-ion battery with high volu-Ceram. Int., 42, 3618–3624. metric energy density. Nat. Commun.,43. Dai, J., Wang, M., Song, M., Li, P., 6.Zhang, C., Xie, A., and Shen, Y. (2016) A 53. Sun, W., Hu, R., Zhang, H., Wang, Y.,novel synthesis of ZnO/N-doped reduced Yang, L., Liu, J., and Zhu, M. (2016) Agraphene oxide composite as superior long-life nano-silicon anode for lithiumanode material for lithium-ion batteries. ion batteries: supporting of grapheneScr. Mater., 112, 67–70. nanosheets exfoliated from expanded44. Yuan, G., Wang, G., Wang, H., and Bai, graphite by plasma-assisted milling.J. (2016) Half-cell and full-cell investiga- Electrochim. Acta, 187, 1–10.tions of 3D hierarchical MoS2/graphene 54. Luo, Z., Xiao, Q., Lei, G., Li, Z.,composite on anode performance in and Tang, C. (2016) Si nanoparti-lithium-ion batteries. J. Alloy. Compd., cles/graphene composite membrane660, 62–72. for high performance silicon anode45. Wu, Z.-S., Zhou, G., Yin, L.-C., Ren, in lithium ion batteries. Carbon, 98,W., Li, F., and Cheng, H.-M. (2012) 373 – 380.Graphene/metal oxide composite elec- 55. Feng, J., Zhang, Z., Ci, L., Zhai, W.,trode materials for energy storage. Nano Ai, Q., and Xiong, S. (2015) ChemicalEnergy, 107–131. dealloying synthesis of porous silicon46. Sethuraman, V.A., Srinivasan, V., and anchored by in situ generated grapheneNewman, J. (2013) Analysis of electro- sheets as anode material for lithium-chemical lithiation and delithiation ion batteries. J. Power Sources, 287,kinetics in silicon. J. Electrochem. Soc., 177 – 183.160, A394–A403. 56. Guo, J. and Wang, C. (2010) A polymer47. Li, X., Zhang, G., Zhang, L., Zhong, scaffold binder structure for high capac-M., and Yuan, X. (2015) Silicon/ ity silicon anode of lithium-ion battery.graphite/carbon nanotubes composite Chem. Commun., 46, 1428– anode for lithium ion battery. Int. J. 57. Koo, B., Kim, H., Cho, Y., Lee, K.T.,Electrochem. Sci., 10, 2802–2811. Choi, N.-S., and Cho, J. (2012) A highly48. de las Casas, C. and Li, W. (2012) A cross-linked polymeric binder for high-review of application of carbon nan- performance silicon negative electrodesotubes for lithium ion battery anode in lithium ion batteries. Angewandtematerial. J. Power Sources, 208, 74–85. Chemie, 51, 8762–8767.49. Ye, M., Hu, C., Lv, L., and Qu, L. (2016) 58. Chan, C.K., Patel, R.N., O’Connell,Graphene-winged carbon nanotubes as M.J., Korgel, B.A., and Cui, Y. (2010)high-performance lithium-ion batteries Solution-grown silicon nanowires foranode with super-long cycle life. J. Power lithium-ion battery anodes. ACS Nano, 4,Sources, 305, 106–114. 1443 – 1450.

5.2 Preparation of Graphene-Based Membranes 137 Transport of ions and molecules in the GO membrane Hydrated ions or molecules Water GO(a) GO Biomedical filtration Desalination or hydrofracking Water, fuel, or chemical purification >2 nm Nanoparticles Covalent bonds or small functional groups Polyelectrolytes or nanofibers 0.3 to 0.7 nm 0.7 to 2 nm(b)Figure 5.4 GO membranes. (a) Water and GO membrane is tunable by adjusting thesmall-sized ions and molecules (compared nanochannel size. (Reprinted from Ref. [44]with the void spacing between stacked with permission from The American Associa-GO nanosheets) permeate superfast in tion for the Advancement of Science, Copy-the GO membrane, but larger species are right 2014.)blocked. (b) The separation capability of theprecisely separate target ions and molecules within a specific size range from bulksolution. In this way, GO membranes may be ideally tailored for applications inwater purification, wastewater reuse, and pharmaceutical and fuel separation, oreven for biomedical applications [44] (Figure 5.4b).GO membranes can be fabricated by different methods, including spin/spray-coating, drop casting, Langmuir–Blodgett (L–B), vacuum filtration orlayer-by-layer (LbL) assembly, among others.The first methodology consists in the spin/spray-coating of a smooth substrate(i.e., copper foil) using a GO suspension [2, 7, 46]. Nair et al. [7] described a proce-dure to fabricate freestanding GO membranes of ∼1 cm in diameter after etchingaway a central part of the copper foil with a HNO3 treatment. Finally, the mem-branes were cleaned with deionized water and dried on a hot plate (<50 ∘C). Densestacking occurs because the face-to-face attractive capillary forces created by thespray- or spin-coating can overcome the repulsive forces between the edges ofGO sheets. Therefore, the initial deposition is mainly governed by capillary inter-actions between the faces of GO sheets, rather than by the electrostatic forcesbetween the GO edges (Figure 5.5).GO dispersions can be also re-assembled into membranes with controlled thick-ness through a drop-casting process. This methodology induced the nanocapillarynetwork formation for selective ion penetration and water purification proper-ties of freestanding GO membranes [47, 48]. In this method, the GO colloidalsuspension is drop-casted onto a substrate with a smooth surface, such as sil-ica or paper, and dried at room temperature. Once deposited, the van der Waals

138 5 Graphene-Based Membranes for Separation Engineering (c) (a) (b) 1 cm 1 μmFigure 5.5 (a) Photo of a 1-mm-thick GO film peeled off of a Cu foil. (b) TEM of the filmcross-section (c) mounted on Cu hole for separation test. (Reprinted from Ref. [7] with per-mission from The American Association for the Advancement of Science, Copyright 2012.)forces are sufficient to keep the GO sheets strongly adhered to the substrate. Theoxygen functional groups, located asymmetrically in the GO membrane, maintaina certain distance during the drop-casting process, creating void spaces betweennon-oxidized regions forming a series of nanocapillaries within the film. The preparation of GO membranes by L–B assembly includes the dispersion ofGO in a volatile organic solvent (i.e., methanol or chloroform) [37, 49] and then,spreading it onto the water surface. As the solvent evaporates, the molecules aretrapped on the water surface, forming a monolayer. The film can be transferred toa solid substrate (e.g., by dip-coating), forming a monolayer coating over a largearea. Bi- and tri-layers of GO films can be prepared by drying the as-formed filmsin an oven, and then depositing another layer of GO under the same conditions.The number of layers and thickness of GO membranes can be readily controlledusing the L–B technique. Vacuum filtration has been frequently used for the fabrication of self-standingGO membranes for gas separation and water treatment [1–3, 50, 51]. The GO sus-pension is filtered through the supports leading to the deposition of the GO mem-brane on the filter. The thickness of the GO membrane is controlled by adjustingthe volume of the GO suspension. The most commonly used filters for preparingmembranes are anodic aluminum oxide (AAO) discs and polymeric membranes(i.e., mixed cellulose ester membranes). The AAO discs are preferred due to theirrigidity, which can promote a tighter packing of the sheets under vacuum filtra-tion, leading to stronger membranes [3, 52]. Membranes fabricated by vacuum filtration have interactions between thenanosheets, such as electrostatic repulsion, van der Waals attractive forces andhydrogen bonding. Compression induced by vacuum sucking was responsible foraligning the GO sheets perpendicularly to the flow direction and narrowing theinter-spacing between the oriented GO sheets [53]. It has been reported [51] thatthe filtration rate affects the pore structure substantially and, consequently, theseparation capacity of the resulting membranes, since it affects the way the GOsheets are stacked on the surface of the porous substrate. Slow filtration leads tosmaller thickness of the GO stacks and thus the space between the aggregatedstacks is of smaller size.

5.2 Preparation of Graphene-Based Membranes 139 It is noteworthy that pristine GO membranes made via solution filtrationdemonstrate outstanding mechanical durability in dry conditions [3]; however,the dispersion of GO membranes in water (i.e., low stability) has been alsoreported due to the electrostatic repulsion between the GO sheets (negativelycharged upon hydration) that can overcome van der Waals attraction or hydrogenbonding that holds the sheets together [20, 21, 54]. Thus, such free-standingmembranes made by simple filtration methods are not likely to survive thecross-flow testing conditions, typical of real-world membrane operations [49]. One way to solve this issue consists in forming stable bonds between GOnanosheets, stabilizing the membrane integrity. These cross-linking spacerscould also act as barriers to form a new channel, which would greatly enhancewater permeability. To this end, few studies have focused on fabricating cross-linked GO membranes in an effort to improve their mechanical integrity [20,52, 55–59] during LbL assembly. The LbL method is ideal to introduce aninterlayer stabilizing force via covalent bonding, electrostatic interaction, or botheffects during layer deposition [44]. The GO membrane thickness can be readilycontrolled by varying the number of LbL deposition cycles. This technique ishighly scalable for membranes, as opposed with the synthesis of monolayergraphene membranes, which requires the manufacturing of large-sized graphenesheets and the punching of nanopores with a narrow size distribution [60]. Yeh et al. [52] also reported that pristine GO membranes can disintegrate inwater, but they can become stable if they are cross-linked by multivalent cationicmetals. They found that metal cations, such as Al3+ (released from an AAO fil-ter disc during the filtration) or Mn2+ (a by-product from GO synthesis), canact as cross-linkers in the sheets and strengthen the final membrane. Significantenhancement in mechanical stiffness and fracture strength of GO membranes canbe also achieved by the addition of a small amount (<1 wt%) of Mg2+ and Ca2+during the filtration [55]. These results can be readily rationalized in terms ofthe chemical interactions between the functional groups of the GO sheets andthe divalent metals ions. Hu et al. [56] reported GO membranes by LbL assem-bling with positively charged polyelectrolytes such as poly(allylamine hydrochlo-ride) (PAH) via electrostatic interaction. Although the binding forces betweenassembled polyelectrolyte layers are mainly contributed by electrostatic interac-tion, hydrophobic interaction and hydrogen bonding may also be involved to alower extent. Hu and Mi [20] prepared a water filtration membrane with layered GOnanosheets on a polysulfone support coated with polydopamine and 1,3,5-benzenetricarbonyl trichloride (TMC) as cross-linkers (Figure 5.6). The ionic fluxof sulfonated polyethersulfone (SPES) membranes can be improved up to 19% bythe incorporation of 10% GO as compared to unmodified SPES membrane. Thestrong interfacial interactions due to GO nanofillers into the SPES matrix improvethe thermal and mechanical properties of the nanocomposite membranes [57].An et al [58]. studied GO membranes covalently cross-linked with borateinspired by cell wall assembly of higher-order plants. The study has reported aremarkable enhancement by over 255% in the stiffness of the graphene film by

140 5 Graphene-Based Membranes for Separation Engineering Polysulfone DopamineTMC solution Polydopamine TMC (b)GO solutionRepeated LbL deposition TMC solution(a) (c)Figure 5.6 Schematic illustration of (a) an the mechanism of reactions between GO andLbL procedure to synthesize the GO mem- TMC. (Reprinted from Ref. [20] with permis-brane; (b) the mechanism of reactions sion from the American Chemical Society,between polydopamine and TMC, and (c) Copyright 2013.)adding 0.94 wt% boron to the final composite GO film. The strong cross-linkingbetween adjacent nanosheets has been attributed to the formation of covalentbonds between borate ions and the hydroxyl groups on the GO nanosheets.5.3Graphene-based Membranes for Separation ApplicationsGraphene-based membranes have shown promising performance in gas separa-tion and water treatment due to their unique “size sieving” effect. Both nanoscaledinterlayer spacing between two individual flakes and selective structural defects onthe flakes have been claimed to be responsible for the observed separation perfor-mance.5.3.1Gas SeparationTheoretical studies have investigated the permeability and selectivity of graphene-based membranes toward different gases [6, 10, 61–63]. In 2012, Nair et al. [7] tested the permeance of several gases, including He,H2, N2, and Ar through self-standing GO membranes with a thickness rangingfrom 0.1 to 10 mm. The leakage rate was monitored by the variations in the inner

5.3 Graphene-based Membranes for Separation Applications 141pressure of the sealed containers over a period of several days. The results revealedthat no noticeable reduction in pressure was observed for any tested gas, includingHe, H2, N2, and Ar, suggesting that the GO membranes were completely imper-meable to these tested gases, which is similar to that previously mentioned forgraphene monolayers. This work was the first report showing that GO membranesmay separate different molecules through interlayer spacing between GO sheets. Koening et al. [27] demonstrated the experimental evidence of gas separationthrough individual porous graphene membranes. Pores were introduced ingraphene by ultraviolet-induced oxidative etching and the molecular transportthrough them was measured using a pressurized blister test as well as mechanicalresonance. The deformation of the film surface due to gas pressure on the feed sidewas measured to quantify the permeability, with a higher deflection indicating alower permeability due to the gas molecules not being able to pass the graphenebarrier. The etched porous graphene membrane was found to be selective for H2and CO2 over larger molecules such as Ar, N2, and CH4, suggesting the presenceof pores slightly larger than 0.3 nm in size (Figure 5.7). This work demonstratedthe great potential of using nanoporous graphene as a promising membranematerial for gas separation by molecular sieving. Kim et al. [2] demonstrated highly permeable and selective thin (i.e., few-layered) defective graphene and GO membranes for separating mixtures of gasesof industrial relevance. They reported that graphene itself displays outstanding gasseparation properties, with O2/N2 separation properties above the so-called upperbound [64] (Figure 5.8). Further studies on GO membranes showed outstandingproperties in terms of CO2 selectivity (e.g., CO2/H2, CO2/N2, and CO2/CH4).The authors reported that the gas-transport properties are quite sensitiveto the preparation method and the presence of water in the GO structures, 150 100 CO2 Before etch 10 H2 After etch 100 Maximum deflection, δ (nm) –dδ/dt (nm min–1) 1 50 Before etch After etch 0.1 0.01 Ar 0 H2 0.01 N2 CH4 CO2 H2 10–3 –50 Ar CO2 Ar –100 CH4 N2 0 CH4(a) 25 50 75 100 125 150 175 10–4 0.28 0.30 0.32 0.34 0.36 0.38 0.40 Time, t (min) (b) Molecular size (nm)Figure 5.7 Comparison of gas permeation H2 and CO2 over larger gas molecules afterrates of pristine and porous graphene mem- etching. The connecting lines show the mea-branes: (a) maximum deflection of the mem- surements before (black) and after (red) etch-brane surface versus time before and after ing. (Reprinted from Ref. [27] with permissionetching; (b) gas permeance versus molec- from Nature Publishing Group, Copyrightular size, revealing selective permeation of 2012.)

142 5 Graphene-Based Membranes for Separation Engineering N2 Defective regions 103O2 Wrinkle Upper bound of polymer membrane Selectivity (CO2/N2) 102 GO (hydrate state 85%) GO (hydrate state 68%) CMS TZPIM GO (dry state) 101 TR PIM Zeolite Silica Possible Polymer 102 103 104 105 diffusion 1010 01 CO2 permeability (Barrer) pathway (b)(a)Figure 5.8 Permeation of gas molecules PIM, polymer of intrinsic microporosity) andthrough laminates of GO membranes. (a) Per- inorganic (CMS, carbon molecular sieve, zeo-meation of gases through defect regions. (b) lite, silica) membranes. (Reprinted from Ref.Performance of GO membranes relative to [2] with permission from The American Asso-other high-performance polymeric (TR, ther- ciation for the Advancement of Science,mally rearranged polymer; TZPIM, tetrazole Copyright 2013.)functional polymer of intrinsic microporosity;as these variables change the organization of the laminate structure, therebychanging the shapes and sizes of the openings available for gas transport [65]. Li et al. [50] prepared GO membranes with a thickness of about 1.8 nm by afacile filtration process. The membranes showed mixture separation selectivitiesas high as 3400 and 900 for H2/CO2 and H2/N2 mixtures, respectively, throughselective structural defects on GO. They suggested that the major transport path-way in few nanometer thick GO films relied mostly on certain structural defectspresent along the GO sheets rather than due to interlayer spacing. Our group [51] fabricated a series of GO membranes by varying the filtrationrate of the starting GO suspension as well as the amount of the surface func-tional groups of the employed GO. The as-produced GO membranes were robustenough for incorporation in various membrane modules as well as to test their per-formance in gas permeability studies. The results showed that the slow filtrationderived membranes were almost impermeable to m-xylene vapor and exhibiteda good separation performance for several gas pairs (H2/N2, H2/CO, H2/CH4,H2/C2H6, H2/C4H10, and H2/SF6). Gas transport through the void space betweenGO stacks, or discontinuities of GO layers, was dominant over transport throughthe inter-layer space.5.3.2Water TreatmentGraphene-based membranes have been developed and applied for water applica-tions, including treatment, purification, and even desalination.

5.3 Graphene-based Membranes for Separation Applications 143 Nair et al. [7] were the first authors that reported one of the most exciting andrevolutionary physical properties of GO membranes as ideal filters for water,opening outstanding opportunities in separation science. They found that GOmembranes can be impermeable to all molecules tested except to water vapor(Figure 5.9). Water molecules can flow through the empty space between the lay-ers in a GO laminate (interlayer d-distance) because this space is what is requiredfor one layer of water molecules. When the other molecules that were testedtried to enter such space, they found that the graphene capillaries were shrunk,or clogged, with water molecules. In addition, the permeation of water throughthe GO membrane was as fast as the evaporation through an open aperture(unimpeded). Cohen-Tanugi et al. [4] showed the high performance of monolayer grapheneto effectively filter salt (NaCl) from water, using classical molecular dynamics.They report the desalination performance of such membranes as a functionof pore size, chemical functionalization, and pressure. In another work [66],the authors reported the mechanical strength of nanoporous graphene as areverse osmosis (RO) membrane using molecular dynamics simulation. Theyshowed that an appropriate substrate with openings smaller than 1 μm wouldallow nanoporous graphene to withstand pressures exceeding 570 bar or 10times more than typical pressures for seawater RO. Moreover, nanoporousgraphene membranes can maintain their ultrahigh permeability even at lowpressures (<100 bar), indicating that they are promising membranes for waterdesalination [67]. 0.9 H2O through Water GO 10–7 H2O through 10–11 open aperture 0.6 Weight loss (g) Permeability (mm × g/cm2 × s × bar) 0.3 H2O through Ethanol Acetone reduced GO Hexane 0 Ethanol 0 Hexane 10–15 Decane Argon(a) H2O Propanol Hydrogen (b) Nitrogen 10 20 He Time (h)Figure 5.9 Permeation through GO mem- indicate the upper limits set by our experi-brane. (a) Weight loss for a container sealed ments). (Inset) Schematic representation ofwith a GO film. No loss was detected for the structure of monolayer water inside aethanol, hexane, and so on, but water graphene capillary with d = 7 Å. (Reprintedevaporated from the container as freely as from Ref. [7] with permission from The Amer-through an open aperture (blue curves). (b) ican Association for the Advancement of Sci-Permeability of GO membrane with respect ence, Copyright 2012.)to water and various small molecules (arrows

144 5 Graphene-Based Membranes for Separation Engineering O’Hern et al. [12] reported selective ionic transport through controlled sub- nanometer pores (created by chemical oxidation of nucleated defects) in graphene membranes. The results revealed that the created pores were selective for cations at short oxidation times, consistent with electrostatic repulsion from negatively charged functional groups terminating the pore edges. At longer oxidation times, the pores allowed transport of salt but prevented the transport of a larger organic molecule, indicative of steric-size exclusion. In a separate study, Qiu et al. [68] reported an approach to create corruga- tion on reduced graphene oxide (rGO) flakes by hydrothermal treatment of the rGO dispersions. The researchers revealed that the amplitude of GO corrugation can be simply controlled by hydrothermal treatment temperature, and the corru- gation could form nanochannels in the rGO membranes. Colloids of Au and Pt nanoparticles with average diameters of ∼13 and ∼3 nm, respectively, were used for a series of filtration tests. As a result, the 90-rGO (rGO hydrothermally treated at 90 ∘C) membrane showed no permeation for both Au and Pt, indicating a chan- nel size smaller than 3 nm. When rGO was treated at higher temperatures (100 and 120 ∘C), the membranes allowed only Pt to pass through, but at a still higher tem- perature (150 ∘C) both Au and Pt nanoparticles could pass through the membrane. The rejection of ions and organic dye molecules by GO membranes has also been intensively studied. The selectivity of this process is mainly attributed to both size exclusion and interactions (including chemical and electrostatic ones) with functional groups. Hu et al. [20] reported that GO membranes display a relatively low rejection of monovalent and divalent salts (6–46%) and moderate/high rejec- tion of organic dyes (46–66% and 93–95% for methylene blue and rhodamine B, respectively). Similar results were also reported by Han et al. [21] with large per- formance for the retention of organic dyes (especially for the charged dyes), which was ascribed to the mechanism of physical sieving and electrostatic interaction. Huang et al. [54] investigated the effects of pH, salt concentration, and pressure on the controlled separation performance of GO membranes. The ionic strength and protonation (deprotonation) of the carboxylic acid groups on the GO sheets were responsible for the variation in the inter-spacing between GO sheets and the pore size of nanochannels, leading to the corresponding separation performance. At low pH values and high salt concentrations, the repulsion forces between the negatively charged GO sheets dramatically decrease and shrink the space between them, resulting in the reduction of the flux and increase in the rejection rate. Furthermore, the collapsed nanochannels at high pressure could be recovered by releasing the applied pressure, due to their unique elastic properties. Sun et al. [47] demonstrated selective ion penetration properties of freestanding GO membranes; they reported fast permeation of sodium salts, slow permeation of heavy-metal salts, and no permeation of Cu ions and organic contaminants such as rhodamine B. The nanocapillaries formed within the GO membranes might be responsible for the permeation of metal cations and the strong coordina- tion between transition-metal ions, and the oxygen-containing functional groups might give rise to the much slower permeation of heavy metal ions through GO membranes compared to sodium salts (Figure 5.10). The authors also reported the

5.3 Graphene-based Membranes for Separation Applications 145Drop-casting induced nanocapillary network formationPenetration of Na+ ions through nanocapillaries Coordination of Cu2+ ionsCoordination of Mn2+ ions Coordination of Cd2+ ionsFigure 5.10 Schematic diagrams of a GO membrane and the interaction with different ions.(Reprinted from Ref. [47] with permission from the American Chemical Society, Copyright2013.)selective trans-membrane transport properties of alkali and alkaline earth cations(Na+, K+, Mg2+, Ca2+, and Ba2+) through a membrane composed of stackedand overlapped GO sheets. The simulation of the transmembrane transport ofcations indicated that the coordinative interactions are responsible for selectivity[48]. The authors [48] also demonstrated the efficient recovery of acids fromiron-based electrolytes using GO membranes. The trans-membrane transportof H+ was two orders of magnitude greater than that of Fe3+. Notably, when theconcentration of FeCl3 sources was reduced to some extent (e.g., 0.01 mol l−1),the Fe3+ cations could be blocked by GO membranes. The mechanism for theeffective separation of H+ from Fe3+ was discussed, indicating that the molecularsieving effect of GO nanocapillaries and the coordination between Fe3+ and GOwere responsible for the effective capture of Fe3+ while the rapid propagation ofprotons through the hydrogen-bonding networks along the water layers formedwithin the interlayer spacing was responsible for the fast migration of H+. Joshi et al. [3] described the permeation of aqueous solutions of ions and neu-tral molecules through GO membranes, reporting “ultrafast” transport proper-ties. They focused on the transport of ions (e.g., NaCl, MgCl2, and K3[Fe(CN)6])as well as water-soluble neutral molecules (glycerol and sucrose). The authorsobserved a remarkably sharp cut-off in the permeation properties as a function ofthe hydration size of the permeating solute. Small hydrated ions, ranging in sizefrom K+ to Mg2+, had similar permeation properties, whereas larger species wereessentially prohibited to permeate through the GO membrane. They estimated

146 5 Graphene-Based Membranes for Separation Engineering that the membrane was essentially impermeable to molecules with a hydration radius larger than approximately 4.5 Å. The authors also reported very high par- tition coefficients for salts with GO membranes, suggesting values for the salt partition coefficient, K (the ratio of the salt concentration in the GO membrane to that in the external solution), as high as 10 or more. Such values are extraordi- nary compared to those obtained for typical polymers, which are widely used in desalination and other applications involving salt transport (Figure 5.11). The use of membranes consisting of polymeric matrices and GO is an alter- native that combines the processability of polymers and the unique properties of GO materials for water treatment. The hydrophilic nature of GO makes it a good candidate to improve antimicrobial properties, increase permeability, and enhance the mechanical strength of polymeric membranes [69–73]. GO mem- branes cross-linked with polydopamine and TMC displayed significant rejection of monovalent and divalent salts and moderate or high rejection of organic dyes [20]. Choi et al. [59] created an assembly of GO nanosheets on a polyamide (PA) thin-film composite (TFC) membrane for desalination driven by RO (Figure 5.12). The GO material was coated onto the PA-TFC membrane surface via LbL deposi- tion of oppositely charged nanosheets, which resulted in increased hydrophilicity and reduced surface roughness of the RO membrane. The altered properties of the RO membrane due to the GO nanosheets resulted in enhanced resistance to protein fouling and increased chlorine resistance. Wang et al. [74] used rGO in polyethersulfone (PES) membranes to fabricate TFC membranes, which were tested in forward osmosis (FO) of salty water. TFC membranes modified with rGO presented higher water fluxes than those prepared with the neat PES support, 101 15 Cl– Mg2+2.0 Propanol 10Ions permeated (10–3 mol cm–2) K+ AsO43– Permeation rate 10–1 (mg h–1 × cm2) Permeation rate (mol h–1 × m2) 5 10–3 [Fe(CN)6]3– PTS4– [Ru(bipy)3]2+1.5 0 012 Feed concentration1.0 (mol L) Mg2+0.5 10–5 Sucrose 0.0 Glycerol(a) 0 5 10 15 20 3.0 3.5 4.0 4.5 5.0 5.5 6.0 Time (h) (b) Hydrated radius (Å) Figure 5.11 (a) Ion permeation through GO mesh. The shown permeation rates are nor- laminates. Permeation through a 5-mm-thick malized per 1 M feed solution and are mea- GO membrane from the feed compartment sured by using 5-mm-thick membranes. with a 0.2 M solution of MgCl2. (Inset) Per- (Reprinted from Ref. [3] with permission from meation rates as a function of C in the feed The American Association for the Advance- solution; (b) Sieving through atomic-scale ment of Science, Copyright 2014.)

5.3 Graphene-based Membranes for Separation Applications 147GO AGO/GO bilayers Polyamide active layer Polysulfone supportAGO Polyester non-wovenFigure 5.12 Schematic illustration of a mul- (AGO) nanosheets. (Reprinted from Ref. [59]tilayered GO coating on a PA thin-film com- with permission from the American Chemicalposite membrane surface via LbL deposition Society, Copyright 2013.)of oppositely charged GO and aminated-GOthe enhanced performance being due to a low structural parameter of the result-ing membrane and the reduction of the internal concentration polarization. TFCmembranes with GO (0.12 wt%) also improved the pure water flux upto 80% com-pared to the neat TFC membrane, without significantly affecting salt selectivity inRO experiments [75]. Similar findings were reported concerning GO–TiO2 (GOT) nanocompositemembranes for water treatment [76, 77] (Figure 5.13) and GO-based polyelec-trolyte membranes assembled to synthesize the cross-linked GO membrane forwater reuse [20, 58]. These cross-linking spacers could also act as barriers to formnew channels that would greatly enhance water permeability. In view of the versatile mass transport characteristics of GO membranes,focus on their pervaporation separation behavior has also been reported.(a) (b)Figure 5.13 Cross-sectional M-GO/GOT SEM images at different magnifications. (a) 1000×;(b) 2500× (inset corresponds to the freestanding GO membrane: M-GO). (Reprinted from Ref.[77] with permission from Elsevier, Copyright 2015.)

148 5 Graphene-Based Membranes for Separation Engineering Tang et al. [78] used the pervaporation process for dehydration of ethanol using free-standing GO membranes assembled by a pressurized ultrafiltration method. Experimental results suggested that the interlayer spacing was determined by both the packing density of GO nanosheets and the water content in the feed solution. The packing density was sensitively affected by the ultrafiltration pressure applied during membrane formation. By tuning the ultrafiltration pressure, a high separation performance with water permeability of 13 800 Barrer (1 Barrer = 3.348 × 10−19 kmol m m−2 s−1 Pa−1) and water/ethanol selectivity of 227 was achieved for dehydration of a 85 wt% ethanol aqueous solution at 24 ∘C. Huang et al. [79] fabricated a GO membrane prepared by vacuum filtration onto a ceramic hollow fiber. This GO membrane exhibited excellent water permeation of dimethyl carbonate/water mixtures through a pervaporation process. The permeate water content reached 95.2 wt% and a high permeation flux of 1702 g m−2 h−1 at 25 ∘C and 2.6 wt% feed water content. Tsou et al. [80] synthesized GO/polyacrylonitrile (mPAN) membranes pre- pared by a pressure-, vacuum-, and evaporation-assisted self-assembly technique. The membranes were applied to dehydrate 1-butanol mixtures by pervaporation. The results showed that the membranes exhibited exceptional pervaporation performance at 30 ∘C (permeate water of 99.6 wt% and permeation flux of 2.54 kg m−2 h−1); moreover, the membrane sustained its operating stability at a relatively high temperature of 70 ∘C and a high water concentration of 99.5 wt% was maintained with a permeation flux of 4.34 kg m−2 h−1. Currently, the mentioned problems associated with nanoporous graphene membranes are expected to be overcome, and the commercialization of this type ™of membranes already started; Perforene (from Lockheed Martin Corporation, USA) being an example. The Perforene membrane was developed by placing holes that are ≤1 nm of size in a graphene membrane. These holes are small enough to trap the ions while dramatically improving the flow-through of water molecules, reducing clogging and pressure on the membrane. The Perforene material works by removing sodium, chlorine, and other ions from seawater and other sources. Furthermore, the graphene-based membranes are seen as being 500× thinner and 99% more energy efficient than those employed in RO [81]. Graphene is also being explored for its use in various stages of oil and gas processes. However, the difficulty of dispersing large graphene flakes in aqueous media creates problems in water-based muds [82]. Instead of graphene, GO provides a more stable material for aqueous dispersions and it maintains the sheet-like morphology that would allow the desired pore-plugging through filter cake formation. Kosynkin et al. [83] showed that adding platelets of GO to a common water-based drilling fluid decreased the losses of the fluid to the surrounding rock, as compared to a standard mixture of clays and polymers used in the drilling industry today. By methylating the GO through an esterification reaction, the stability of GO in saline environments is increased, making it an ideal candidate for the next generation of fluid-loss-control additives. Another work has been patented [84], which is related to the recovery of oil from subterranean reservoirs and, more particularly, to new and improved

5.4 Conclusions 149secondary recovery operations using flood water including a dispersion of ahydrophilic graphite oxide for mobility control [85]. McCoy et al. [86] haverevealed the possibility of GO in stabilizing crude oil in water emulsion. Athighly acidic pH values, fully reversible flocculation of emulsion droplets canbe achieved, whereas when adjusted to high pH, the flocculation is irreversible,which can be interpreted as a permanent chemical change in GO. The interfacialcharge of the GO and oil–water interface is the over-riding drive for the excep-tional stability of acidic GO emulsions. Nguyen et al. [87] reported the use ofGO as a thermal stability additive for polymer solutions in polymer-enhanced oilrecovery. Baker Hughes Incorporated, an oilfield service company, patented the use ofgraphene as a component in fluids used for retrieving oil and gas [88], since thegraphene can infiltrate the rock and act as a seal, preventing leakage that is morefluid. On the other hand, Lockheed Martin Corp. is testing nanofilters using itspatented Perforene (graphene sheets with precisely sized holes as small as 1 nm)for water remediation in the oil and gas industry. The company states that theultimate goal is water desalination, but more feasible and immediate uses can befound in the oil and gas industry, in particular to clean drilling wastewaters, wherethe requirements in terms of the quality of the graphene and hole sizes are lesschallenging. This goal only requires 50–100 nm sized holes, compared to the 1 nmholes needed for desalination [89].5.4ConclusionsIn summary, membranes based on graphene and graphene derivatives are stand-ing up as promising candidates for precise and selective gas separation and waterpurification processes due to their high permeance, outstanding mechanical prop-erties, and facile fabrication. The structure and preparation of graphene-based membranes such asnanoporous graphene membranes, free-standing GO membranes and chemicallycross-linked GO membranes, among others, have been briefly summarized. Ithas been demonstrated that graphene-based membranes can exhibit mechanicalstrength, high permeance, and excellent selectivity. Graphene-based membranes show excellent selectivity and superior perfor-mance for multiple gas mixtures as well as for water/wastewater treatment.However, several challenges still exist in their path to industrial application,including the ability needed to synthesize large quantities of graphene materialwith enough mechanical strength to be stable under higher hydrostatic pressures,as well as the precise design essential of GO membranes, including the pattern ofoxidized regions, the interlayer space, the pore size, and the number of graphenelayers. These studies are expected to attract a great deal of interest from theindustry, including the oil and gas industry, due to the importance of separationprocesses in their sustainability.

150 5 Graphene-Based Membranes for Separation Engineering Acknowledgments This work was financially supported by Project POCI-01-0145-FEDER-006984 – Associated Laboratory LSRE-LCM funded by FEDER funds through COM- PETE2020 - Programa Operacional Competitividade e Internacionalização (POCI) – and by national funds through FCT - Fundação para a Ciência e a Tecnologia. LMPM and AMTS acknowledge the FCT Investigator program (IF/01248/2014 and IF/01501/2013, respectively), with financing from the European Social Fund and the Human Potential Operational Program. SMT acknowledges financial support from FCT grant SFRH/BPD/108981/2015.References1. Dikin, D.A., Stankovich, S., Zimney, 8. Aghigh, A., Alizadeh, V., Wong, H.Y., E.J., Piner, R.D., Dommett, G.H.B., Islam, M.S., Amin, N., and Zaman, M. Evmenenko, G., Nguyen, S.T., and Ruoff, (2015) Recent advances in utilization of R.S. (2007) Preparation and characteri- graphene for filtration and desalination zation of graphene oxide paper. Nature, of water: a review. Desalination, 365, 448 (7152), 457–460. 389 – 397.2. Kim, H.W., Yoon, H.W., Yoon, S.-M., 9. Mahmoud, K.A., Mansoor, B., Mansour, Yoo, B.M., Ahn, B.K., Cho, Y.H., Shin, A., and Khraisheh, M. (2015) Functional H.J., Yang, H., Paik, U., Kwon, S., Choi, graphene nanosheets: the next genera- J.-Y., and Park, H.B. (2013) Selective gas tion membranes for water desalination. transport through few-layered graphene Desalination, 356, 208–225. and graphene oxide membranes. Science, 10. Jiang, D.-E., Cooper, V.R., and Dai, S. 342 (6154), 91–95. (2009) Porous graphene as the ultimate3. Joshi, R.K., Carbone, P., Wang, F.C., membrane for gas separation. Nano Kravets, V.G., Su, Y., Grigorieva, I.V., Lett., 9 (12), 4019–4024. Wu, H.A., Geim, A.K., and Nair, R.R. 11. Suk, M.E. and Aluru, N.R. (2010) Water (2014) Precise and ultrafast molecular transport through ultrathin graphene. J. sieving through graphene oxide mem-4. branes. Science, 343 (6172), 752–754. 12. Phys. Chem. Lett., 1 (10), 1590–1594. Cohen-Tanugi, D. and Grossman, O’Hern, S.C., Boutilier, M.S.H., Idrobo, J.C. (2012) Water desalination across J.-C., Song, Y., Kong, J., Laoui, T., Atieh, nanoporous graphene. Nano Lett., 12 (7), M., and Karnik, R. (2014) Selective ionic 3602 – 3608. transport through tunable subnanometer5. Konatham, D., Yu, J., Ho, T.A., and pores in single-layer graphene mem- Striolo, A. (2013) Simulation insights branes. Nano Lett., 14 (3), 1234–1241. for graphene-based water desalina- 13. Sint, K., Wang, B., and Král, P. (2008) tion membranes. Langmuir, 29 (38), Selective ion passage through functional- 11884 – 11897. ized graphene nanopores. J. Am. Chem.6. Du, H., Li, J., Zhang, J., Su, G., Li, X., Soc., 130 (49), 16448–16449. and Zhao, Y. (2011) Separation of hydro- 14. Sun, C., Boutilier, M.S.H., Au, H., gen and nitrogen gases with porous Poesio, P., Bai, B., Karnik, R., and graphene membrane. J. Phys. Chem. C, Hadjiconstantinou, N.G. (2014) Mecha- 115 (47), 23261–23266. nisms of molecular permeation through7. Nair, R.R., Wu, H.A., Jayaram, P.N., nanoporous graphene membranes. Grigorieva, I.V., and Geim, A.K. (2012) Langmuir, 30 (2), 675–682. Unimpeded permeation of water through 15. Zhu, Y., Murali, S., Cai, W., Li, X., Suk, helium-leak–tight graphene-based mem- J.W., Potts, J.R., and Ruoff, R.S. (2010) branes. Science, 335 (6067), 442–444. Graphene and graphene oxide: synthesis,

References 151properties, and applications. Adv. Mater., (2010) DNA translocation through22 (35), 3906–3924. graphene nanopores. Nano Lett., 10 (8),16. Eda, G. and Chhowalla, M. (2010) 3163 – 3167.Chemically derived graphene oxide: 27. Koenig, S.P., Wang, L., Pellegrino, J., andtowards large-area thin-film electronics Bunch, J.S. (2012) Selective molecularand optoelectronics. Adv. Mater., 22 sieving through porous graphene. Nat.(22), 2392–2415. Nanotechnol., 7 (11), 728–732.17. Huang, H., Ying, Y., and Peng, X. (2014) 28. Cheng, Y.C., Kaloni, T.P., Zhu, Z.Y., andGraphene oxide nanosheet: an emerg- Schwingenschlögl, U. (2012) Oxidationing star material for novel separation of graphene in ozone under ultravioletmembranes. J. Mater. Chem. A, 2 (34), light. Appl. Phys. Lett., 101 (7), 073110.13772 – 13782. 29. Liu, L., Ryu, S., Tomasik, M.R.,18. Xu, Q., Xu, H., Chen, J., Lv, Y., Dong, C., Stolyarova, E., Jung, N., Hybertsen,and Sreeprasad, T.S. (2015) Graphene M.S., Steigerwald, M.L., Brus, L.E., andand graphene oxide: advanced mem- Flynn, G.W. (2008) Graphene oxida-branes for gas separation and water tion: thickness-dependent etching andpurification. Inorg. Chem. Front., 2 (5), strong chemical doping. Nano Lett., 8417 – 424. (7), 1965–1970.19. Hegab, H.M. and Zou, L. (2015) 30. Wu, S., Yang, R., Shi, D., and Zhang,Graphene oxide-assisted membranes: G. (2012) Identification of structuralfabrication and potential applications in defects in graphitic materials by gas-desalination and water purification. J. phase anisotropic etching. Nanoscale, 4Membr. Sci., 484, 95–106. (6), 2005–2009.20. Hu, M. and Mi, B. (2013) Enabling 31. Russo, C.J. and Golovchenko, J.A. (2012)graphene oxide nanosheets as water Atom-by-atom nucleation and growth ofseparation membranes. Environ. Sci. graphene nanopores. Proc. Natl. Acad.Technol., 47 (8), 3715–3723. Sci., 109 (16), 5953–5957.21. Han, Y., Xu, Z., and Gao, C. (2013) 32. Brodie, B.C. (1859) On the atomicUltrathin graphene nanofiltration mem- weight of graphite. Philos. Trans. R. Soc.brane for water purification. Adv. Funct. Lond., 149, 249–259.Mater., 23 (29), 3693–3700. 33. Staudenmaier, L. (1898) Verfahren zur22. Bunch, J.S., Verbridge, S.S., Alden, J.S., Darstellung der Graphitsäure. Ber. Dtsch.van der Zande, A.M., Parpia, J.M., Chem. Ges., 31 (2), 1481–1487.Craighead, H.G., and McEuen, P.L. 34. Hummers, W.S. and Offeman, R.E.(2008) Impermeable atomic membranes (1958) Preparation of graphitic oxide. J.from graphene sheets. Nano Lett., 8 (8), Am. Chem. Soc., 80 (6), 1339–1339.2458 – 2462. 35. Stankovich, S., Piner, R.D., Chen, X.,23. Leenaerts, O., Partoens, B., and Peeters, Wu, N., Nguyen, S.T., and Ruoff, R.S.F.M. (2008) Graphene: a perfect nanobal- (2006) Stable aqueous dispersions ofloon. Appl. Phys. Lett., 93 (19), 193107. graphitic nanoplatelets via the reduction24. Garaj, S., Liu, S., Golovchenko, J.A., and of exfoliated graphite oxide in the pres-Branton, D. (2013) Molecule-hugging ence of poly(sodium 4-styrenesulfonate).graphene nanopores. Proc. Natl. Acad. J. Mater. Chem., 16 (2), 155–158.Sci. U S A, 110 (30), 12192–12196. 36. Paredes, J.I., Villar-Rodil, S.,25. Merchant, C.A., Healy, K., Wanunu, Martínez-Alonso, A., and Tascón, J.M.D.M., Ray, V., Peterman, N., Bartel, (2008) Graphene oxide dispersions inJ., Fischbein, M.D., Venta, K., Luo, organic solvents. Langmuir, 24 (19),Z., Johnson, A.T.C., and Drndic´, M. 10560 – 10564.(2010) DNA translocation through 37. Cote, L.J., Kim, F., and Huang, J. (2009)graphene nanopores. Nano Lett., 10 (8), Langmuir−Blodgett assembly of graphite2915 – 2921. oxide single layers. J. Am. Chem. Soc.,26. Schneider, G.F., Kowalczyk, S.W., Calado, 131 (3), 1043–1049.V.E., Pandraud, G., Zandbergen, H.W., 38. Lerf, A., He, H., Forster, M., andVandersypen, L.M.K., and Dekker, C. Klinowski, J. (1998) Structure of graphite

152 5 Graphene-Based Membranes for Separation Engineeringoxide revisited. J. Phys. Chem. B, 102 sheets by pH-assisted selective sedi-(23), 4477–4482. mentation. J. Am. Chem. Soc., 133 (16),39. Dreyer, D.R., Park, S., Bielawski, C.W., 6338 – 6342.and Ruoff, R.S. (2010) The chemistry of 50. Li, H., Song, Z., Zhang, X., Huang, Y., Li,graphene oxide. Chem. Soc. Rev., 39 (1), S., Mao, Y., Ploehn, H.J., Bao, Y., and Yu,228 – 240. M. (2013) Ultrathin, molecular-sieving40. Chen, C., Yang, Q.-H., Yang, Y., Lv, W., graphene oxide membranes for selectiveWen, Y., Hou, P.-X., Wang, M., and hydrogen separation. Science, 342 (6154),Cheng, H.-M. (2009) Self-assembled 95 – graphite oxide membrane. 51. Romanos, G., Pastrana-Martínez, L.M.,Adv. Mater., 21 (29), 3007–3011. Tsoufis, T., Athanasekou, C., Galata, E.,41. Chen, H., Müller, M.B., Gilmore, Katsaros, F., Favvas, E., Beltsios, K.G.,K.J., Wallace, G.G., and Li, D. (2008) Siranidi, E., Falaras, P., Psycharis, V., andMechanically strong, electrically conduc- Silva, A.M.T. (2015) A facile approachtive, and biocompatible graphene paper. for the development of fine-tuned self-Adv. Mater., 20 (18), 3557–3561. standing graphene oxide membranes and42. Wei, N., Peng, X., and Xu, Z. (2014) their gas and vapor separation perfor-Understanding water permeation in mance. J. Membr. Sci., 493, 734–747.graphene oxide membranes. ACS Appl. 52. Yeh, C.-N., Raidongia, K., Shao, J., Yang, Q.-H., and Huang, J. (2015) On the ori-Mater. Interfaces, 6 (8), 5877–5883. gin of the stability of graphene oxide43. Boukhvalov, D.W., Katsnelson, M.I., and membranes in water. Nat. Chem., 7 (2),Son, Y.-W. (2013) Origin of anomalous 166 – 170.water permeation through graphene 53. Putz, K.W., Compton, O.C., Segar,oxide membrane. Nano Lett., 13 (8), C., An, Z., Nguyen, S.T., and Brinson, 3930 – 3935. L.C. (2011) Evolution of order dur-44. Mi, B. (2014) Graphene oxide mem- ing vacuum-assisted self-assembly of graphene oxide paper and associated branes for ionic and molecular sieving. polymer nanocomposites. ACS Nano, 5 Science, 343 (6172), 740–742. (8), 6601–6609.45. Huang, H., Song, Z., Wei, N., Shi, L., 54. Huang, H., Mao, Y., Ying, Y., Liu, Y., Mao, Y., Ying, Y., Sun, L., Xu, Z., andPeng, X. (2013) Ultrafast viscous water Sun, L., and Peng, X. (2013) Salt con-flow through nanostrand-channelled centration, pH and pressure controlledgraphene oxide membranes. Nat. Com- separation of small molecules throughmun., 4, 2979. lamellar graphene oxide membranes.46. Pham, V.H., Cuong, T.V., Hur, S.H., Shin, Chem. Commun., 49 (53), 5963–5965.E.W., Kim, J.S., Chung, J.S., and Kim, E.J. 55. Park, S., Lee, K.-S., Bozoklu, G., Cai,(2010) Fast and simple fabrication of a W., Nguyen, S.T., and Ruoff, R.S. (2008)large transparent chemically-converted Graphene oxide papers modified bygraphene film by spray-coating. Carbon, divalent ions—enhancing mechanical48 (7), 1945–1951. properties via chemical cross-linking.47. Sun, P., Zhu, M., Wang, K., Zhong, ACS Nano, 2 (3), 572–578.M., Wei, J., Wu, D., Xu, Z., and Zhu, 56. Hu, M. and Mi, B. (2014) Layer-by-layerH. (2013) Selective ion penetration of assembly of graphene oxide membranesgraphene oxide membranes. ACS Nano, via electrostatic interaction. J. Membr.7 (1), 428–437. Sci., 469, 80–87.48. Sun, P., Wang, K., Wei, J., Zhong, M., 57. Gahlot, S., Sharma, P.P., Gupta, H.,Wu, D., and Zhu, H. (2014) Effective Kulshrestha, V., and Jha, P.K. (2014)recovery of acids from iron-based Preparation of graphene oxide nano-electrolytes using graphene oxide mem- composite ion-exchange membranes forbrane filters. J. Mater. Chem. A, 2 (21), desalination application. RSC Adv., 47734 – 7737. (47), 24662–24670.49. Wang, X., Bai, H., and Shi, G. (2011) 58. An, Z., Compton, O.C., Putz, K.W.,Size fractionation of graphene oxide Brinson, L.C., and Nguyen, S.T. (2011)

References 153Bio-inspired borate cross-linking in Preparation and characterization ofultra-stiff graphene oxide thin films. Adv. HPEI-GO/PES ultrafiltration mem-Mater., 23 (33), 3842–3846. brane with antifouling and antibacterial59. Choi, W., Choi, J., Bang, J., and Lee, properties. J. Membr. Sci., 447, 452–462.J.-H. (2013) Layer-by-layer assembly of 70. Lee, J., Chae, H.-R., Won, Y.J., Lee,graphene oxide nanosheets on polyamide K., Lee, C.-H., Lee, H.H., Kim, I.-C.,membranes for durable reverse-osmosis and Lee, J.-m. (2013) Graphene oxideapplications. ACS Appl. Mater. Inter- nanoplatelets composite membrane withfaces, 5 (23), 12510–12519. hydrophilic and antifouling properties60. Wang, E.N. and Karnik, R. (2012) Water for wastewater treatment. J. Membr. Sci.,desalination: graphene cleans up water. 448, 223–230.Nat. Nanotechnol., 7 (9), 552–554. 71. Ganesh, B.M., Isloor, A.M., and Ismail,61. Schrier, J. (2010) Helium separation A.F. (2013) Enhanced hydrophilicity andusing porous graphene membranes. J. salt rejection study of graphene oxide-Phys. Chem. Lett., 1 (15), 2284–2287. polysulfone mixed matrix membrane.62. Hauser, A.W. and Schwerdtfeger, P. Desalination, 313, 199–207.(2012) Nanoporous graphene mem- 72. Zhao, C., Xu, X., Chen, J., and Yang,branes for efficient 3He/4He separation. F. (2013) Effect of graphene oxide con-J. Phys. Chem. Lett., 3 (2), 209–213. centration on the morphologies and63. Trinh, T.T., Vlugt, T.J.H., Hägg, M.-B., antifouling properties of PVDF ultrafil-Bedeaux, D., and Kjelstrup, S. (2013) tration membranes. J. Environ. Chem.Selectivity and self-diffusion of CO2 and Eng., 1 (3), 349–354.H2 in a mixture on a graphite surface. 73. Liang, B., Zhan, W., Qi, G., Lin, S.,Front. Chem., 1, 38. Nan, Q., Liu, Y., Cao, B., and Pan, K.64. Sanders, D.F., Smith, Z.P., Guo, R., (2015) High performance grapheneRobeson, L.M., McGrath, J.E., Paul, oxide/polyacrylonitrile composite per-D.R., and Freeman, B.D. (2013) Energy- vaporation membranes for desalinationefficient polymeric gas separation applications. J. Mater. Chem. A, 3 (9),membranes for a sustainable future: 5140 – 5147.a review. Polymer, 54 (18), 4729–4761. 74. Wang, Y., Ou, R., Wang, H., and Xu,65. Kim, H.W., Yoon, H.W., Yoo, B.M., Park, T. (2015) Graphene oxide modifiedJ.S., Gleason, K.L., Freeman, B.D., and graphitic carbon nitride as a modifierPark, H.B. (2014) High-performance for thin film composite forward osmosisCO2-philic graphene oxide membranes membrane. J. Membr. Sci., 475, 281–289.under wet-conditions. Chem. Commun., 75. He, L., Dumée, L.F., Feng, C., Velleman,50 (88), 13563–13566. L., Reis, R., She, F., Gao, W., and Kong,66. Cohen-Tanugi, D. and Grossman, L. (2015) Promoted water transportJ.C. (2014) Mechanical strength of across graphene oxide–poly(amide) thinnanoporous graphene as a desalina- film composite membranes and theirtion membrane. Nano Lett., 14 (11), antibacterial activity. Desalination, 365,6171 – 6178. 126 – 135.67. Cohen-Tanugi, D. and Grossman, 76. Xu, C., Cui, A., Xu, Y., and Fu, X. (2013)J.C. (2014) Water permeability of Graphene oxide–TiO2 composite fil-nanoporous graphene at realistic pres- tration membranes and their potentialsures for reverse osmosis desalination. application for water purification. Car-J. Chem. Phys., 141 (7), 074704. bon, 62, 465–471.68. Qiu, L., Zhang, X., Yang, W., Wang, 77. Pastrana-Martínez, L.M., Morales-Torres,Y., Simon, G.P., and Li, D. (2011) Con- S., Figueiredo, J.L., Faria, J.L., and Silva,trollable corrugation of chemically A.M.T. (2015) Graphene oxide basedconverted graphene sheets in water and ultrafiltration membranes for photocat-potential application for nanofiltration. alytic degradation of organic pollutantsChem. Commun., 47 (20), 5810–5812. in salty water. Water Res., 77, 179–190.69. Yu, L., Zhang, Y., Zhang, B., Liu, 78. Tang, Y.P., Paul, D.R., and Chung, T.S.J., Zhang, H., and Song, C. (2013) (2014) Free-standing graphene oxide

154 5 Graphene-Based Membranes for Separation Engineering thin films assembled by a pressurized 84. Rodewald, P.G. (1976) Oil Recovery by ultrafiltration method for dehydration of Waterflooding Employing Graphite Oxide ethanol. J. Membr. Sci., 458, 199–208. for Mobility Control, http://www.google79. Huang, K., Liu, G., Lou, Y., Dong, Z., (accessed 19 Shen, J., and Jin, W. (2014) A graphene February 2016). oxide membrane with highly selective molecular separation of aqueous organic 85. Rodewald, P.G. (1976) Oil recovery by solution. Angew. Chem. Int. Ed., 53 (27), waterflooding employing graphite oxide 6929 – 6932. for mobility control. US3998270.80. Tsou, C.-H., An, Q.-F., Lo, S.-C., De Guzman, M., Hung, W.-S., Hu, C.-C., 86. McCoy, T.M., Pottage, M.J., and Tabor, Lee, K.-R., and Lai, J.-Y. (2015) Effect of R.F. (2014) Graphene oxide-stabilized microstructure of graphene oxide fab- oil-in-water emulsions: pH-controlled ricated through different self-assembly dispersion and flocculation. J. Phys. techniques on 1-butanol dehydration. J. Chem. C, 118 (8), 4529–4535. Membr. Sci., 477, 93–100.81. Wood, J.A., Schwartz, R., Lill, P.J. (2014) 87. Ba Dung, N., Trung Kien, N., Dryer and water recovery/purification Truong Han, B., Duy Khanh, P., unit employing graphene oxide or perfo- Xuan Loc, D., and Phuong Tung, N. rated graphene monolayer membranes. (2015) The impact of graphene oxide WO2014176541A1. particles on viscosity stabilization for82. Luo, J., Cote, L.J., Tung, V.C., Tan, diluted polymer solutions using in A.T.L., Goins, P.E., Wu, J., and Huang, J. enhanced oil recovery at HTHP off- (2010) Graphene oxide nanocolloids. shore reservoirs. Adv. Nat. Sci.: Nanosci. J. Am. Chem. Soc., 132 (50), Nanotechnol., 6 (1), 015012. 17667 – 17669.83. Kosynkin, D.V., Ceriotti, G., Wilson, 88. Monteiro, O. and Quintero, L. (2012) K.C., Lomeda, J.R., Scorsone, J.T., Graphene-containing fluids for oil Patel, A.D., Friedheim, J.E., and Tour, and gas exploration and production. J.M. (2012) Graphene oxide as a high- WO2012129302A2. performance fluid-loss-control additive in water-based drilling fluids. ACS Appl. 89. Mater. Interfaces, 4 (1), 222–227. lockheed-martin-tests-nanofilters- oil-and-gas-wastewater-management (accessed February 2015).

1556Graphene Coatings for the Corrosion Protection of BaseMetalsRobert V. Dennis, Nathan A. Fleer, Rachel D. Davidson, and Sarbajit Banerjee6.1Introduction to CorrosionThe unfortunate and relentless problem of corrosion of base metals has plaguedmankind for thousands of years. In industrialized societies, the massive costs ofmaintaining and repairing infrastructure can be attributed to a large measure tothe weathering and corrosion of structural components [1–4]. The complexity ofcorrosion processes makes estimation of the true costs of addressing such degra-dation phenomena rather difficult; however, several studies have attempted toprovide some accounting of the financial impact of corrosion on the economy.An influential report from the United States Federal Highway Administration,published in 2002, estimated that the direct cost due to corrosion to the UnitedStates economy in 1998 was USD276 billion (amounting to as much as 3.1% ofthe nation’s gross domestic product) [1]. This staggeringly high monetary cost haslikely only escalated over the last 15 years with diminished spending on infras-tructure and, furthermore, this number does not include indirect costs associatedwith the inhibition and control of corrosion, which are likely just as high, if nothigher, than the direct costs [1–3]. Fundamentally, corrosion of a metal is caused by a series of electrochemicalreactions where concurrent metal dissolution (oxidation) and the reduction ofoxygen in the presence of water result in the loss of metal and the re-formation of amore thermodynamically stable metal oxide. The generalized anodic and cathodicreactions can be written as follows: M(s) → M+(aq.) + e− (6.1)O2 (g) + 2H2O(l) + 4e− → 4OH−(aq.) (6.2)eventually yielding a metal oxide upon further reaction. The metal substrate itselfserves as the conduit for the electrons; charge compensation between the anodicand cathodic half-cells is further facilitated by ion transport through an exter-nal medium. Mitigating corrosion thus fundamentally comes down to inhibitingthis sequence of half-cell reactions by either impeding electron or ion transportGraphene Technology: From Laboratory to Fabrication,First Edition. Edited by Soroush Nazarpour and Stephen R. Waite.© 2016 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2016 by Wiley-VCH Verlag GmbH & Co. KGaA.

156 6 Graphene Coatings for the Corrosion Protection of Base Metals thereby providing a barrier precluding oxygen and water diffusion, or incorporat- ing a sacrificial layer that reacts preferentially instead of the metal surface. While detailed mechanistic understanding is not yet available, graphene coatings that exhibit promise for corrosion inhibition and are the focus of this contribution are thought to mobilize a combination of these modes. As a good first approximation, all coatings that are designed to inhibit cor- rosion of metals can be classified according to the following four mechanisms: (i) cathodic protection; (ii) anodic passivation; (iii) electrolytic inhibition; and (iv) active corrosion inhibition [5–9]. Several particularly effective coatings incorpo- rate multiple modes of action. Figure 6.1 schematically represents the four main modes of corrosion inhibition as well as an additional mode, “self-healing,” which is a broader concept that has been differentiated from active corrosion inhibition for the purposes of this discussion. The concept of cathodic protection is based on the idea that deposition of a more electropositive metal will polarize the substrate Modes of corrosion inhibition Barrier protection Cathodic protection Barrier coating Hydrophobic/superhydrophobic Air Prevents corrosive coating prevents water from H2O/O2/ species from contacing the substrate Zn2+ O2electrolytes reaching the substrate >90° Metal such as Zn Water contact can act as a sacrificial H2O angle anode, polarizing the OH– substrate and oxidizing before the underlying metal ZnImpermeable e– Metal substrate Metal substrate Active corrosion inhibition Anodic passivationPassivating layer Sparingly soluble chromates Strongly binding ligands(conversion coating leach out to form insoluble coordinate to the exposedor anodized coating) precipitates, passivating the surface metal surface, inhibiting further corrosionImpermeable to Chromate inhibitorscorrodant species Layered double hydroxide (LDH) loaded with corrosion inhibitors Topcoat PrimerMetal substrate Metal substrate Metal substrate Metal substrate Rupture of nanocrystal Rupture and release Micelle filled Micelle encapsulation of monomers from with healing offers small particles, releasing Monomers micro and nanocontainers media healing agents on demand Metal substrate Breakdown of coating “Self-healing” Metal substrate Metal substrateVascular system Sol-gel coating repairing Capsules contain sol-gelprovides healing the vold in the coating precursors such as alkoxysilanesmaterial to theruptured area, Metal substrate O OOrepairing the O Si O Si O Si Ocoating surface O OO O Si O Si O Si O O OO Figure 6.1 Schematic depiction of different modes of corrosion inhibition, including barrier protection, cathodic protection, anodic passivation, active corrosion inhibition, and “self- healing” [9]. (IOP Publishing. Reproduced with permission. All rights reserved.)

6.1 Introduction to Corrosion 157metal with the coating thus serving as a sacrificial anode that is preferentially oxi-dized. The most common example of this mode of action is galvanization of steelor other metal substrates by hot-dip or electroplating processes. Other metalssuch as aluminum and magnesium have also commonly been alloyed with zincvia interdiffusion or used independently as sacrificial barrier coatings [10]. It is ofutmost importance that the sacrificial metal and the substrate be properly coupledso that the polarization of the substrate is sufficient to prevent pitting corrosionwhile also avoiding “overprotection” that can lead to hydrogen embrittlement oralkaline attack [5]. Anodic passivation of a metal substrate inhibits corrosion by using a pas-sivating layer to coat the metal surface, which has the effect of suppressingthe redox reactions listed in Eqs (6.1) and (6.2). This mode essentially involvesdeposition of a barrier film on the metal surface that is either impervious toion diffusion or only allows the selective diffusion of specific ions [6, 11]. Thisapproach has been commonly used for a number of decades in the form ofchromate conversion coatings and anodized aluminum. In this approach, anatural oxide layer is typically combined with some sort of passivation layer,which together create a bipolar precipitate that strongly retards ion transportthrough the barrier. Additionally, closely aligned or alternating layers of denselypacked oxides and porous oxides deposited during an anodization or con-version coating step redirect ion transport between the anodic and cathodicsites [6, 12, 13]. A third mode of corrosion resistance, electrolytic inhibition, also involvesshutting down ion transport pathways between the anodic and cathodic sites ofthe metal by using a low-ionic-conductivity matrix or diffusion barrier. Thesecoatings are typically barriers that attempt to limit the transport of corrodantelectrolyte species to the metal by increasing the tortuosity of the conductionpathways and reducing the movement of charge. Finally, active corrosion inhi-bition addresses the inevitable scenario of coating failure. Active corrosioninhibition involves the incorporation of components that can be selectivelyreleased upon failure or in response to external stimuli, thereby reconstitutinga protective barrier at the metal interface [5, 6, 14, 15]. Some examples of theseactive corrosion inhibitors include strongly binding ligands (with high bindingconstants for formation of dative bonds at metal surfaces) or sparingly solubleoxide precursors (with low solubility product, Ksp, values) that can precipitate,even at very low concentrations, to form barrier layers. Typically, such activespecies are encapsulated within appropriate polymeric or porous inorganiccontainers that in turn are embedded within the coating; the active inhibitors arereleased upon coating failure or the initiation of corrosion [5, 15–18]. The broadumbrella of “self-healing” coatings also includes the incorporation of monomersand catalyst particles that can help to actually reconstruct the existing coatingupon failure without necessarily addressing corrosion inhibition. Figure 6.1schematically depicts the main modes of corrosion inhibition with examplesof each.

158 6 Graphene Coatings for the Corrosion Protection of Base Metals During the last couple of decades there has been a renewed push toward the development of novel coating systems spurred by several critical imperatives. An ever-aging global infrastructure requires a re-evaluation of current coating tech- nologies with increasing realization that many conventional coating materials rep- resent a hazard to human health and the environment. One of the single most important forces driving the development of new coating technologies is the strin- gent regulatory environment that addressess the potent carcinogenicity and envi- ronmental concerns of hexavalent chromium, which has been used extensively in chromate conversion coatings and hard-chrome electroplating processes for a number of decades. Hexavalent chromium-based coatings have been commonly used in the industry due to their excellent corrosion resistance and self-healing properties (depicted in Figure 6.1); such coatings also provide excellent sheen and metallic luster that gives it its distinctive aesthetic appeal [5, 19–25]. The exceptional mobility of hexavalent chromium ions that make it ideal for corro- sion inhibition also allow for the facile environmental transport of these species [19, 22, 26, 27]. Given the increasing realization of the long-term environmental legacy of hexavalent chromium, its use is stringently regulated around the globe [19, 22, 23, 25, 28, 29]. Similarly to hexavalent chromium, volatile organic com- pounds (VOCs) found in most polymer and paint coatings (as organic solvents and sometimes toxic curing agents) represent a major occupational safety haz- ard, particularly during application of the coatings [30–32]. Consequently, there is a global push towards sustainable technologies that incorporate earth-abundant and non-toxic components while at the same time providing protection over pro- tracted periods of operation. Sacrificial coatings such as galvanized zinc have been the “gold standard” for corrosion-resistant coatings for more than a century due to the ease of plating and the excellent barrier and sacrificial coating properties that they bestow upon the underlying base metal substrate [33, 34]. Sacrificial coatings such as zinc owe their corrosion-inhibition properties to the difference in reduction potentials between the sacrificial metal and the substrate metal whereby the coated metal is oxidized preferentially, preventing oxidation of the substrate (as schematically illustrated in Figure 6.1) [33]. Unfortunately, the longevity of the corrosion inhibition is directly proportional to the coating thickness and in a sense these coatings function on the basis of continuous dissolution or failure over their operational lifetime. In other words, ensuring longevity of protection often requires considerable cost and weight penalties with the latter being particularly consequential for aerospace and transportation applications since increased weight oftentimes correlates to increased fuel consumption [9, 35]. Since base metals are for the most part a commodity business with small margins, the recent market unpredictability and the dramatic fluctuations in the price of zinc have been a cause for serious con- cern in the steel industry and have substantially cut into profit margins [9]. On the other end of the spectrum, high-performance coatings are more critical than ever for advanced lightweight alloys that are increasingly finding use in aerospace and transportation applications. The highly heterogeneous nature of these alloys, oftentimes characterized by intermetallic or elemental precipitates, make them

6.2 Bare Graphene as a Protective Barrier 159particularly susceptible to corrosion by establishing surface domains with distinc-tive reduction potentials that can each serve as the cathodic and anodic halves ofcorrosion cells. To compound this problem, zinc does not have the ability to pro-tect aluminum and aluminum alloys from corrosion since aluminum lies lower inthe galvanic series than zinc and thus new, more electroactive, coating systems areurgently required [36]. Given the inadequacies of current technologies, the exploration of entirelynew coating paradigms has become of utmost importance. As a carbon-basedelectroactive and highly conducting material that can be interfaced with metals,graphene is an attractive non-metallurgical candidate for protecting base metalseither by itself or as the active element of polymer, metal matrix, or ceramiccomposites [4, 37]. In this chapter, we review current mechanistic understandingof how graphene inhibits corrosion and provide illustrative examples of the useof graphene as a component in- or as a standalone coating for the corrosioninhibition of base metals. The objective of this contribution is to contrastapproaches for corrosion protection based on the use of graphene, explore themechanistic underpinnings of the protection bestowed by graphene coatings,capture a snapshot of this rapidly developing area of research, and to examinethe potential of this material as an alternative to conventional metallurgical filmsor polymeric coatings. Additionally, the authors’ perspective on the outlook forgraphene-based corrosion-resistant coatings has been included with a focus onthe obstacles to commercialization of graphene: price, consistency, and quality.6.2Bare Graphene as a Protective Barrier6.2.1Some Electronic Structure Considerations at Graphene/Metal InterfacesThe conduction and valence bands of graphene, derived from pz orbitals, adopta conical configuration and intersect at the Fermi level (EF) [38, 39]. Remarkably,the bands show a linear dispersion as a function of energy ±1 eV from the pointof intersection (the Dirac point). Such a “slim hourglass” electronic structure hassome peculiarities that render it particularly useful for protecting metal surfaces.The low density of states near EF imply that metals (among other species) canreadily participate in charge transfer interactions with graphene depending onthe relative alignment of their work functions and the extent of overlap of thegraphene π-cloud with metal orbitals of the appropriate symmetry [38, 40].Extensive details of electronic structure consequences of interfacing graphenewith metals and dielectrics have been reviewed elsewhere and are summarizedhere only to provide a perspective of mechanisms for mitigating corrosion [38,41–43]. As a first approximation, charge transfer between metal surfaces andgraphene induces a potential barrier at the graphene/metal interface and theresulting polarization impedes the electron transfer processes necessary for

160 6 Graphene Coatings for the Corrosion Protection of Base Metals corrosion depicted in Eqs (6.1) and (6.2). Unlike in a bulk solid, charge transfer and other perturbations propagate across the 2D geometric structure and can profoundly alter the electronic structure of graphene [38]. The nature of the metal/graphene interface is thus of paramount importance in determining the extent of polarization and consequently the degree of protection afforded to the metal substrate by graphene. Several different types of graphene/metal interactions can be distinguished. For low work-function and highly electropositive metals such as Li, Na, K, and Cs, graphene serves as an electron acceptor and a rigid shift of the band structure is observed as a result of electron doping [9, 38]. In contrast, a broad class of transition- and post-transition metals such as Cu, Ag, Au, and Pt exhibit interactions reminiscent of physisorption accompanied by charge transfer and development of an interfacial dipole as illustrated in Figure 6.2. For these metals, ab initio density functional calculations predict that while at distant separations, the directionality of charge transfer and the magnitude of the interfacial dipole are well predicted by the relative alignments of work functions, at closer separations, exchange repulsion terms assume greater significance [44, 45]. At a Cu/graphene interface, the direction of charge transfer flips from n- to p-type doping with increasing separation between graphene and the underlying Cu(111) surface [46]. In other words, the repulsion between the itinerant electrons in the π-cloud of graphene and the electron gas of the metal contribute significantly to the surface potential difference (denoted as Δv in Figure 6.2) that develops at the interface. This potential difference can thus impede redox processes involved in corrosion. As a third type of metal/graphene interface, for metals such as Ni, Co, Pd, and Ti, the high degree of epitaxial matching of crystal lattices with graphene as well as the strong hybridization of the transition metal dz2 orbitals with the graphene π-cloud profoundly reshapes the electronic structure of graphene, opening up a bandgap at the Fermi level and removing spin degeneracy [45–47]. The interfaced graphene layer thus acquires some carbidic character and is rendered a semi- conductor [9, 38]. A potential difference now exists between the surficial (and sub-surficial) metal layers that are hybridized with the graphene and constitute a semiconductor and the underlying metallic layers that are relatively unperturbed by interfacing with graphene. The semiconductor/metal interface thus established within the metal gives rise to a Schottky barrier to the tunneling of electrons and this potential barrier can further impede electron transfer at the metal/graphene interface. The height of the barrier depends on the pinning of the Fermi level of the carbide-like semiconductor formed at the graphene/metal interface [48]. Indeed, the Schottky barrier represents a major challenge with making ohmic contacts to semiconducting carbon nanotubes. In other words, both physisorption and cova- lent hybridization of graphene on the metal surface give rise to interfacial potential barriers that serve to protect against oxidation of the metal. Another potential mechanism, illustrated in Figure 6.2, derives from the high electrical conductivity of graphene where room-temperature mobilities can readily surpass 10 000 cm2 V−1 s−1. The much higher electrical conductivity of graphene as compared to the underlying metal substrate provides an alternative

6.2 Bare Graphene as a Protective Barrier 161 The impermeability of graphene prevents High surface areacorrosive species from reaching the substrate graphene flakes block pores and conduction pathways for water and ions Metal substrate Metal substrate Barrier Ionic Tortuous protection conduction path M = M+ + e– H2O AnodeO2 + 2H2O + 4e– = 4OH– Electron Corrosion conduction of metals Cathode Metal substrateΔV Physisorption: interfacial dipole Alternative Chemisorption: Schottky barrier electronic pathwayA surface potential discontinuity, ΔV, can Graphene, typically being more inhibit corrosion by impeding charge conductive than the substrate metals, transfer necessary for corrosion allows for electrons to transport from anodic sites away from cathodic sites, disrupting the corrosion cell ΔV Anode Metal substrate Cathode Metal substrateFigure 6.2 Schematic depiction of the four barrier protection; by requiring a tortuousmain modes of corrosion inhibition by path for ion permeation; by formation of agraphene. This graphic depicts the ways potential barrier at the graphene/metal inter-by which graphene can help to impede or face (either a Schottky barrier or interfacialentirely shut down the electrochemical pro- dipole); and by providing an alternative elec-cesses related to corrosion: by providing tronic pathway.

0 h 234.5 h 1752 h 6.3 Graphene Nanocomposites for Corrosion Inhibition 167(a) Galvanized steel 3144 h(b) Uncoated low-alloy steel(c) PEI coating(d) 2 wt% UFG/MWCNT/PEI coating(e) 20 wt% UFG/PEI coatingFigure 6.5 Digital photographs of salt-water immersion measurements on uncoated low-alloy steel, a PEI coating, and a 20 wt% UFG/PEI coating [37]. American Ceramic SocietyBulletin. Used with permission.tortuous path for ion permeation and further likely yield a potential barrier at themetal interface (Figure 6.2) [37]. In an alternative approach developed by the authors of this work, grapheneoxide was used both as the active filler material and as the curing agent and cova-lently linked to the host polymer matrix, thereby achieving excellent dispersionat high loading levels [68]. A commercially available epoxy resin, Araldite 506,was used as the matrix. This resin typically requires a curing agent and/or ele-vated temperatures to initiate the reaction; however, the hydroxyl and carboxylicacid groups of graphene oxide can react with the epoxide groups of the Araldite,thereby constituting a cross-linked network. This approach enabled the incor-poration of graphene loadings up to 50 wt% within the epoxy matrix. Sampleswith high loading levels of graphene oxide exhibited excellent corrosion resistanceupon exposure to 3.5 wt% aqueous solutions of NaCl even after 2160 h of exposure.Lower loading levels (e.g., 5 wt%) did not perform as well in the extended expo-sure tests, likely as a result of the lower cross-linking density (which renders thematrix more permeable to water and corrodant species) as well as the relativelylow amount of the electro-active filler material [68]. Several analogous approaches have been developed to improve the interfacialchemistry between graphene and the polymer matrix including that of Chang

168 6 Graphene Coatings for the Corrosion Protection of Base Metals et al. demonstrated that functionalizing graphene sheets with 4-aminobenzoic acid allows for their facile incorporation within PANI [65]. PANI by itself is one of most promising polymeric systems used for corrosion protection of steels as a result of its high electroactivity and facile charge transfer with steel substrates [72]. Unfortunately, it can be difficult to achieve good dispersion of graphene within PANI; however, excellent dispersion was achieved with the mediation of 4-aminobenzoic acid and authors report excellent corrosion-resistant barrier properties. Tafel analysis for this system showed a significant drop in corrosion current density and a shift to a more positive potential from the bare steel or that of the PANI coating alone (from 3.70 μA cm−2 and −647 mV for PANI to 0.38 μA cm−2 and −537 mV for a 0.5 wt% graphene/PANI coating), which in turn also decreased the estimated corrosion rate by approximately two orders of magnitude [65]. The incorporation of graphene within PMMA has been achieved using a “grafting from” approach by first binding an initiator for atom-transfer radical polymerization, (N-(2-aminoethyl)-2-bromo-2-methylpropanamide, NABM), and subsequently polymerizing methyl methacrylate (MMA) from the surface of graphene [69]. Graphene oxide loadings of up to 81 wt% within PMMA are achieved by this method; the corrosion current density of a coated copper sheet decreased by 3–4 orders of magnitude upon coating with the graphene/PMMA composite. The coating was able to provide corrosion protec- tion to the copper surface even up to 100 h of exposure to the 3.5% NaCl solution [69]. As a concluding example, Okafor et al. found that graphene that was incorpo- rated within a hybrid polymer of epoxy ester, siloxane, and urea afforded excellent corrosion protection properties for Al coupons [67]. Even at concentrations as low as 1–2 wt% of graphene it was possible to achieve a significant diminution in corrosion, as corroborated by electrochemical impedance spectroscopy (EIS) and Tafel analysis of potentiodynamic polarization experiments. Tafel analysis of polarization data indicates a drop in the corrosion current density of about two orders of magnitude as compared to the neat hybrid polymer coating [67]. From a practical perspective, graphene/polymer nanocomposites likely represent the most viable option for large-scale applications given the drawbacks of single- and few-layered graphene as stand-alone coatings summarized in the preceding section. The widespread use of nanocomposites incorporating carbon black, microstructured carbon, carbon fibers, and, increasingly, carbon nanotubes, renders the incorporation of graphene within polymers a relatively facile “drop- in” solution for a diverse range of paints and coatings. The static dissipation properties arising from the formation of a percolative network are an added bonus, particularly for packaging applications in the semiconductor industry. 6.4 Graphene/Metal Nanocomposites for Corrosion Inhibition The incorporation of graphene within metallurgical thin films yields inter- esting metal matrix composites some of which hold promise for corrosion

6.4 Graphene/Metal Nanocomposites for Corrosion Inhibition 169protection [73]. Metallurgical films and alloys are extremely effective at pre-venting corrosion of various metal substrates and can serve as sacrificial anodeswhen used as coatings. Alternatively, metals that lie higher in the reductivepotential series can also serve to protect substrates by creating a barrier thatrequires an increased potential to initiate oxidation. Incorporation of graphene(or carbon nanotubes) can increase the formability of the metal coating andpotentially offer enhanced electrical conductivity. The primary drawback is thatdispersion of graphene within metals remains a formidable challenge and thedissimilar graphene/metal interface is prone to debonding, which in turn can leadto incipient porosity and serve as the nucleation point for initiation of corrosion[73]. This drawback limits the amount of graphene that can be incorporatedwithin such composites. Corrosion studies have been performed on several graphene/metal compositematerials including Ni/graphene [74, 75], Zn/graphene [76], and Sn/graphene[77]. These coatings were all synthesized via combined electrophoretic depositionand electroplating from dispersions of chemically modified graphene or exfo-liated graphene in plating baths. The substrate used in all cases was mild steel.Based on analysis of X-ray diffraction data, the incorporation of graphene bringsabout a substantial change in the texture and grain size of the electrodepositedfilms. Table 6.2 collates the percentage decrease in grain size reported by variousauthors [74–77]. While the values for each trial cannot be compared directly, thegeneral trends suggest a pronounced decrease of grain size upon incorporation ofgraphene. The decrease in grain size can be attributed to an increased density ofnucleation sites on the metal surface as a result of the presence of graphene, whichcould potentially inhibit grain growth. Changes in texture could also potentiallyresult from changes in the preferred crystallographic growth planes duringdeposition as a result of deposition onto graphene and not metal surfaces. Kumarand Berlia have postulated that graphene could limit access of solution phasemetal to the substrate surface, thereby limiting particle growth [76, 77]. Severalof these studies also indicate that the surfaces of the coatings exhibit protrusionsof hillock structures upon the incorporation of graphene. These protrusionsTable 6.2 Average grain size calculated using the Scherrer equation and the reduction ingrain size upon incorporation of graphene [74–77]. Grain size (nm) Reduction in grain size (%) Metal CompositeNi [74] 30 20 33Ni [75] 35 ∼19 46Ni [75] 35 ∼16 54Zn [76] 70 11Sn [77] 79.46 62 5.6 75

170 6 Graphene Coatings for the Corrosion Protection of Base Metals Table 6.3 Summary of Tafel analysis results for Ni/graphene, Zn/graphene, and Sn/graphene composite coatings [74–77]. Icorr (A cm−2) Composite Ecorr (V) Composite Metal MetalNi [74] 15.8 6.687 −0.492 −0.398Ni [75] 19.1 3.02 −0.2665 −0.2512Ni [75] 19.1 0.398 −0.2665 −0.2346Zn [76] 19.86 6.82Sn [77] 1.365 0.815 0.915 0.920 −0.573 −0.537suggest the partial segregation of graphene domains and can potentially be sitesfor initiation of failure. The metal matrix composites show an appreciable increase of microhardness,which can be attributed to the high strength of graphene as well as the small grainsize achieved in the electrodeposited films. The diminution in grain size preventsthe build-up and movement of dislocations [75]. Kumar et al. suggest that despitethe presence of hillocks, the overall density of pits is reduced in Zn/graphene coat-ings upon graphene incorporation, likely as a result of the ability of graphene tobridge or fill gaps [76]. The reduced density of pits is thought to be favorable forthe protection of the underlying metal substrates since such defects can serve toinitiate corrosion. In each study, Tafel analysis was performed on electrodepositedbare metal and the metal/graphene composite films. The results of these studiesare summarized in Table 6.3. In all cases, the corrosion current density (Icorr) wasobserved to substantially decrease for the composite material and the corrosionpotential (Ecorr) became more positive. The corrosion rate for the Zn/graphenecomposite decreased fourfold as compared to the bare metal; similarly, the corro-sion rate for the Sn/graphene composite was about 60% of the value for the bareSn film. Graphene oxide has also been used instead of reduced graphene oxide withincobalt/graphene oxide composite coatings deposited by electrodeposition ontomild steel [78]. Graphene oxide can be used with a wider variety of solventswhen compared to reduced graphene oxide and the oxygen groups present onthe surface allow for a greater range of coordinative interactions with a metalsubstrate. Analogous to the results summarized in Table 6.2, the grain size ofcobalt was reduced to 20 ± 2 nm for a cobalt/graphene oxide composite coatingcompared to a value of 50 ± 5 nm for the bare cobalt film [78]. These authors alsoobserved a change in the preferred growth orientation of the cobalt films uponincorporation of graphene. Tafel plot analysis indicated corrosion potentials of−0.3149 and −0.3597 V for the bare cobalt and cobalt/graphene oxide compositecoatings, respectively. The corrosion current decreased from 9.70 × 10−6 A cm−2for the bare cobalt films to 3.04 × 10−6 A cm−2 for the cobalt/graphene oxidecoatings. The corrosion rate also decreased similarly for the composite, going

6.5 Graphene/Ceramic Nanocomposites for Corrosion Inhibition 171from 4.98 × 10−2 mm yr−1 for the bare metal coating to 1.56 × 10−2 mm yr−1 forthe cobalt/graphene oxide coating [78]. Graphene/metal nanocomposites thus show promising performance for corro-sion inhibition. Graphene is observed to decrease the grain size of electroplatedmetal films and to increase their hardness. By providing more nucleation sites andenabling homogeneous deposition with reduced pit density, the nanocompositecoatings increase the resistance of the electrodeposited films to corrosive attack.Further optimization of graphene/metal interfaces is clearly required to mitigatea major probable reason for failure.6.5Graphene/Ceramic Nanocomposites for Corrosion InhibitionSeveral reports in the literature suggest the attachment of silica or alumina prior toincorporation within an epoxy resin to not only disperse the material more read-ily, but also in some cases to prevent the graphene from having any deleteriouseffects on the corrosion resistance [58, 66, 70, 71]. Some research has suggestedthat because graphene is so electro-active it could actually increase the degreeof corrosion for the underlying metal by increasing the number of active cath-ode sites throughout the coating matrix, essentially setting up a graphene/metalcouple that causes galvanic corrosion [56, 70, 79]. Consequently, a number of dif-ferent approaches have been developed to incorporate graphene within a numberof commercially available epoxy resins wherein the graphene is first encapsulatedby an insulating ceramic layer [66, 70, 71, 79, 80]. The enhancement in corrosionresistance that is observed for such composites is predicated entirely on the highsurface area and aspect ratio of the graphene, which implies that any corrosivespecies must take a tortuous path to reach the metal surface (Figure 6.2). As arepresentative example of this approach, Sun et al. showed that graphene encapsu-lated by a nanometer-sized layer of SiO2 created an effective barrier-type coatingwhen dispersed within a polymer matrix, thereby significantly increasing the cor-rosion protection afforded to the underlying copper substrates [70]. Alternatively, Aneja et al. recently described an approach where theyused silica not as an electrically insulating barrier but instead as a means toattach the graphene to a steel surface [58]. In this work, the authors used (3-aminopropyl)triethoxysilane (APTES) to create a functionalized graphene surfacethat could then be reacted with the steel surface to create a graphene/silica com-posite that was bonded to hydroxyl groups on the steel substrate. The subsequentdeposition of an epoxy layer further enhances the corrosion-resistant propertiesthrough its barrier characteristics. Remarkably, potentiodynamic polarizationmeasurements indicate that these silica-functionalized graphene coatings aresuperior even to samples that have undergone conventional pretreatment withchrome. The corrosion current density of the silica/graphene coating wasapproximately three orders of magnitude lower than that of bare steel alone andan order of magnitude lower than the chromium-pretreated sample. In addition,

172 6 Graphene Coatings for the Corrosion Protection of Base Metals a pronounced shift to more positive corrosion potentials was observed for the silica/graphene/epoxy composites. Salt fog testing of these samples as well as EIS data corroborate the results of Tafel analysis, confirming the formation of an excellent corrosion-resistant coating [58]. Another study by Khalil et al. examined the corrosion inhibition afforded by a nickel/graphene/anatase-TiO2 coating on mild steel [81]. This coating design seeks to combine the desirable properties of graphene/metal composites discussed in the preceding section with the possibility for a further increase of stability and increase in hardness through the incorporation of ceramic materials. Graphene/TiO2 composites were prepared separately prior to being electrodeposited from a plating bath onto mild steel alongside metallic nickel. A coating thickness of 20 μm was achieved with a concentration of 0.4 g L−1 graphene-TiO2 particles in the plating bath at a current density of 10 mA cm−2. X-ray diffraction analysis indicates an average particle size of 24 nm for a bare nickel coating and 20 nm for the nickel/graphene/TiO2 composite deposited under the same conditions. No major changes in the preferred orientation of growth were observed in the diffraction data. Tafel analysis showed a decrease in corrosion current density and corrosion rate with an increase in coating thick- ness. When comparing films of the maximum thickness, the corrosion current density decreased from 1.53 × 10−6 A cm−2 for the nickel/graphene composite to 3.46 × 10−8 A cm−2 for the nickel/graphene/TiO2 composite. The corrosion rate also decreased from 0.0183 mm yr−1 for nickel/graphene to 0.0004 mm yr−1 for the nickel/graphene/TiO2 coating. Nyquist plots showed an increase in polarization resistance from 3.90 × 103 Ω for nickel/graphene on mild steel to 3.31 × 104 Ω for nickel/graphene/TiO2. The inclusion of ceramic components can thus enhance the corrosion resistance of the first two classes of coatings discussed in this chapter. For graphene/polymer composites, ceramic components can facilitate the immobilization of graphene to SiO2 surfaces or mitigate graphene/steel galvanic couples. For graphene/metal composites, the inclusion of ceramic components can further increase the hard- ness of the coatings and tortuosity of the ion permeation pathways. 6.6 Summary and Future Outlook In this chapter, we have attempted to capture a snapshot of a rapidly emerging discipline that could potentially yield some of the first large-scale commercial products incorporating graphene. Protecting base metals is an urgent imperative not just to prolong the longevity of infrastructure but also to facilitate the use of base metals in many emerging applications related to clean energy. Concerns regarding ecological toxicity of conventional coating materials have created interesting opportunities for the adoption of new technologies. The interesting results obtained for graphene composites suggest that this material could be ideally poised for widespread industrial deployment. It is clear that the use of

6.6 Summary and Future Outlook 173graphene by itself is unlikely to be practical over prolonged periods of operationgiven the inevitable presence of extended defects and, therefore, the rationaldesign of polymer, ceramic, and metal composites remains a critical imperative. Mechanistic understanding of how graphene reacts with metal substratesremains incomplete, particularly with regard to the influence of additionalpolymeric, metal, or ceramic matrices. In this work, we have discussed severalmechanistic possibilities including the development of interfacial potentialbarriers as a result of exchange repulsions or covalent hybridization, impositionof a highly tortuous path for ion permeation as a result of its high surface area,disruption of electron transport from the anode to the cathode as a result of itshigh electrical conductivity, and the establishment of an impermeable barrier as aresult of its tightly packed covalently bonded structure (Figure 6.2). It is notewor-thy that several graphene/polymer and graphene/metal matrix composites showexcellent corrosion inhibition not just under potentiodynamic and EIS testingbut also over prolonged exposure to accelerated testing environments. A major challenge in the discipline is the wide diversity of materials that aredesignated as being graphene, spanning a broad range of thicknesses, lateraldimensions, and extent of functionalization. The inconsistencies noted in theliterature possibly derive in large measure from the widely heterogeneousmaterials used within coatings. With increased quality and consistency ofavailable graphene materials, it is expected that more definitive answers willbecome available regarding the modes of action. The availability of higherquality and consistent graphene materials will also facilitate the development ofsystematic structure–function correlations (including through high-throughputexperimental methods coupled with the appropriate data analytics), which willsubsequently inform optimization of coating formulations much in the same waythat integrated computational materials engineering approaches are currentlybeing used for the development of new alloys. A second major problem pertains to the mode of application. Most studies thusfar use laboratory equipment and testing under controlled conditions. For deploy-ment on an industrial scale, methods such as wet casting, electrodeposition, androll-to-roll printing will need to be developed (which further requires the avail-ability of large amounts of high-quality samples). The results of field testing are yetto become publicly available although it is clear that several companies have nowadvanced graphene coatings to pilot-plant and field-testing scales. As with otherareas of coatings research, the development of multifunctional and “self-healing”coatings is a particularly attractive frontier. The authors hope that this contribu-tion will provide further impetus to this nascent discipline that shows exceptionalpromise for rapid commercialization.AcknowledgmentsWe gratefully acknowledge Dr Tapan K. Rout and Dr Anil Gaikwad of Tata Steelfor helpful discussions. Our work on graphene nanocomposite coatings was

174 6 Graphene Coatings for the Corrosion Protection of Base Metals funded by the New York State Pollution Prevention Institute, Tata Steel, and the Indo-US Science and Technology Forum.References 1. Koch, G.H., Brongers, M.P.H., 16. Shchukin, D.G., Zheludkevich, M., and Thompson, N.G., Virmani, Y.P., and Möhwald, H. (2006) J. Mater. Chem., 16, Payer, J.H. (2002) Corrosion Costs and 4561. Preventive Strategies in the United States, FHWA, Washington, D.C. 17. Lamaka, S.V., Zheludkevich, M.L., Yasakau, K.A., Serra, R., Poznyak, S.K., 2. National Research Council (2011) and Ferreira, M.G.S. (2007) Prog. Org. Research Opportunities in Corrosion Coat., 58, 127–135. Science and Engineering, The National Academies Press, Washington, DC. 18. Zheludkevich, M.L., Shchukin, D.G., Yasakau, K.A., Möhwald, H., and 3. Winkleman, A., Svedberg, E.B., Schafrik, Ferreira, M.G.S. (2007) Chem. Mater., R.E., and Duquette, D.J. (2011) Adv. 19, 402–411. Mater. Process., 169, 26–31. 19. Kendig, M.W. and Buchheit, R.G. (2003) 4. Böhm, S. (2014) Nat. Nanotechnol., 9, Corrosion, 59, 379–400. 741 – 742. 20. Groshart, E. (1997) Met. Finish., 95, 5. Presuel-Moreno, F., Jakab, M.A., 70 – 72. Tailleart, N., Goldman, M., and Scully, J.R. (2008) Mater. Today, 11, 14–23. 21. Bibber, J.W. (2007) J. Appl. Surf. Finish., 2, 273–275. 6. Hughes, A.E., Cole, I.S., Muster, T.H., and Varley, R.J. (2010) NPG Asia Mater., 22. Bibber, J.W. (2008) Met. Finish., 106, 2, 143–151. 41 – 46. 7. Figueira, R.B., Silva, C.J.R., and Pereira, 23. Bibber, J.W. (2011) in Management of E.V. (2014) J. Coatings Technol. Res., 12, Hazardous Residues Containing Cr (VI) 1 – 35 (ed M.J. Balart Murria), Nova Science Publishers, Inc., Hauppauge, NY, USA, 8. Zheludkevich, M.L., Salvado, I.M., and pp. 37–49. Ferreira, M.G.S. (2005) J. Mater. Chem., 15, 5099–5111. 24. Chen, T., Li, W., and Cai, J. (2011) RSC Adv., 1, 607–610. 9. Dennis, R.V., Patil, V., Andrews, J.L., Aldinger, J.P., Yadav, G.D., and Banerjee, 25. Osborne, J.H. (2001) Prog. Org. Coat., S. (2015) Mater. Res. Express, 2, 032001. 41, 280–286.10. Prosek, T., Nazarov, A., Bexell, U., 26. Shchukin, D.G., Zheludkevich, M., Thierry, D., and Serak, J. (2008) Corros. Yasakau, K., Lamaka, S., Ferreira, M.G.S., Sci., 50, 2216–2231. and Möhwald, H. (2006) Adv. Mater., 18, 1672 – 1678.11. Worley, B.C., Ricks, W.A., Prendergast, M.P., Gregory, B.W., Collins, R., 27. Katz, S.A. and Salem, H. (1993) J. Appl. Cassimus, J.J., and Thompson, R.G. Toxicol., 13, 217–224. (2013) Langmuir, 29, 12969–12981. 28. Anon (1998 Edited by Agency for Toxic12. Cole, I.S., Muster, T.H., Lau, D., Wright, Substances and Disease Registry,) , N., and Azmat, N.S. (2010) J. Elec- Department of Health and Human trochem. Soc., 157, C213–C222. Services, Atlanta, GA.13. Sato, N. (1989) Corrosion, 45, 354–368. 29. Baral, A. and Engelken, R.D. (2002) Env-14. Clark, W.J., Ramsey, J.D., McCreery, iron. Sci. Policy, 5, 121–133. R.L., and Frankel, G.S. (2002) J. Elec- 30. Jagtap, S.D., Tambe, S.P., Choudhari, trochem. Soc., 149, B179–B185. R.N., and Mallik, B.P. (2014) Prog. Org.15. Grigoriev, D., Akcakayiran, D., Coat., 77, 395–402. Schenderlein, M., and Shchukin, D. (2014) Corrosion, 70, 446–463. 31. Kou, Y., Wang, J., Lv, M., and Jian, X. (2010) Polym. Int., 59, 107–111. 32. Elmore, J.D., Kincaid, D.S., Komar, P.C., and Nielsen, J.E. (2002) J. Coating. Technol., 74, 63–72.

References 17533. Marder, A.R. (2000) Prog. Mater Sci., 45, 50. Leenaerts, O., Partoens, B., and Peeters,191 – 271. F.M. (2008) Appl. Phys. Lett., 93, 193107.34. Hosking, N.C., Ström, M.A., Shipway, 51. Bunch, J.S., Verbridge, S.S., Alden, J.S.,P.H., and Rudd, C.D. (2007) Corros. Sci., van der Zande, A.M., Parpia, J.M.,49, 3669–3695. Craighead, H.G., and McEuen, P.L.35. Rout, T.K., Gaikwad, A.V. and (2008) Nano Lett., 8, 2458–2462. Dingemans, T. (2011) Patent Specifi- 52. Prasai, D., Tuberquia, J.C., Harl, R.R., cation 2011035920A1 Jennings, G.K., and Bolotin, K.I. (2012)36. Dennis, R.V., Viyannalage, L.T., Aldinger, ACS Nano, 6, 1102–1108. J.P., Rout, T.K., and Banerjee, S. (2014) Ind. Eng. Chem. Res., 53, 18873–18883. 53. Singh Raman, R.K.,37. Dennis, R.V., Viyannalage, L.T., Gaikwad, Chakraborty Banerjee, P., Lobo, D.E., A.V., Rout, T.K., and Banerjee, S. (2013) Gullapalli, H., Sumandasa, M., Kumar, Am. Ceram. Soc. Bull., 92, 18–24. A., Choudhary, L., Tkacz, R., Ajayan,38. Schultz, B.J., Dennis, R.V., Lee, V., P.M., and Majumder, M. (2012) Carbon, and Banerjee, S. (2014) Nanoscale, 6, 50, 4040–4045. 3444 – 3466.39. Novoselov, K.S., Geim, A.K., Morozov, 54. Kirkland, N.T., Schiller, T., Medhekar, S.V., Jiang, D., Katsnelson, M.I., N., and Birbilis, N. (2012) Corros. Sci., Grigorieva, I.V., Dubonos, S.V., and 56, 1–4. Firsov, A.A. (2005) Nature, 438, 197 – 200. 55. Singh Raman, R.K. and Tiwari, A. (2014)40. Sarkar, S., Moser, M.L., Tian, X., Zhang, JOM, 66, 637–642. X., Al-Hadeethi, Y.F., and Haddon, R.C. (2014) Chem. Mater., 26, 184–195. 56. Schriver, M., Regan, W., Gannett, W.J.,41. Batzill, M. (2012) Surf. Sci. Rep., 67, Zaniewski, A.M., Crommie, M.F., 83 – 115. and Zettl, A. (2013) ACS Nano, 7,42. Bekyarova, E., Sarkar, S., Wang, F., Itkis, 5763 – 5768. M.E., Kalinina, I., Tian, X., and Haddon, R.C. (2013) Accounts Chem. Res., 46, 57. Keil, P., Frahm, R., and 65 – 76. Lützenkirchen-Hecht, D. (2010) Corros.43. Dahal, A. and Batzill, M. (2014) Sci., 52, 1305–1316. Nanoscale, 6, 2548–2562.44. Lee, V., Park, C., Jaye, C., Fischer, 58. Aneja, K.S., Bohm, S., Khanna, A., D.A., Yu, Q., Wu, W., Liu, Z., Bao, J., and Bohm, M. (2015) Nanoscale, 7, Pei, S.-S., Smith, C., Lysaght, P., and 17879 – 17888. Banerjee, S. (2010) J. Phys. Chem. Lett., 1, 1247–1253. 59. Sahu, S.C., Samantara, A.K., Seth, M.,45. Giovannetti, G., Khomyakov, P.A., Parwaiz, S., Singh, B.P., Rath, P.C., and Brocks, G., Karpan, V.M., van den Brink, Jena, B.K. (2013) Electrochem. Commun., J., and Kelly, P.J. (2008) Phys. Rev. Lett., 32, 22–26. 101, 026803.46. Khomyakov, P.A., Giovannetti, G., 60. Lee, V., Whittaker, L., Jaye, C., Baroudi, Rusu, P.C., Brocks, G., van den Brink, K.M., Fischer, D.A., and Banerjee, S. J., and Kelly, P.J. (2009) Phys. Rev. B, 79, (2009) Chem. Mater., 21, 3905–3916. 195425.47. Abtew, T., Shih, B.-C., Banerjee, S., 61. Lih, E.T.Y., Zaid, R.b.M., Ling, T.L., and Chong, K.F. (2012) Int. J. Chem. Eng. Appl., 3, 453. 62. Bierwagen, G., Brown, R., Battocchi, D., and Hayes, S. (2010) Prog. Org. Coat., 67, 195–208. 63. Orloff, N.D., Long, C.J., Obrzut, J., Maillaud, L., Mirri, F., Kole, T.P., McMichael, R.D., Pasquali, M., Stranick,and Zhang, P. (2013) Nanoscale, 5, S.J., and Alexander Liddle, J. (2015) Sci.1902 – 1909. Rep., 5, 17019.48. Tersoff, J. (1984) Phys. Rev. Lett., 52, 64. Kim, H., Abdala, A.A., and Macosko,465 – 468. C.W. (2010) Macromolecules, 43,49. Zhang, D., Wang, D.Z.-R., Creswell, R., 6515 – 6530.Lu, C., Liou, J., and Herman, I.P. (2015) 65. Chang, C.-H., Huang, T.-C., Peng, C.-W.,Chem. Mater., 27, 5032–5039. Yeh, T.-C., Lu, H.-I., Hung, W.-I., Weng,

176 6 Graphene Coatings for the Corrosion Protection of Base Metals C.-J., Yang, T.-I., and Yeh, J.-M. (2012) Pfefferkorn, F.E., and Li, X.-C. (2012) Carbon, 50, 5044–5051. Scripta Mater., 67, 29–32.66. Ma, Y., Di, H., Yu, Z., Liang, L., Lv, L., 74. Kumar, C.M.P., Venkatesha, T.V., and Pan, Y., Zhang, Y., and Yin, D. (2015) Shabadi, R. (2013) Mater. Res. Bull., 48, Appl. Surf. Sci., 360, 936–945 1477 – 1483.67. Okafor, P.A., Singh-Beemat, J., and 75. Jiang, K., Li, J., and Liu, J. (2014) RSC Iroh, J.O. (2015) Prog. Org. Coat., 88, Adv., 4, 36245–36252. 237 – 244. 76. Kumar, M.K.P., Singh, M.P., and68. Patil, V., Dennis, R.V., Rout, T.K., Srivastava, C. (2015) RSC Adv., 5, Banerjee, S., and Yadav, G.D. (2014) 25603 – 25608. RSC Adv., 4, 49264–49272. 77. Berlia, R., Kumar, M.K.P., and Srivastava,69. Qi, K., Sun, Y., Duan, H., and Guo, X. C. (2015) RSC Adv., 5, 71413–71418. (2015) Corros. Sci., 98, 500–506. 78. Liu, C., Su, F., and Liang, J. (2015) Appl.70. Sun, W., Wang, L., Wu, T., Pan, Y., and Surf. Sci., 351, 889–896. Liu, G. (2015) J. Mater. Chem. A, 3, 79. Yang, Z., Sun, W., Wang, L., Li, S., Zhu, 16843 – 16848. T., and Liu, G. (2015) Corros. Sci., 103,71. Yu, Z., Di, H., Ma, Y., Lv, L., Pan, Y., 312 – 318 Zhang, C., and He, Y. (2015) Appl. Surf. 80. Sun, W., Wang, L., Wu, T., Pan, Y., and Sci., 351, 986–996. Liu, G. (2014) Carbon, 79, 605–614.72. Rout, T.K., Jha, G., Singh, A.K., 81. Khalil, M.W., Salah Eldin, T.A., Hassan, Bandyopadhyay, N., and Mohanty, O.N. H.B., El-Sayed, K., and Abdel Hamid, Z. (2003) Surf. Coat. Tech., 167, 16–24. (2015) Surf. Coat. Tech., 275, 98–111.73. Chen, L.-Y., Konishi, H., Fehrenbacher, A., Ma, C., Xu, J.-Q., Choi, H., Xu, H.-F.,

1777Graphene Market ReviewMarko Spasenovic7.1IntroductionThe first production of graphene in the Geim laboratory in 2004 [1], with thenow famous scotch-tape method, locked the interest of a large body of scientificresearchers. Nearly all of graphene’s spectacular predicted properties have sincebeen confirmed experimentally. The scotch-tape, or exfoliation method, yieldsgraphene of a very high quality, but of a small size, with an area generally on theorder of a few tens of micrometers squared. One would typically be lucky to findseveral flakes of that size on a substrate that is 1 cm2. The extremely low yield ofmechanical exfoliation has not been a problem for testing fundamental science;however, it would be too small for any conceivable application. Already in 2006, researchers succeeded in growing large-area graphene bychemical vapor deposition (CVD) [2], the first method of mass-producinggraphene. CVD yields almost 100% coverage of metal films with monolayergraphene. The method of CVD growth on metal and subsequent transfer ofgraphene to various useful substrates has rapidly overcome many hurdles [3, 4],recently yielding a continuous film of graphene on a 100 m roll [5]. The growingnumber of graphene researchers coupled with the wide accessibility of CVDgrowth chambers quickly led to a blossoming of the number of companiesmarketing graphene sheets. The end user has for the most part been the labresearcher, with almost no excursion to industrial space. The primary reason forthe relatively slow inclusion of industry has of course been the price of graphene,which is still higher than that of silicon, and orders of magnitude higher thanthe price of competing transparent conductor material, indium tin oxide (ITO)[6]. Since graphene-based devices could not beat the price of devices basedon more mature materials, the young graphene industry has been looking forthose applications where graphene can offer novel benefits or unprecedentedfunctionality. In more recent years, graphene obtained with chemical methods,such as liquid phase exfoliation [7] or the modified Hummers method, hasbecome an instant hit with lab researchers as well as a myriad of companies.Although chemically exfoliated graphene is of lower quality than sheet graphene,Graphene Technology: From Laboratory to Fabrication,First Edition. Edited by Soroush Nazarpour and Stephen R. Waite.© 2016 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2016 by Wiley-VCH Verlag GmbH & Co. KGaA.

178 7 Graphene Market Review it is cheaper to produce and is readily obtained in large volumes [8]. Graphene obtained with these methods is typically in the form of a powder, which means it can be used as an additive in composites, opening a door to a market of a size much larger than the graphene research market [9], spreading across the auto- motive, sporting, construction, aeronautic, and other traditional large industries, but also affecting emerging technologies such as 3D printing, wearable clothing, and printed electronics. This most recent development pushed graphene closer to real-world applications and ignited a new direction for the market to expand into, enabling a step-up in production volumes and a further reduction of the price of graphene. In addition to few-layer graphene (also dubbed as “graphene nanoplatelets” (GNPs)) some chemical methods yield carbon nanostructures of various shapes. The wide variety of available chemistry results in as wide a variety of graphene-related products, in fact an entire family of “graphenes.” There are thus almost as many variants of graphene as there are graphene producers, setting standardization as priority number one for the coming years, but also creating confusion and adding to the difficulty of predicting any future market volumes. One fact is certain: graphene is a disruptive technology that holds enormous potential for commercialization, the realization of which has just begun. It is also an enabling technology, that will infiltrate other key sectors such as 3D printing, healthcare, automotive, aircraft, sports, construction, and others. This chapter will cover the historical development of the early graphene mar- ket (Section 7.2), the current state of the market (Section 7.3), and some future projections (Section 7.4). 7.2 Graphene Market: Past and Present The early graphene market (2009–2012) revolved around feeding the appetite of the rapidly growing graphene research community. Probably the first company to offer large area graphene in its product catalog was Graphene Supermarket [10], established in 2009 in Calverton, NY. The Supermarket was the sales outpost of Graphene Laboratories, a research start-up run by a few scientists turned entrepreneur. Graphene Supermarket offered CVD-grown graphene on copper substrates and on the most commonly used SiO2/Si substrate, which allows tuning the carrier concentration by back gating. The initial offering was basic, the only other item in the catalog being clean SiO2/Si wafers for customers to do their own exfoliation; however, due to the exploding demand and the Supermarket’s ability to deliver high-quality graphene in a hassle-free manner, the recipe worked like a charm. This simple yet effective recipe soon spread to Europe and the Far East, with the establishment of Graphenea (Spain) in 2010 [11] and Graphene Square (South Korea) in early 2012 [12]. These three managed to quickly master CVD growth and establish themselves as the default graphene suppliers across three continents, setting any latecomers in a difficult position of having to invent advanced applications for an immature material. Still,

7.2 Graphene Market: Past and Present 179with the low entry cost of CVD chambers (of the order of USD100k) and quickmaturing of the technology of CVD growth, even these three companies had tokeep inventing to alleviate the effect of research labs and universities doing theirown growth. In fact, in the short term of the early graphene market, it turned out to be moreprofitable to provide CVD chambers themselves than the graphene made in thosechambers, as exemplified by the success of planarTECH [13]. planarTECH startedin 2012 to sell CVD chambers specially tuned for graphene, and already in 2014surpassed USD1 million in revenue (the entire graphene market turnover in 2013was 6.6 million EUR [14], according to some, and 20 million EUR, according toothers (IDTechEx presentations)). planarTECH had aimed to double their revenuein 2015 (J. Patrick Frantz, private communication). Since similar chambers can beused to grow other monolayer materials such as transition metal dichalcogenides,which are gaining popularity, marketing CVD machines should prove to be solidbusiness in the medium term as well. In the long term, however, graphene holds enormous potential (see Section10.3), but the trick for success lies in recognizing the correct applications for eachvariant of graphene and aligning to the needs of the target customer accordingly.The two oldest graphene producers, Graphene Laboratories and Graphenea areproving to be extremely versatile and inventive in that regard, having significantlyexpanded their product catalogs and engaged in various collaborations with otherindustries. Since these two are excellent examples of successful companies in theearly graphene market, we will focus more on their business strategies before mov-ing to examples of smaller companies. As of date, Graphenea’s main focus is stillon the research community, although an increasing percentage of business is com-ing from sales to companies (Jesus de la Fuente, private communication). Withthe number of graphene-related scientific publications per year soaring in thethousands, this might be a good strategy; however, Graphenea will have to keepfollowing the brutally fast research developments to stay on top. What the com-pany has done so far is to perfect the process of transfer of graphene from copperto custom substrates, offering graphene on PET, glass, SiO2/Si, TEM grids, andmicrocavities as standard in the catalog, with a promise to be able to transfer onalmost any substrate that a customer provides. Following the developments related to powdered forms of graphene, Graphe-nea has introduced various products featuring graphene oxide, either in a watersuspension or as dry powder. These products are in demand, as powderedgraphene is easily integrated into polymer matrices for enhancement of thermaland mechanical properties. It will most likely be powdered graphene (includingGO) that keeps Graphenea sailing in the medium term, as the construction,automotive, aerospace, and sporting industries pick up on the hi-tech func-tionality that graphene can offer. At the end of 2013, Graphenea announced amillion-euro investment in the company by Repsol [14], a leading oil and gasbusiness. Graphenea has managed to increase its turnover by 50% each year sinceestablishment. To keep on top of cutting edge research, Graphenea collaborateswith leading graphene scientists, which gives the company an excellent knowledge

EUR/cm2180 7 Graphene Market Review base as well as a good reputation. Graphenea is also the leading graphene supplier in the billion-euro Graphene Flagship consortium, a 10-year program of the European Commission designed to help bring graphene to commercial products. Becoming part of the Flagship provided access to hundreds of European research groups for Graphenea, further strengthening the company’s position on the European market. The business strategy for Graphenea is to follow the evolution of graphene as it moves from basic research to applications, and in that regard it is natural that the user base has been gradually shifting from research labs to other businesses in end-user industries. In 2014, the company had more than USD1 million in revenue, and a positive cash flow. The premise is that the price of high quality sheet graphene will soon drop below the price of main competing materials (see Figure 7.1), which will give a boost to sales of pure graphene. Graphene Laboratories has taken a different approach by expanding their product catalog to include other ultrathin materials such as boron nitride, MoS2, WS2, as well as materials and devices supplementary to graphene production, like coronene for better CVD growth or the SpeedMixer for mixing nanomaterials and creating paste. In February 2012, Graphene Labs signed an agreement with Lomiko Metals Inc, a Canada-based mining company, to build entire graphene application chains. The agreement obliged Lomiko to provide high-quality natural graphite and to raise USD2 million over a period of 18 months. In effect, the companies worked together to develop graphene technology and to perfect the use of natural graphite in graphene production. This type of joint venture between graphene specialists and graphite mines was also seen elsewhere, for example in the partnerships of Grafoid and Focus 1000 100 10 1 0.1 0.01 2010 2012 2014 2016 2018 2020 2022 Figure 7.1 The price of graphene.

7.2 Graphene Market: Past and Present 181Graphite, Mason Graphite and Group Nanoxplore, and American Graphite andCheaptubes. Although it seemed at the time that many more graphite minerswould go this way, others seem to have shied away from major graphene ventures.For example, Northern Graphite, another large mining company based in Canada,was happy to sign a research agreement with Turkish Grafen Chemical Industries,with the only stake being patent rights. Overall, the graphite mining industryhas been an important investor in the graphene market for the past 5 years,mainly through joint ventures with and takeovers of graphene specialist firms.In this period, graphite mines were one of the very few options for investing in“graphene stock,” albeit rather indirectly. By now several pure graphene compa-nies have gone public, making the graphene market more interesting to smalltimeinvestors. Graphene Labs have used their presence on the ultrathin materials market asa springboard to hop into 3D printing, a fast-emerging tech that relies on novelmaterials where graphene would feature as an additive to the printing mass, eitheras conductive material or as a mechanical strengthener. The founders of GrapheneLabs spun out Graphene 3D Lab Inc., and in 2014 demonstrated a 3D printedgraphene-based battery, that can be printed in any shape and size and can there-fore be incorporated into custom designs. At the end of that year the companyacquired Boots Industries, a Canadian 3D printer manufacturer. Following up, thecompany has launched a conductive graphene-enhanced filament, fully compati-ble with commercial 3D printers. A conductive 3D printing filament can be used,for example, to print EMI and RF shielding, wearable electronics, keyboards, MIDIcontrollers, and so on. In June 2015, Graphene 3D Labs announced the doubling of their productioncapacity for functional filaments, triggered by growing demand on their graphenefilament. The company also announced the introduction of at least three newmaterials in 2015. Clearly, this is a successful line of business with huge growthpotential, and Graphene Labs has been playing it right from the onset. LikeGraphenea, Graphene Laboratories announced revenue in excess of USD1million in 2014, and a positive cash flow coupled with a lack of debt. Perhaps the biggest producer (west of China) of GNPs, tiny flakes of multilayergraphene, is Dayton based Angstron Materials. Claiming approximately 300 tpaproduction capacity and several dozen patents on GNPs, this is also the companythat holds the first patent on GNPs. “GNPs” is a generic name given to a multitudeof graphene nanomaterials of arbitrary shape and lateral dimensions on the orderof 100 nm, with a thickness of several nanometers. This kind of graphene has beengaining popularity in recent years, due to the ease of production and the chemicalreactiveness at the edges, which makes it suited for integration into other materialmatrices. GNPs are heavily being used in thermal, mechanical, barrier, and energyapplications. Most producers have been moving up the value chain, offering additivemasterbatches, inks and coatings, and partnering with other industries to pushgraphene into products. For example, XG Sciences have delved deep into energystorage applications, partnering with Samsung Ventures Investment Corporation

182 7 Graphene Market Review and having already tested graphene-based batteries and electrodes. Vorbeck has gone even further, offering ready-made RFID tags, enhanced rubber, and flexible battery straps made of graphene. Similarly, in parallel to sales of various forms of GNPs, Italian Directa Plus partnered with Vittoria SpA to bring to the market the world’s first graphene-enhanced bicycle tires, with the next step going in the direction of race car pneumatics. The sports requisite industry in general provides a natural habitat for graphene products, because sports equipment needs to be lightweight, durable, and often flexible, all of which can be improved with the addition of a little bit of graphene to the original polymer matrix. In that respect, the first commercial product containing graphene was a tennis racquet from HEAD, sported already in 2013 by Novak Djokovic´ and Maria Sharapova. Although several companies claim hundreds of tpa capacity and a few claim kilotonnes, it is estimated that no more than a kilotonne per year of combined production of GNPs and graphene oxide will be required in the next few years [15], indicating current production overcapacity. Graphene manufacturers are hence now focusing on batch-to-batch consistency and cutting production costs [15]. Overproduction is best illustrated by the resultant effect on stock of some publicly traded graphene businesses, shown in Figure 7.2. It is likely that the graphene market has just passed the peak of hype and is entering the trough of disillusionment of the Gartner hype cycle (Dr Khasha Ghaf- ferzadeh, IDTechEx presentations). However, some have managed to have short- term stock market success, as exemplified by the UK-based Haydale. The market capitalization for all pure-play graphene companies is well below USD50 million, setting them in the “nano cap” category. It seems that as of the end of 2015, the graphene market is at a tipping point between fundamental research- driven and application-driven (Figure 7.3). The graphene market and its key players are best summarized in Figure 7.4 where development of graphene is tracked from source to end product through the value chain. Although only key businesses are shown on the chart, the Graphene Tracker business directory contains nearly 90 entries, and there are an estimated 44 graphene producers on the market [15]. Most fit in the “Graphene Based Com- pounds” category, and at least half of those are moving up the value chain, with their own end products in the pipeline. Due to the oversupply and rapid growth of the number of companies in this sector, moving up the value chain seems like the only good strategy for long-term survival. Also, a common strategy for graphene businesses has been to find partners further up the value chain, in a joint effort to propel graphene materials to higher value. Examples of this type of partner- ship abound, for instance in Angstron’s ventures with Stryke Industries (military), Haydale’s with planarTECH (graphene manufacturing equipment), Graphene 3D Lab’s with Polymaker (3D printing materials), XG Sciences’ with Samsung (elec- tronics), Perpetuus’ with OXIS (battery electrodes), and Graphene NanoChem’s with Emery (oleochemicals).

7.2 Graphene Market: Past and Present 1832.5 GGG: Graphene 3D Labs2. AGM: Applied graphene materials PLC400300200100 80 GRPH: Graphene nanochem 6040200 Jan 2015 Apr 2015 Jul 2015 Oct 2014Figure 7.2 Stock value of some publicly traded graphene companies (Google Finance.)200 HAYD: Haydale150100500 Jan 2015 Apr 2015 Jul 2015 Oct 2014Figure 7.3 Haydale stock price.

184 7 Graphene Market ReviewAixtron Graphenea ResearchCVD equipment Graphene squareplanarTECH Graphene supermarket BGT materials CVD machines CVD graphene End products Graphite ore Graphene BASF based IBMNorthern graphite compounds NokiaZenyatta ventures SamsungFocus graphite Vorbeck materials SandiskLomiko XG sciences MicronMason graphite Angstron HEAD Applied graphene materials AMO GmbH Haydale ... GRAnPH Nanotech CrayoNano Grafen Graphenano ...Figure 7.4 Graphene industry diagram, highlighting key businesses. ( Market InitiativesIn an unprecedented effort to turn graphene expertise into market value, largefunding and coordination initiatives are being assembled worldwide. By far, thelargest of these is the pan-European Graphene Flagship, which aims to coordi-nate research and commercialization efforts of 142 entities in 23 countries, as ofMarch 2015. The Flagship is both a funding and coordination action, with theEU contribution amounting to 1 billion EUR over 10 years (2012–2022) and asmuch expected to be contributed by national governments and businesses. TheGraphene Flagship is run from Chalmers University of Technology in Sweden,directed by Prof. Jari Kinaret. The director is supported by a management team,and the entire undertaking is divided into 16 work packages, each coordinated byits own leader. The Flagship is in close contact with the European Commission,the executive body of the European Union, and the two have signed a partner-ship agreement that obliges them to support each other. Regional, national, andtransnational graphene-related projects can apply to become part of the GrapheneFlagship, which allows for absorbing the entire graphene community of Europeinto one body. The Flagship is one of two Future and Emerging Technology (FET) initiatives(the other being the Human Brain Project), the success of which will decidewhether the EU will embark on other such long-term projects in the future.

7.4 Market and Application Projections 185 On the North-American side, the effort is not as co-ordinated as the Flag-ship consortium, but it rather reflects a traditional American approach. Thecommercialization effort is being driven by the Graphene Stakeholders Asso-ciation (GSA), a non-profit organization entailing more than 40 members. Theorganization pushes forward standardization, characterization analysis, andenvironmental health and safety issues, in an effort to legislate graphene. Throughon-going collaboration with key stakeholders, the GSA’s “Center of Excellence”initiative seeks to accelerate the development of standards and facilitate thecommercialization process of graphene and other 2D materials. In the far East,the China Innovation Alliance of the Graphene Industry (CGIA) is the Chinesecounterpart of the GSA. Also a non-profit, this alliance is a national organizationsupported by the government of China. The CGIA is a consortium of severaldozen members, including industrial enterprises, academic institutions, andresearch organizations. The South Korean government announced a USD40million 6-year investment plan for graphene companies in 2013. Finally, Malaysiaannounced the Malaysian “National Graphene Action Plan” in July 2014. A “whitepaper” was published at the time, but little news followed since. More local intensified efforts to commercialize graphene are evident in theUnited Kingdom, where government support to bridging research and industryhas materialized through the National Graphene Institute (NGI) and the Cam-bridge Graphene Centre. The Cambridge Centre was initially funded by a £12million government grant, coupled with an additional £13 million from industrialpartners, including Nokia, Plastic Logic, Dyson, Philips, and BAE Systems. TheNGI in Manchester is a research and incubator center housed in a £61 millionbuilding of over 7600 m2. The first company to move into the NGI was BGTMaterials, formerly known as Bluestone Global Tech.7.4Market and Application ProjectionsProjections of the value of the graphene market vary almost as much as theforms of graphene. Most recent reports from IDTechEx predict a market valueof about USD200 million in 2026, and include 17 specific application sectors.The largest application sectors, according to this report, will be energy storageand supercapacitors, composites, and research, with inks and coatings followingclosely. In contrast, a report from Chisult Insight [16] predicts a USD300 million mar-ket size for 2020 in China only. Allied Market Research predicts USD150 millionglobally for the same year; Markets and Markets predicts USD1.5 billion in 2020from electronics only [17], while BCC Research predicts about USD1 billion intotal [18]. Experienced market analysts, including IDTechEx and Lux Research,are increasingly cautious about their predictions, with Lux Research’s publishingan extremely negative outlook in June 2015 [19].

186 7 Graphene Market Review Graphene-based technology market readinessTime to market (years) DNA sequencing 5 10+ Flexible OLED Biosensors Fast electronics Photodetectors Thermal composites Polymer reinforcer for aerospace and automotive Batteries and supercapacitors RFID tags and antennae Functional fluids, ex. drillingNow Sports equipment Principles and concepts Proof of concept and prototyping Commercialization technology readiness levelFigure 7.5 Time to market and technology readiness level of some applications (author’sown predictions). Applications that demand fast electronic performance, which were first thoughtto be the most promising ones for graphene, will surprisingly grow very littlein the short term, due to the fierce competition that graphene faces there fromestablished materials such as silicon and ITO. In general, applications that requirehigh-quality graphene are expected to take longer to reach the market, as evi-denced by recent application predictions (see Figure 7.5 and Refs [20] and [21]).IDTechEx predicts that even in the year 2026, more than 75% of the market sharewill belong to composites, energy storage (supercapacitors), and inks and coatings,whereas sensors, logic, transparent conductive films and research will altogethertake up less than 15% of the market share [21].).7.5ConclusionAlthough still primarily research-driven, the graphene market is naturally beingpopulated by graphene for industrial use. The effort to commercialization isespecially being helped by large co-ordinated initiatives on all sides of the globe.Drawing from the experience with carbon nanotubes and other nanotechnolo-gies, the community is pursuing material substitution and additives in parallel todiscovering disruptive, groundbreaking applications. The graphene productionmarket is following controlled expansion and is now worth several million USD,with projected rapid expansion in the coming years due to the material’s diverserange of applications.

References 187References1. Novoselov, K.S., Geim, A.K., Morozov, 9. (accessed 13 January S.V., Jiang, D., Zhang, Y., Dubonos, 2015). S.V., Grigorieva, I.V., and Firsov, A.A. (2004) Electric field effect in atomically 10. (accessed 13 thin carbon films. Science, 306 (5696), January 2015). 666 – 669. 11. (accessed 13 Jan-2. Somani, P.R., Somani, S.P., and Umeno, uary 2015). M. (2006) Planer nano-graphenes from camphor by CVD. Chem. Phys. Lett., 12. (accessed 13 430 (1), 56–59. January 2015).3. Li, X., Cai, W., An, J., Kim, S., Nah, 13. (accessed 13 Jan- J., Yang, D., Piner, R., Velamakanni, uary 2015). A., Jung, I., Tutuc, E., Banerjee, S.K., Colombo, L., and Ruoff, R.S. (2009) 14. Repsol Press Release, http://www.repsol Large-area synthesis of high-quality and .com/es_en/corporacion/prensa/notas- uniform graphene films on copper foils. de-prensa/ultimasnotas/05122013- Science, 324 (5932), 1312–1314. repsol-cdti-invierten-un-millon-euros- empresa-graphenea.aspx (accessed 134. Bae, S., Kim, H., Lee, Y., Xu, X., Park, January 2015). J.-S., Zheng, Y., Balakrishnan, J., Lei, T., Kim, H.R., Song, Y.I., Kim, Y., Kim, K.S., 15. Zurutuza, A. and Marinelli, C. (2014) Özyilmaz, B., Ahn, J.H., Hong, B.H., and Nat. Nanotechnol., 9, 730–734. Iijima, S. (2010) Roll-to-roll production of 30-inch graphene films for transpar- 16. ent electrodes. Nat. Nanotechnol., 5 (8), 402474-research-on-global-china- 574 – 578. graphene-industry-2015.html (accessed 29 February 2016).5. Kobayashi, T., Bando, M., Kimura, N., Shimizu, K., Kadono, K., Umezu, 17. http://www.researchandmarkets N., Miyahara, K., Hayazaki, S., Nagai, .com/publication/mystl29/graphene_ S., Mizuguchi, Y., Murakami, Y., and electronics_market_by_materials_dev Hobara, D. (2013) Production of a (accessed 29 February 2016). 100-m-long high-quality graphene trans- parent conductive film by roll-to-roll 18. chemical vapor deposition and trans- research/advanced-materials/graphene- fer process. Appl. Phys. Lett., 102 (2), technologiesapplications-markets- 023112. avm075c.html (accessed 29 February 2016).6. price (accessed 13 January 2015). 19. 2015/07/case-closed-graphene-is-the-7. Du, W., Jiang, X., and Zhu, L. (2013) next-carbon-nanotube/ (accessed 29 From graphite to graphene: direct February 2016). liquid-phase exfoliation of graphite to produce single- and few-layered pristine 20. Ferrari, A.C. et al. (2015) Science and graphene. J. Mater. Chem. A, 1, 10592. technology roadmap for graphene, related two-dimensional crystals,8. Wolf, E.L. (2014) Chapter 2: Practi- and hybrid systems. Nanoscale, 7, cal productions of graphene, supply 4598 – 4810. and cost. Applications of Graphene, SpringerBriefs in Materials. 21. reports/graphene-2d-materials- and-carbon-nanotubes-markets- technologies-and-opportunities-2015- 2025-000440.asp (accessed 02 May 2016).

1898Financing Graphene VenturesStephen R. Waite∗Graphene is dead. Long live graphene. -Andre GeimThe discovery of graphene in 2004 has generated a good deal of interest amonginvestors, entrepreneurs, and established companies alike. Graphene is viewedtoday as one of the world’s most remarkable materials and for some very goodreasons. Graphene possesses unique properties (as discussed elsewhere in thisbook) that make it a promising candidate for use in a range of applications in keyeconomic segments, including aerospace, consumer products, energy, electron-ics, communications, health care, and transportation. The astonishing propertiesof graphene have captivated the interest of many investors and entrepreneurs whosee the potential to add significant value in commercial applications using thematerial. Over the past decade, numerous companies have been started to capitalize onthe opportunities associated with graphene. While these companies vary in theirpurpose and objective, they all share a common trait in needing to raise capital tofund operations, research, and development. In this chapter, we explore the issuesrelated to financing companies innovating with graphene.1) This chapter comprises five major sections plus an Appendix. The first sectionprovides a brief overview of the graphene company landscape. The second sectionexplores what we call “The Art of Raising Capital,” in both theory and practice.Raising capital is paramount for all graphene ventures and this section provides aprimer on funding channels for graphene ventures. The third section examines theshifting financial landscape and discusses the implications for seed and early stagegraphene ventures. This section includes observations gleaned from interviewswith various executives in regard to raising capital for their graphene venturesin the current financing environment. The fourth section discusses the financing* Co-founder and Co-Executive Director of the Graphene Stakeholders Association, author of Quan- tum Investing (Thomson/Texere).1) It should be noted and kept in mind that while the focus of this chapter is on financing graphene ventures, many of the observations apply to seed and early stage companies innovating with other 2D and 3D advanced nanomaterials.Graphene Technology: From Laboratory to Fabrication,First Edition. Edited by Soroush Nazarpour and Stephen R. Waite.© 2016 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2016 by Wiley-VCH Verlag GmbH & Co. KGaA.

8.2 The Art of Raising Capital 197 Another promising channel of funding for graphene ventures is strategicfunding from larger, more established corporations. Strategic partnershipswith established corporations can provide not only financial capital, but alsoaccess to customers as well as helping build an ecosystem to support futuredevelopment and innovation. Customer development is paramount for success inthe marketplace. A well thought out and properly executed strategic partnershipwith an established corporation can yield great benefits to graphene ventures andset the stage for a liquidity event down the road, through either an acquisitionor IPO. Early stage graphene ventures are encouraged to seek out win–winstrategic partnerships with established corporations globally as a means offunding operations and developing customers. The other prominent sources of funding for graphene ventures are reverse merg-ers and IPOs. Reverse mergers are financing vehicles that supply capital and liq-uidity for existing and prospective investors and are executed in lieu of an IPO.With a reverse merger, a company backs into a shell that is already publicly tradedon a stock exchange case at the OTC market in the United States Some shells havecash on the balance sheet. The cash can be used to help execute a reverse merger.Some investors have a dim view of companies that engage in reverse mergers.There has been quite a bit of financial shenanigans in the form of smoke and mir-rors that have gone on with reverse mergers over the years that has tarnished theirreputation with investors. That said, a well-executed and managed reverse mergercan be a boon to an early stage company developing innovative products withgraphene and other advanced nanomaterials. IPOs are another source of capital for graphene ventures. IPOs are viewed asattractive by many emerging technology companies, primarily because such offer-ings can and often do provide access to a large amount of capital while creatingliquidity for existing and prospective investors. IPOs have a long and illustrioushistory with emerging technology companies in the United States and elsewhere.Some of the biggest and most successful technology companies, including Apple,Intel, Microsoft, ARM Holdings, Google, Amgen, and many others completedsuccessful IPOs and continue trading on public securities exchanges today. IPOs has traditionally been viewed as an ideal long-term source of capital giventhat successful IPOs can take care of present capital requirements and provideaccess to future capital through secondary and follow on offerings. It is frequentlythe case that an emerging technology will tap the public markets for additionalfunding several times early on in its evolution. Financial markets around the worldhave established stock exchanges where public securities can trade following anIPO. Major banks have extensive resources to help companies execute an IPO andoffer research and other banking support following the completion of a deal. Adiverse group of investors – retail and institutional – can and do frequently par-ticipate in an IPO. There have been a handful of graphene companies that havecompleted successful IPOs/reverse mergers in recent years, as discussed further. As can be gleaned from the foregoing discussion, there are many differentsources of funding available for graphene and other advanced nanomaterialsventures. The art of financing such ventures involves being opportunistic and

198 8 Financing Graphene Ventures resourceful with respect to all the various types of funding sources highlighted earlier. Funding opportunities vary by region and by country. Entrepreneurs seeking to seed or fund early stage graphene ventures are encouraged to explore all channels and sources of available capital in their respective region and country. Resourceful entrepreneurs often have alternative sources of capital lined up before needing it and are in constant communication with potential investors giving them updates on the state of the business. In terms of type of financing, equity capital in the form of common or preferred stock is the preferred choice of financing for an emerging technology company developing graphene-enabled products. Debt financing can be employed, but this type of funding is discouraged for graphene ventures due to the unpredictable nature of future company cash flows. Debt financings are far better suited toward later stage companies in more stable industry segments that are generating free cash flow. That said, it is quite common to see emerging technology companies using convertible debt – that is, debt that converts into equity at a later stage of development depending on meeting certain criteria as specific in the funding doc- uments – to fund early stage development. A relatively new and quite attractive option for graphene venture is convertible equity. This type of funding is suitable to the risk and development profile of early stage graphene ventures and makes for an attractive alternative source of funding relative to convertible debt. It is important for a company to have a capital structure that is properly aligned and supportive of the business. The exhibit in the following text illustrates a hypothetical evolution of funding associated with an early stage graphene venture in the current financial environ- ment. Clearly, financial environments will shift over time and it is important for companies to remain opportunistic and take advantage of as many funding chan- nels as possible. The exhibit shows a relationship between the type of funding and the product development cycle – a cycle that typically extends out over a decade or more for an advanced materials company. As the exhibit shows, graphene ventures are likely to be predominately funded by founders, F&F early on, along with government grant monies. Some companies may attempt an IPO and/or reverse merger for funding if the channel is avail- able. The funding at this stage of evolution fosters important activities related to customer and product development. Success with product and customer devel- opment will embolden executives to seek additional funding from deeper pockets (e.g., HNW and Angel investors). Bringing in bigger sums of capital facilitates further product development and the expansion of important operational activ- ities such as sales and marketing. At this time, executives may be successful in securing strategic partnerships that bring in additional capital and assist with cus- tomer development. Product development may be accelerated through additional government/research grants. Validation from customers will further embolden executives to seek additional funding, typically from institutional investors that can write big checks. At this juncture in the company’s evolution, there is clear market demand for the venture’s graphene-enabled technology or application and sales and marketing activity ramp up to support commercialization. Successful

8.3 Shifting Financial Landscape for Graphene Ventures 199product commercialization is likely to lead executives to consider raising morecapital through a public offering or a reverse merger. There may be an oppor-tunity for a liquidity event through an acquisition by a larger, more establishedcompany. Founders and executives will often seek liquidity for shareholders as away to enhance the company’s attractiveness to investors and provide a means toreap the rewards of their hard work. The hypothetical financing cycle posits a tight relationship between fundingand customer and product development. It is often the case that the relationshipbreaks down along the way for various reasons that frequently lie outside the con-trol of the company. It is tempting for founders and executives to rush to do anIPO or reverse merger before customers and products have been developed. Suchbehavior can backfire in the case of advanced materials companies given the rel-atively long product development lead times. Investors that participate in an IPOor reverse merger that is completed early on while a company is in the early phaseof research and development may lose patience as executives struggle to createvalue. Chief among the reasons for a breakdown in the relationship are shifts inthe financial environment and investment climate. The next section of this chapterexplores the shifting financial landscape and discusses the implications for com-panies innovating with graphene. Many of the companies mentioned in the first section of this chapter are engagedin collaborations with larger, more established companies. Strategic partnershipscan be highly effective in accomplishing two important tasks associated with earlystage graphene ventures: raising capital and acquiring customers. There is growinginterest in graphene for commercial use by companies such as Samsung, LockheedMartin, IBM, and many others. Such collaborations can, and often do, providefinancing and access to customers, which is critical to the success of an early stagetechnology venture. Such financings will be highlighted and discussed furtherbelow.8.3Shifting Financial Landscape for Graphene VenturesThe most prevalent sources of funding for seed or early stage graphene venturesover the past decade have been F&F, HNW, Angel, Government, and Strategic.Notably absent from the funding of graphene ventures is venture capital and IPOsand reverse mergers. To be sure, as noted earlier a few graphene start-ups havebeen able to complete IPOs during the past couple of years, including HaydalePLC, Applied Graphene Materials, and Graphene 3D Labs (see Table “GrapheneCompany IPOs”). As the exhibit shows, despite the relatively large amount of hype surroundinggraphene, there has not been a large demand for graphene-related IPOs to date.There are many reasons for this. For starters, venture capital funds have beenincreasingly migrating toward social media/software deals and away from fundingadvanced nanomaterials companies. We will have more to say about this trend.

200 8 Financing Graphene VenturesGraphene company IPOsCompany Offering date Amount raised (USD) ExchangeApplied Graphene Materials Nov 2013 17 mn AIMCientifica Oct 2013 389k AIMHaydale PLC Apr 2014 11 mn AIMGraphene 3D Labsa) Aug 2014 1.5 mn TSX VEa) Reverse merger transaction that included private placement of USD6 million shares at USD0.25.Source: Graphene-info. The financing environment for early stage graphene ventures has shifted inrecent years. In the period leading up to the announcement of the Nobel Prizethere was appreciable interest by investors that were piqued by the remark-able properties of graphene and its potential in a wide range of industries.Interest in funding graphene start-ups escalated in 2010, following Geim andNovoselov’s Nobel Prize award for the discovery of the 2D nanomaterial. Thehype surrounding graphene intensified following the Nobel Prize award. In2011–2012, graphene came to be known as “the wonder material” and a posterchild for the next disruptive technology, a title previously held by its cousin,carbon nanotubes (CNTs). Interest in graphene ventures came from both retailand institutional investors. Retail investors were captivated by media storiesportraying graphene as “a wonder material.” Institutional investors saw thepotential for game-changing applications with graphene. During the first couple of years following Geim and Novoselov’s Nobel Prizeaward the funding environment for seed and early stage graphene companies wasfairly robust. Numerous graphene companies were seeded during this time. Fund-ing for graphene ventures could be completed within a year with nothing morethan a CEO, a Scientific Founder, and a business plan. Investor interest was highfrom both institutional and retail sources of capital. However, the funding envi-ronment for graphene ventures began to shift after Bayer MaterialScience (BMS)announced its exit from the CNT business in spring of 2013.2) The BMS news cast a cloud over the commercialization prospects of graphene.Institutional investors began to ask some tough questions due to the lack of com-mercialization success with CNTs. After all, CNTs were discovered in the early1990s. After two decades, it would seem reasonable to see evidence of adoption ofthe advanced nanomaterials. Furthermore, if a company like BMS cannot succeedin the nanocarbon business, why should anyone think a start-up could make itin graphene? These relevant questions require an astute response from the execu-tives associated with fledgling graphene ventures (see “Lessons Learned” section).2) See “Bayer MaterialScience exits carbon nanotube business,” at link: http://www.plasticstoday .com/articles/bayer-materialscience-exits-carbon-nanotube-business0508201301.

8.3 Shifting Financial Landscape for Graphene Ventures 201 The wide range of potential applications for graphene along with the factthat there has been relatively little experience innovating with 2D materialsposes unique investment challenges. These challenges are all the more acutein a financing environment where investor interest in advanced materials andnanotechnology is overshadowed by the attraction of social media and othersoftware-related investment opportunities today. This subject is discussed fur-ther in the next section of this chapter. Lessons Learned BMS’s exit from the CNT business in May 2013, while casting a cloud over graphene’s commercialization potential and leading to a shift in the financing environment for graphene ventures, has been instructive. There are some valuable lessons associated with BMS’s experience that graphene ventures can learn. Over the past two decades, CNT production, with a few exceptions, followed a traditional structure of the materials value chain. There was an implicit assumption that one could simply produce the materials evermore cheaply and others would add value to them along the way. This assumption turned out to be costly. Commercializing advanced nanomaterials requires new tools and processes that enable dispersion for use with other conventional industrial polymers. Producers of nanomaterials are required to work closely with customers and strategic partners to create valued-added materials with desirable properties at a price/performance ratio that is attractive for the customer. Advanced nanomaterials such as CNTs are inelastic with respect to their demand. The demand for CNTs is tied to their ability to be used, not to their price. No one wants nanocarbon per se. What companies seek are better materials – materials improved by the addition of the most active form of carbon available. In an interview with Ross Kozarsky of Lux Research, Peter Kruger, former head of BMS, admitted that the company should have focused more on application development and working with customers than with raw CNT production. This is precisely the path Nanosys CEO Jason Hartlove and his team took that led to successful market penetration with its quantum dot nanomaterials. Nanosys was long recognized for its deep portfolio of nanotechnology intel- lectual property. After taking the helm at Nanosys, Hartlove and his team eval- uated all of the IP and saw a near-term path to commercialization with the company’s proprietary quantum dot nanomaterials. Hartlove’s strategy was to create a high valued-added product with its proprietary quantum dots. Within a few of years of Hartlove taking over the helm, Nanosys was able to bring ™to market a next generation quantum dot display technology called QDEF .

202 8 Financing Graphene Ventures The QDEF technology produces vibrant displays that are being used in various consumer electronics products today. The Nanosys model for commercializing advanced nanomaterials stands in contrast to the failed approach by BMS with CNTs. In sum, BMS’ failure with CNTs has been an instructive and a valuable lesson in how not to commer- cialize advanced nanomaterials. Given the lessons being learned, it would be misguidance to extrapolate the failed fortunes of CNTs to graphene, as some investors were prone to do after BMS exited the CNT business. It is the failures of the past that lead to the successes of tomorrow for those who are paying atten- tion. Entrepreneurs innovating with graphene and the investors backing them are advised to take note of the important lessons learned in commercializing advanced nanomaterials. Start-ups intent on commercializing graphene and other advanced 2D and 3D nanomaterials face many hurdles in the capital markets today. Chief among the hurdles is the relatively long length of time and large amount of capital required to meaningfully create a new market or penetrate an existing market (see the “Nantero Case Study” in the Appendix to better appreciate the time, capital, and effort to bring an innovative nanomaterials product to market). By contrast, early stage social media companies can leverage the hundreds of millions of computer devices connected to the Internet and ramp up in a matter of months or a couple of years. Witness the ascent social media companies such as Uber that are able raise hun- dreds of millions and billions of venture capital today. According to CB Insights, Uber has raised USD5.5 billion since the company’s founding in March 2009. The most recent round of venture financing commanded a USD50 billion PE valua- tion.3) Eye-popping numbers such as these would make any venture investor stand up and take notice. By contrast, there are advanced nanomaterials companies that have been plugging away for over a decade and have raised a fraction of what Uber has raised in terms of capital and at valuations that are a small percentage of what Uber is commanding. To be sure, not all social media and software ventures are akin to Uber. Other so- called early stage “on demand” companies in the United States have raised USD5.9 billion, slightly more than Uber alone. Nonetheless, billions of dollars of early stage venture capital is flowing into the social media segment today due to the favorable business climate and dynamics for such enterprises. This poses a major challenge for graphene start-ups today. There is a large and growing number of social media start-ups vying for capital today that have the potential to leverage the hundreds of millions of electronic devices in the market today and ramp up very quickly. 3) See CB Insights blog post at link demand-economy/.

8.4 The Graphene Financing Road Ahead 2038.4The Graphene Financing Road AheadDue to the success of social media companies such as Uber, Facebook, LinkedIn,Twitter, and others, social media start-ups are commanding the majority of atten-tion of venture investors in Silicon Valley and elsewhere. Entrepreneurs seekingto commercialize graphene or some other advanced nanomaterial must recog-nize and take into consideration the competition for capital coming from socialmedia ventures – and more generally, software ventures – today. A typical VCor Angel investor is far more interested in funding a social media business thathas Uber-like potential than an advanced nanomaterials start-up. VC Steve Blanknoted this attraction in an illuminating blog post titled, “Why Facebook is KillingSilicon Valley.” In his blog post, Blank wrote: “If investors have a choice of investing in a blockbuster cancer drug that will pay them nothing for fifteen years or a social media application that can go big in a few years, which do you think they’re going to pick? If you’re a VC firm, you’re phasing out your life science division. As investors funding clean tech watch the Chinese dump cheap solar cells in the U.S. and put U.S. startups out of business, do you think they’re going to continue to fund solar? And as Clean Tech VC’s have painfully learned, trying to scale Clean Tech past demonstration plants to industrial scale takes capital and time past the resources of venture capital. A new car company? It takes at least a decade and needs at least a billion dollars. Compared to IOS/Android apps, all that other stuff is hard and the returns take forever. Instead, the investor money is moving to social media. Because of the size of the market and the nature of the applications, the returns are quick – and huge. New VC’s, focused on both the early and late stage of social media have transformed the VC landscape.”4) Blank’s assessment of the investing landscape in Silicon Valley is sobering. Soft-ware has been eating the world and venture capital deal flow in Silicon Valley andelsewhere has been increasingly concentrated in social media for all the reasonsnoted in his blog post. That said, it remains to be seen if these trends will con-tinue in the years ahead. Investment fads come and go, as all experienced investorsknow. Social media is hot now, but may not always be hot. There are already signsthat the ebullience in social media investors is leading to over-inflated valuationsand resource constraints that could turn the tide of increasing concentration inthe software segment. As Blank notes, the unwritten manifesto for Silicon Valley VCs for decades hasbeen: “We choose to invest in ideas, not because they are easy, but because they arehard, because that goal will serve to organize and measure the best of our energiesand skills, because that challenge is one that we are willing to accept, one we are4) See Steve Blank blog post, “Why Facebook is Killing Silicon Valley,” at link 2012/05/21/why-facebook-is-killing-silicon-valley/.

204 8 Financing Graphene Ventures unwilling to postpone, and one which we intend to win.” Graphene has tremen- dous commercial potential in the marketplace. There is hope that 1-day VCs in Silicon Valley and elsewhere see attractive investment opportunities once again in funding start-ups and emerging companies seeking to innovate with advanced nanomaterials. Entrepreneurs seeking to start-up graphene ventures need to be especially cre- ative and resourceful in how they go about funding and operating their companies in an environment where software is eating the world. It will be incumbent upon all emerging graphene companies to tap into as many investment channels as pos- sible and formulate strategies that foster accelerated product development times and strategic partnerships that promote rapid customer development as well as access to financial capital. There are signs today that governments around the world are increasingly com- mitted to fostering innovation associated with advanced nanomaterials such as graphene. The Graphene Flagship in Europe is a prime example. Government funding initiatives announced in recent years have been geared toward basic scien- tific research – research that is essential and important, as basic scientific research has the potential to spill over into the development of commercial applications. There is heightened interest among various governments around the world to fund the development of graphene-enabled applications for various defense and mil- itary applications. Entrepreneurs tapping into government funding sources for such applications are always encouraged to seek to leverage any success in that segment into the enterprise and consumer segments. After all, there is no sub- stitute for customer financing. Successful and sustainable businesses, graphene and otherwise, are funded by customers that benefit from using the product or application. To the extent government funding is available, we encourage graphene start-up companies and emerging graphene ventures to continue to monitor the invest- ment landscape carefully. There have been some constructive and noteworthy developments in US capital markets that seek to foster greater access to capital for emerging technology companies. In March 2015, the Securities and Exchange Commission (SEC) announced new rules to provide investors with more invest- ment choices. The new rules update and expand Regulation A, an existing exemption from registration for smaller issuers of securities. The rules are mandated by Title IV of the Jumpstart Our Business Startups (JOBS) Act. The updated exemption will enable smaller companies to offer and sell up to USD50 million of securities in a 12-month period, subject to eligibility, disclosure, and reporting requirements. SEC Chair Mary Jo White noted that it is important for the Commission to con- tinue to look for ways to facilitate capital raising by smaller companies.5) With Title IV of the 2012 JOBS Act, the fastest growing private companies in America can conduct IPOs and raise up to USD50 million from any investor. 5) See press release, “SEC Adopts Rules to Facilitate Smaller Companies’ Access to Capital,” at link,

8.4 The Graphene Financing Road Ahead 205Historically, investing in start-ups and small businesses in the United States wasreserved for just accredited investors, or just the wealthiest 2% of America. Now,after 3 years of anticipation, investing in pre-IPO companies will be opened up tothe other 98%. The new SEC legislation is a breath of fresh air and a welcome devel-opment for emerging graphene ventures. The new SEC legislation could providefurther impetus to the crowdfunding platforms that have emerged in recent yearsthat are friendly to start-up companies. It remains to be seen how entrepreneurstake advantage of the new rules and how receptive investors are in funding earlystage advanced nanomaterials companies in US capital markets. Meanwhile, emerging graphene companies in the United Kingdom and Europewill continue to raise funds through small capitalization listings on the AIM. TheAIM has been friendly to entrepreneurs raising capital for graphene-enabled inno-vations and is likely to remain so in the future. Of course, graphene entrepreneursalways need to be mindful that the window for small cap IPOs opens and shutsdepending on gyrations in the financial markets and the appetite for such dealsamong investors.SummaryThe discovery of graphene in 2004 opened the door to a wide range of commer-cialization opportunities. The commercialization of graphene requires funding.The funding of graphene ventures is more of an art than a science. As noted ear-lier, there are numerous sources of capital available to seed or fund an early stagegraphene venture. A resourceful executive will be as opportunistic as possibleand try to tap into as many channels of funding as possible to produce and/orcommercialize graphene-enabled products and applications. Government fund-ing remains an attractive source of capital for fledgling graphene ventures, particu-larly in the United Kingdom, Europe, and Asia. The European Union has ploughedover USD1 billion into start-up companies during the past year in an attempt toaccelerate the high-tech economy and stimulate economic dynamism. The EUprogram is targeted to address the chronic complaint from entrepreneurs thatfunding is far more difficult to raise in Europe than elsewhere. In North America,HNW and Angel funding are likely to continue to be prominent sources of cap-ital for early stage graphene ventures. Globally, crowdfunding is a relatively newchannel that could play a bigger role in funding graphene ventures in the future. The past decade has seen a migration of venture capital funding away fromadvanced nanomaterials companies and toward social media/software deals. Itis unclear if this migration will continue. Nevertheless, the relatively long devel-opment cycles of advanced nanomaterials-enabled products and applicationspresent funding challenges as many investors’ time horizons have appeared toshorten with the rise of high-frequency trading, hedge funds, and the like. IPOsand reverse mergers are attractive financing options for graphene ventures today,but investor demand for such IPOs with the bigger, more established banks hasnot been very enthusiastic to date. Regulatory changes in the United States are

206 8 Financing Graphene Ventures encouraging and may foster greater access to capital for graphene ventures in the months ahead. Today, there is capital available to fund seed and early stage graphene ventures. Experience suggests being artful in funding such ventures, pursuing diligently as many sources of capital as possible. Importantly, while a founder or executive may have a well thought out plan of how to finance a graphene venture, it is important to remain opportunistic and take what the market will give you. The fact of the matter is that the volatility of capital markets is unpredictable and what a company can do funding-wise from week to week, month to month, quarter to quarter and year to year can shift significantly and sometimes drastically. It is interesting to note how many founders and executives, when faced with a funding decision that will result in having a smaller share of something potentially big or a large share of something that will likely be small, will choose the latter option. Experience has shown that pride and greed can be major impediments to a successful venture. Founders and executives of graphene ventures are strongly advised to be mindful of the pride and greed factors as they go forward. Lastly, there is a strong case to be made for graphene ventures collaborating with larger, established companies where such strategic alignments make good business sense. Well-crafted strategic partnerships help foster the commercialization of innovative graphene-enabled products to the benefit of all major constituents including, most importantly, the customer. Appendix Nantero Case Study – The Funding and Evolution of a Nanomaterials Start-up This case study explores the founding and funding of Nantero, Inc. Nantero is pio- neering the use of CNTs in memory devices. While silicon may continue to be the dominant material for the production of microprocessors and other types of semi- conductors over the next 5 years or so, scientists and researchers are exploring the use of other materials – materials that are highly scalable down to 1 or 2 nm and that would keep Moore’s law chugging along for the foreseeable future. One mate- rial that appears promising and has received a great deal of attention over the past decade is CNTs. CNTs are incredibly strong elastic cylinders of carbon atoms that bear a striking resemblance to a tube of rolled-up chicken wire. One nanotube is just 1/50 000th the diameter of a human hair. CNTs possess unique structural and electrical prop- erties. They are 117 times stronger than steel, half the density of aluminum, and have thermal and electrical conductivity properties that make the material use- ful for a variety of applications in a wide range of industries from aerospace and transportation to energy and electronics. Since their discovery in 1991, CNTs have been a focus of intense research in the United States and overseas, being viewed as a potential replacement for silicon in semiconductors in the future. The high electrical conductivity, thermal conductivity, and tensile strength of CNTs make them highly attractive for electronic device applications. These properties enable

8.4 The Graphene Financing Road Ahead 207performance breakthroughs both through incorporation into existing semicon-ductor products and in the development of next-generation products.The Founding of NanteroOne of the companies pioneering the use of CNTs in computer chips is Nan-tero. Nantero was founded in Woburn, MA, in 2001 by Dr Thomas Rueckes, GregSchmergel, and Dr Brent Segal. Dr Rueckes is the inventor of the innovative nano-™electromechanical NRAM design and the company’s Chief Technology Officer.Greg Schmergel is a successful entrepreneur and the company’s CEO and Dr Segalwas the COO. Rueckes, Schmergel, and Segal assembled a complete scientificteam consisting of specialists in multiple disciplines in the fields of nanotechnol-ogy and semiconductors to develop Nantero’s nanotechnology. Nantero was launched with the objective of developing and commercializing DrRueckes’ NRAM invention – a high-density nonvolatile random access memoryusing CNTs. Nantero’s founders envisioned NRAM as a universal memory devicethat would replace all existing forms of memory, such as DRAM, SRAM, and flashmemory. Additionally, the company’s founders sought to develop a proprietaryway to manufacture the nanotechnology memory chip and integrate it with stan-dard semiconductor processes. Development of such processes was seen as a wayto advance Nantero’s CNT memory device into the market more quickly. At the time of founding, Nantero envisioned a market opportunity for its NRAMchip in excess of USD70 billion per year. NRAM would enable instant-on comput-ers and replace DRAM and flash memory in devices such as smartphones, tablets,laptops, MP3 players, digital cameras, and enterprise systems of all kinds. Thereare also other applications in the networking segment. Nantero’s business modelis based on licensing the company’s proprietary CNT technology and processesto manufacturers for use in various applications. As such, Nantero is a next gen-eration fabless semiconductor company, one focused primarily on IP licensing,similar to ARM Holdings.Series A: Financing RoundIn October 2001, Nantero announced its first round of investment to developa nanotube-based universal memory. The amount raised in the Series A roundwas USD6 million. The proceeds of the capital raised were to be used to developNantero’s core NRAM technology and processes. Investors in the Series A roundincluded venture firms Draper Fisher Jurvetson, Stata Venture Partners, and Har-ris & Harris Group. Other investors included Alex d’Arbeloff, Chairman of MITand founder of Teradyne, who also joined Nantero’s Board of Directors. It should be noted that the financial environment in 2001 was not particu-larly conducive to raising significant sums of capital for emerging technologycompanies – especially companies pioneering the development of innovative

208 8 Financing Graphene Ventures nanotechnologies. In the year leading up to the completion of Nantero’s Series A round, the NASDAQ stock index – a bellwether for emerging technology stocks – had declined sharply, driven by a collapse of valuations in the previously high-flying dot com segment. Post-Series A: Funding Evolution In the spring of 2003, Nantero announced that it had achieved a milestone that involved creating an array of 10 billion suspended nanotube junctions on a single silicon wafer. This development showed that nanotubes could reliably be posi- tioned in large arrays and was able to scale to make even larger arrays. Nantero’s process also resulted in substantial redundancy for the memory, because each memory bit depends not on one single nanotube, but upon a large number of nan- otubes that resemble a fabric. The highly conductive single-layer nanotube fabrics were viewed as having a wide range of applications beyond memory chips, includ- ing for transistors, interconnects, and sensors. Creating this enormous array of suspended nanotubes using standard semiconductor processes brought Nantero closer to its end goal of mass-producing NRAM chips. Nantero’s innovative design for NRAM involved the use of suspended nanotube junctions as memory bits, with the “up” position representing bit zero and the “down” position, representing bit one. Bits are switched between states through the application of electrical fields. The wafer was produced using only standard semiconductor processes, maximizing compatibility with existing semiconductor factories. Nantero’s proprietary method for achieving this result involved deposit- ing a very thin layer of CNTs over the entire surface of the wafer, and then using lithography and etching to remove the nanotubes that are not in the correct posi- tion to serve as elements in the array. During 2003, Nantero added Dr Mohan Rao to its Scientific Advisory Board. Dr Rao was one of the world’s leading VLSI chip designers, having served previ- ously as Senior Vice President, Semiconductor Group, at Texas Instruments. He held over 100 patents worldwide on various aspects of memory, including SRAM, DRAM, and system-on-chip. Series B: Financing Round In September 2003, Nantero announced the completion of a Series B round of financing. The company raised USD10.5 million to further advance the develop- ment of its innovative NRAM technology. The Series B round was led by Charles Rivers Ventures, a venture firm with over three decades of experience in high technology. Bruce Sachs and Bill Tai, both Partners at Charles River Ventures, joined Nantero’s Board of Directors. Returning existing institutional investors included Draper Fisher Jurvetson, Stata Venture Partners, and Harris & Harris Group.

8.4 The Graphene Financing Road Ahead 209Post-Series B: Funding EvolutionAround the same time as its Series B funding round, Nantero announcedcollaboration with ASML, the leading global provider of lithography systemsfor the semiconductor industry headquartered in Veldhoven, the Netherlands.Nantero had developed a completely CMOS-compatible manufacturing processfor NRAM. The joint work with ASML was geared toward proving that Nantero’sproprietary process was achievable using popular lithography equipment. Thecollaboration demonstrated that ASML’s equipment is fully capable of handlingnanotubes using Nantero protocols and of carrying out Nantero’s new manufac-turing steps without any modifications. This was a significant milestone for thecompany and set it on course to advance commercialization of its CNT memorydevices. In the spring of 2004, Nantero announced that OB Bilous joined the Company’sBoard of Directors. At the time Mr Bilous was Chairman of the Board of Interna-tional SEMATECH and held 15 patents and had published numerous articles andpresented in many conferences and technical meetings in the area of semicon-ductor technology. Mr Bilous was previously VP of Worldwide Manufacturing forIBM in microelectronics. In the summer of 2004, Nantero announced that it had been issued a seminalpatent covering CNT films and fabrics by the US Patent and Trademark Office. Thepatent relates to a CNT film comprising a conductive fabric of CNTs deposited ona surface. The CNT film is highly useful in a variety of applications, including inthe semiconductor industry wherein the CNT film is deposited on a silicon sub-strate. The CNT film is a major innovation enabling cost-effective, high-volumeproduction of CNT-based devices and other products. With the addition of thenew patent, Nantero had grown its patent portfolio to 10 granted US patents anda pipeline of over 40 additional patents pending. Also in the summer of 2004, Nantero announced a CNT development projectwith LSI Logic to develop semiconductor process technology, expediting the effec-tive use of CNTs in CMOS fabrication. During the same time, the company alsoannounced a joint evaluation of CNT-based electronics with BAE Systems to eval-uate the potential to develop CNT-based electronic devices for use in advanceddefense and aerospace systems. The project involved research and developmentof a variety of next-generation electronic devices that can be built leveraging theunique properties of CNTs and using Nantero’s proprietary methods and pro-cesses for the design and manufacture of nanotube-based electronics. In the fall of 2004, Nantero announced that it had been awarded USD4.5million to develop CNT-based radiation-hard non-volatile RAM with CNI andCase-Southwest Missouri State University. The radiation-resistant nanotechnol-ogy was targeted for use in space for US defense purposes. In early 2005, Nanteroannounced it was actively seeking manufacturing partners in Europe and Asia tofurther advance the commercialization of its NRAM in the consumer electronicssegment. Germany, France, Italy, the Netherlands, Japan, and Korea were amongthe countries earmarked for possible licensing of the NRAM technology.

210 8 Financing Graphene Ventures Series C: Financing Round Nantero announced the completion of a USD15 million Series C round of fund- ing in the first quarter of 2005. Globespan Capital Partners was the lead investor in the round and Ullas Naik, a Managing Director with Globespan Capital Part- ners based in their Boston office, joined Nantero’s Board of Directors. Returning existing institutional investors included Charles River Ventures, Draper Fisher Jurvetson, Stata Venture Partners, and Harris & Harris Group. Post-Series C: Funding Evolution In spring of 2006, Nantero announced it had fabricated and successfully tested a 22-nm NRAM memory switch. The switch demonstrated that NRAM is scalable to numerous process technology nodes over several decades. Nantero’s NRAM switches were tested by writing and reading data using 3 ns cycle times, giving it the potential to match the fastest memories in production at the time. The NRAM switches were fabricated using the company’s proprietary CNT fabric. The results demonstrated that NRAM can be the standalone and embedded memory of choice as the technology combines the nonvolatility of flash with the speed of SRAM and the density of DRAM. The tests showed that NRAM can be scaled for many future generations and we believe the scaling will continue down to below the 5 nm technology node. Also in the spring of 2006, Nantero announced collab- oration with ON Semiconductor to jointly develop CNT technology, continuing ongoing work to integrate CNTs in CMOS fabrication. In fall 2006, Nantero announced a major corporate milestone: that it has resolved all of the major obstacles that had been preventing CNTs from being used in mass production in semiconductor fabs. This was a significant devel- opment for the company. Nanotubes were widely acknowledged to hold great promise for the future of semiconductors, but most experts had predicted it would take a decade or two before they would become a viable material. This was due to several historic obstacles that prevented their use, including a previous inability to position them reliably across entire silicon wafers and contamination previously mixed with the nanotubes that made the nanotube material incompatible with semiconductor fabs. Nantero has developed a method for positioning CNTs reliably on a large scale by treating them as a fabric that can be deposited using methods such as spin- coating, and then patterned using lithography and etching, all common CMOS processes present in every semiconductor fab. Nantero also developed a method for purifying CNTs to the standards required for use in a production semiconduc- tor fab, which means consistently containing less than 25 parts per billion of any metal contamination. With these innovations, Nantero became the first company in the world to introduce and use CNTs in mass production semiconductor fabs. The summer of 2007 saw Nantero providing license rights for biomedical sen- sors to Alpha Szenszor, Inc., a newly formed company also based in Woburn, Mas- sachusetts. Alpha Szenszor was founded to develop a suite of sensor products in

8.4 The Graphene Financing Road Ahead 211the medical field, including portable and cost-effective detectors for infectious dis-eases such as HIV. Alpha Szenszor’s co-founders include Steve Lerner, an industryveteran with over 28 years of product development and industrialization expe-rience. Additionally during this time, Nantero announced it was working withHewlett-Packard (HP) to explore the use of HP inkjet technology and the com-pany’s CNT formulation to create flexible electronics products and develop lowcost printable memory applications. Nantero used HP’s thermal inkjet pico-fluidicsystem (TIPS) research and development tool to evaluate the company’s inkjettechnology for printable memory applications that can be used in a wide range ofapplications including low cost RFID tags. During summer of 2008, Nantero announced collaboration with SVTC Tech-nologies to accelerate the commercialization of nanotube-based electronicsproducts. The collaboration was part of SVTC’s broader mission to enable com-mercialization of new process and device developments in the semiconductor,MEMS, and related nanotechnology domains with support for a direct pathbetween the work completed in SVTC’s facilities to high-volume manufacturing.Together, Nantero and SVTC offered CNT device development capabilities forcustomers targeting a wide range of applications including photovoltaics (solarcells), LEDs, sensors, MEMS, and other semiconductor-based devices. Also in 2008 Nantero announced a significant milestone with the acquisition ofits government business unit by Lockheed Martin. Additionally, Lockheed Martinentered an exclusive license arrangement with Nantero for government-specificapplications of Nantero’s extensive intellectual property portfolio. Approximately30 employees, including Nantero’s co-founder and COO Dr Brent Segal joinedLockheed Martin as part of the purchase of Nantero’s government business unit.Lockheed Martin’s Advanced Technology Center, a unit of Lockheed Martin SpaceSystems Company, would manage the Nantero unit going forward. At the time,Lockheed Martin was a leader in the research, development, and application ofnanotechnology to future military and intelligence applications. Deal terms wereundisclosed. The year 2008 also saw Nantero ranked Number 54 in Inc. Magazine’s“Annual List of America’s 500 Fastest-Growing Private Companies, with three-yearsales Growth of 2833%. Along with the ranking of 54 in the Inc 500, Nantero was#2 in the computer and electronics category and #2 for Boston area companies. In late 2009, Lockheed Martin announced that it tested a radiation-resistant ver-sion of Nantero NRAM devices on a NASA Shuttle mission. NASA incorporatedthe NRAM technology into special autonomous testing configurations installedinto a carrier at the aft end of the payload bay. It was launched into space as partof STS-125, the May 2009 mission of the Space Shuttle Atlantis that successfullyserviced the Hubble Space Telescope. The experiment was a proof-of-concept thatenabled the testing of launch and re-entry survivability, as well as basic function-ality of the CNT memory in orbit throughout the shuttle mission. The NRAMdevices were early prototype parts, and performed the same before, during, andafter completion of the mission. This mission represents an important first step inthe development of high-density, non-volatile, carbon-nanotube-based memoriesfor spaceflight applications.

212 8 Financing Graphene Ventures Series D: Financing Round In 2013, Nantero announced the closing of a Series D round of funding that raised more than USD15 million. The round also included existing investors Charles River Ventures, Draper Fisher Jurvetson, Globespan Capital Partners, Stata Ven- ture Partners, and Harris & Harris Group. The round also included new strategic corporate investors. One of those investors is Schlumberger, the world’s leading supplier of technology, integrated project management and information solutions to customers working in the oil and gas industry worldwide. Additionally, Michael Raam was added to the Advisory Board. Mr Raam was most recently the CEO of SandForce, Inc., the successful SSD controller company and then VP/GM of the Flash Components Division for LSI after LSI acquired SandForce. Also, near the end of 2009, Nantero announced that Dr Tsugio Makimoto, a Japanese semiconductor industry pioneer, joined its Advisory Board. Previously, Dr Makimoto was a Corporate Advisor of Sony Corporation in charge of semicon- ductor technology. It was noted that Dr Makimoto’s deep knowledge and experi- ence in the semiconductor industry, both in Japan and globally, would assist in helping the company continue to add new customers in Asia. Post-Series D: Funding Evolution In 2014, a team of researchers in Japan independently verified that Nantero’s NRAM, had excellent properties and could be used for various applications rang- ing from main memory to storage. Ken Takeuchi, professor at the Department of Electrical, Electronic and Communication Engineering, Faculty of Science and Engineering of Chuo University, led the Japanese team that conducted the NRAM research. The details of the new technology were announced in a lecture titled “23% Faster Program and 40% Energy Reduction of CNT Non-volatile Memory with Over 1011Endurance” (lecture number: T11-3) at 2014 Symposia on VLSI Technology and Circuits in Honolulu, Hawaii. D. Takeuchi was previously leading Toshiba’s NAND flash memory circuit design for 14 years. Series E: Financing Round In spring of 2015, Nantero announced the completion of a USD31.5 million Series E financing round. The round included new investors and participation from exist- ing investors Charles River Ventures, Draper Fisher Jurvetson, Globespan Capital Partners, and Harris & Harris Group. The company noted that the round was sub- stantially oversubscribed. Nantero noted that it intended to use its new funding to continue the acceleration of NRAM as the leading next-generation memory for both storage class memory and as a replacement for flash and DRAM. In addition to the funding, Nantero noted that it added two new company advi- sors. Among the new advisors were Dr Stefan Lai, a former Intel senior executive who co-invented the EPROM tunnel oxide (ETOX) flash memory cell and led the

8.4 The Graphene Financing Road Ahead 213company’s phase change memory (PCM) team, and Dr Yaw Wen Hu, a formerExecutive VP and current Board member of Inotera Memories, where he over-saw new DRAM technology transfer and development of Wafer Level Packaging.Before that, Dr Hu was an Executive VP and Chief Operating Officer for SiliconStorage Technology (SST), where he was responsible for SuperFlash technologydevelopment, working with a team from a brand new memory cell concept to highvolume product shipment and its establishment as the choice of technology forEmbedded Flash applications. In summer 2015, Nantero announced that arrays of its new generation of super-fast, high-density memory (NRAM) were independently tested by Chuo Univer-sity and the results showing excellent performance and reliability will be presentedin a technical paper at the 2015 International Conference on Solid State Devicesand Materials. Nantero announced that former TSMC executive Dr Shang-Yi Chi-ang had joined the company’s advisory board. With more than 40 years experiencein the semiconductor industry, including as co-COO and EVP at TSMC, Dr Chi-ang has contributed to the research and development of CMOS, NMOS, Bipolar,DMOS, SOS, SOI, GaAs lasers, LED, E-Beam lithography, and silicon solar cells.In addition, Nantero also announced the addition of Lee Cleveland to the execu-tive management team. Formerly in charge of flash design at Spansion and AMD,Mr Cleveland is VP of Design and responsible for leading Nantero’s complete chipdesign team. The company also announced the appointment of renowned memoryindustry executive Ed Doller to its Advisory Board. Mr Doller was previously VP& Chief Strategist of the NAND Solutions Group at Micron, where he also servedas VP & GM Enterprise Storage and as VP & Chief Memory Systems Architect.SummaryThe development of Nantero spans 15 years of intensive research and develop-ment in the area of advanced nanomaterials. With one CNT being just 1/50 000ththe diameter of a human hair, these tiny cylinders are far stronger than steel,half the density of aluminum, and have better thermal and electrical conductiv-ity properties than any other material scientists are aware of today. As a pioneerin nanotechnology, Nantero is the first company to develop semiconductor prod-ucts using this material in production of CMOS fabs. Nantero’s innovative NRAMtechnology has several characteristics that make it attractive as a next-generationtechnology for standalone and embedded usage. These include:• CMOS compatible: Works in standard CMOS fabs with no new equipment needed• Limitless scalability: Designed to scale below 5 nm in the future• High endurance: Proven to operate for orders of magnitude more cycles than flash• Faster Read and Write: Same as DRAM, 100s of times faster than NAND• High reliability: will retain memory for >1000 years at 85 ∘C or more than 10 years at 300 ∘C

214 8 Financing Graphene Ventures • Low power: Zero in standby mode, 160x lower write energy per bit than NAND • Low cost: Simple structure, can be 3D multi-layer and multi-level cell (MLC) One of the key advantages of Nantero’s NRAM technology is the ability to con- solidate DRAM with flash memory, since NRAM is both as fast as DRAM and nonvolatile like flash. This ability will undoubtedly be attractive to manufacturers seeking to create smaller, more powerful devices in the future. Potential future applications of the NRAM technology include: Virtual Screens, Next-Generation Enterprise Systems, Rolled-up Tablets, Instant-On Laptops, 3D Video Phones, and other products needing huge amounts of fast memory, targeting a wide range of markets such as consumer electronics, mobile computing, wearables, Internet of Things, enterprise storage, government/military, space, and automotive. Nantero has raised over USD87 million of capital over the past 15 years and earned over USD70 million in revenue. The company’s nanotechnology is under development in multiple world-class manufacturing facilities and the company has more than a dozen major corporate partners actively working on NRAM. Sample NRAMs have demonstrated production-ready yields (i.e., >99.999%). Per- formance for both standalone and embedded memory applications are superior to anything in the market. Nantero is currently sampling memory test chips to cus- tomers at a time when there is a large demand for new high-density standalone memory and for high reliability, scalable embedded memory. Nantero has the potential to scale its business significantly beyond NRAM through licensing of its extensive CNT patent portfolio for other applications in a wide variety of sectors. The company’s business model is similar to that of ARM Holdings, PLC, a peer that has risen to prominence over the years and today is generating over a half a billion of revenue per annum. There is considerable opportunity for Nantero and its CNT technology in the USD330 billion-plus semiconductor industry as chips scale further down into the quantum realm with advanced nanomaterials that have properties that extend beyond silicon. The age of CNT electronics has begun. From this case study we see the time, effort, and patience required to commer- cialize a transformative deep science-enabled technology such as NRAM relative to a social media application circa 2012–2014. Nantero’s path to commercial- ization extends over a period of 15 years. The company has progressed with the development of its innovative nanotechnology without having access to the public capital markets. While a significant amount of private investment capital has been invested in Nantero – over USD87 million to date – the company would have no doubt benefitted from having access to the public capital markets to accelerate the commercialization of its technology. As such, Nantero was unable to tap the pub- lic markets for growth capital as many of its silicon brethren did in the decades of the 1970s, 1980s, and 1990s. Whether the company will be able to tap the public capital markets for funding in the future remains to be seen.

215Indexa 15 clean/green energy 115active corrosion inhibitors 156, 157 CNT, see carbon nanotubes (CNTs)additive manufacturing, see 3-D printing composites, fabrication 77additives 63 conventional polymer composites 64alkylated graphene oxide (AGO) 87 corrosion 155aluminum 8 – complexity 155Angel investors 195 – inhibition 157anodes, graphene 122 – metal 155anodic passivation 157 – resistance 157atomic force microscope (AFM) 6 corrosion inhibitionatomic transfer radical polymerization – different modes 156 – graphene 161 (ATRP) 70, 86 – graphene/ceramic nanocomposites 171automotive and aircrafts 99 – graphene/metal nanocomposites 168azomethineylide 70 – graphene nanocomposites 164 corrosion-resistant coating 165b 116 covalent modification 67batteries– description 115 e– evolution of 116 electrochemical exfoliation, graphite 38– lithium-ion, see lithium-ion batteries electrochemical impedanceBayer Material Science (BMS) 4, 200Bluestone Global Tech (BGT) 185 spectroscopy (EIS) 168bolaamphiphile 40 EMI shielding 97bottom-up approach 53, 65 energy consumption evolution 114bulk nanomaterials possessing 2 environmental, health, and safety (EH&S)c procedures 7carbon nanotubes (CNTs) 1, 200 entrepreneurs 192– – inner surface 82 equity capital 198cathodes, graphene 126 esterification 68cathodic protection 156 – reaction, PVC 68chemically converted graphene (CCG)/PS ethylene vinyl acetate (EVA) 90 exfoliation, graphite 38 composite 89chemical vapor deposition (CVD) 65 f– graphene 27 few-layer graphene 20China Innovation Alliance of the Graphene filler 63 flammability reduction 99 Industry (CGIA) 185Graphene Technology: From Laboratory to Fabrication,First Edition. Edited by Soroush Nazarpour and Stephen R. Waite.© 2016 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2016 by Wiley-VCH Verlag GmbH & Co. KGaA.

216 Index – lithium-sulfur 127 – market and application projections 185 forward osmosis (FO) 146 – market past and present 178 fossil fuels 115 – mechanical cleavage 32 fuels, availability 113 – metal interface 160 functionalized GO (fGO) 67 – MO composite formation 124 functionalized graphene nanosheets – NanoXplore 14, 17 – nitrogen-doped 124 (f-GNSs) 68 – nomenclature 20 functionalized graphene sheets – oxo-functionalized 50 – patent filings 3 (FGS)/waterborne polyurethane (WPU) – PMMA composite 168 composites 72 – Poisson’s ratio 66 Future and Emerging Technology (FET) – polydispersity 20 initiatives 184 – preparation and properties 65 – as protective barrier 159 g – reductive functionalization 40 galvanization, steel 157 – selected properties 121 (GA)-PDMS composites (GAPC) 92 – size 20 Gartner’s Hype Cycle 9 – sodium-ion batteries (NIBs) 127 gas barrier 95 – as standalone corrosion resistant gas separation 140 germanane 2 coating 162 GO, see graphene oxide (GO) – startups 190 – dispersions 137 – structure 20 – nanosheets 139 – supermarket 178 – preparation 138 – synthesis 19 GO/sodium carboxymethyl cellulose – thermal and electrical properties 64 – 3D Labs 181 (NaCME) 94 – 3D printing 14 GO – TiO2 (GOT) 147 – ventures 193 government funding 196 – ventures, shifting financial grapheme, oxidative functionalization 43 graphene 1, 66 landscape 199 – anodes 122 – Young’s modulus 66 – applications 3 Graphene Flagship 184 – and batteries 11 graphene flakes (GFs) 85 – bi-layer 20 Graphene Laboratories 180 – cathodes 126 graphene nanoplatelets (GNPs) 181 – characterization by Raman graphene oxide (GO) 44, 133 Graphene Stakeholders Association spectroscopy 22 – chemical-vapor-deposition 27 (GSA) 185 – commercialization 191 graphene-based membranes – co-ordinated market initiatives 184 – preparation 134 – corrosion inhibition, – separation applications 140 graphene/ceramic nanocomposites, corrosion nanocomposites 164 – delamination from graphite 31 inhibition 171 – delamination of graphite to 19 graphene/metal nanocomposites, corrosion – discovery 3 – electrical and thermal conductivity 66 inhibition 168 – epitaxial growth, SiC 26 graphene/polymer composites – financing road ahead 203 – characterization 74 – functionalization 65, 66 – application 94 – GO as precursor 49 – dynamic mechanical properties 93 – heat management 13 – electrical properties 88 – Li-ion batteries 120 – mechanical properties 77 – lithium-air 127 – preparation 71 – lithium-ion batteries 121 – lithium-polymer 127

Index 217– thermal properties 84 mGraphenea 179 melt mixing technique 72graphenide 40 modified Hummer’s method 66graphite molecular precision manufacturing– electrochemical exfoliation 38– liquid phase exfoliation (MPM) 1, 5– – shear mixing 36 molybdenum disulfide (MoS2) 1– – smart surfactants 38– – sonication 35 n– – stirred media mills 33 nano cap category 182– lithium-ion batteries 119 nanographene synthesis, small– mechanical cleavage 32graphite oxide (GrO) 44, 134 molecules 52graphitic acid 44, 46 nanomaterials – characterization 5h – effect 4hexadecyl-functionalized low-defect graphene – graphene 7 Nantero 207, 208, 210 nanoribbons (HD-GNRs) 81 National Physical Laboratory (NPL) 6hexavalent chromium based coatings 158 new tools, new standards, new protocols, andHigh Net Worth (HNW) investors 194Hummers’ method 47 new processes (TSPPs) 1, 5 N-methylpyrrolidone (NMP) 165 non-covalent surface modification 70 non-renewable energies 114i 192 oin-situ polymerization 71 193 octa(aminophenyl) silsesquioxaneinitial public offerings (IPOs)investor groups classification (OAPS) 79 octadecylamine (ODA) 90j overprotection 157Jumpstart Our Business Startups oxo-functionalized graphene 43, 47, 50 (JOBS) Act 204l player-by-layer assembly (LbL) 73 PANI 168LIB 12 Parthian battery 115linear low density polyethylene (LLDPE) PMMA grafted GO (G-PMMA) 72 Poisson’s ratio, graphene 66 matrix 67, 92 poly(allylamine hydrochloride) (PAH) 139liquid phase exfoliation, graphite poly(vinylidene fluoride) (PVDF) 71– shear mixing 36 polyamic acid (PAA) 165– smart surfactants 38 poly(glycidyl methacrylate) functional segment– sonication 35– stirred media mills 33 polymer chains (Py-PGMA) 71lithium-ion batteries polymer composites 63– advantages 117 polymer nanocomposites 63– carbon materials 118 polystyrene (PS) 67– electrolyte 118 polyvinyl pyrrolidon (PVP) 82– graphene 120, 121 Private Equity (PE) investors 196– graphite 119 PS-polyacrylamide (PS-PAM)– materials used 118– negative electrode 118 copolymer 70– operation principles 116 PVA functionalized GO (f-PVA-GO) 78– positive electrode 118 1-pyrenebutyric acid (PBA) 85– schematic structure 117LLDPE, see linear low density polyethylene r 26 Raman spectra, graphene (LLDPE) matrix reduced GO 48 renewable energies 114

218 Index 3D printing, graphene 14 transmission electron microscopy s sacrificial coatings 158 (TEM) 74, 75 salt-water immersion 167 trough of disillusionment 10 scanning electron microscopy (SEM) 74 TSPPs, see new tools, new standards, new scanning probe microscopy (SPM) 6 Schottky barrier 160 protocols, and new processes (TSPPs) scotch tape method 32 turbine blades 100 Securities and Exchange Commission 2D nanomaterials 2 – development 8 (SEC) 204 2D peak 23 self-healing 156 – coatings 157 u 6 sensor 97 ultrasonic force microscopy (UFM) silicene 2 United States Federal Highway single-layer graphene nanosheets Administration 155 (SLGNs) 86 slim hourglass electronic structure 159 v sodium-ion batteries (NIBs) 127 Venture Capital 195 solution mixing 72 volatile organic compounds (VOCs) 158 sulfanilic acid azocromotrop (SAC) 71 sulfonated polyethersulfone (SPES) w water treatment 142 membranes 139 wonder material 3 surfactants 38 synthesize modified filler (SATPGO) 83 x x-ray diffraction (XRD) 74 t Tafel analysis 168, 170 y Tafel plot 166 Young’s modulus, graphene 66 thermogravimetric analysis (TGA) 72 3D nanomaterials, development 8


Click to View SlideBook Version
Previous: Biotechnology An Illustrated Primer Next: Biotechnology of Microbial Enzymes