Electrochromic Materials and Devices

The Editors

Prof. Dr. Roger J. Mortimer†

Loughborough University

Department of Chemistry

Loughborough

LE11 3TU Leicestershire

UK

Dr. David R. Rosseinsky

University of Exeter

School of Physics

EX4 4QL Exeter

UK

Rev Dr. Paul M. S. Monk

St. Barnabas' Vicarage

1 Arundel Street

OL4 1NL Clarksfield, Oldham

UK

Cover

Sprayed films of electrochromic polymers developed at the University of Florida and Georgia Institute of Technology with the John Reynolds Research Group. Artistic concept and photography by Aubrey Dyer, Keith Johnson, and Justin Kerszulis.

All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.

Library of Congress Card No.: applied for

British Library Cataloguing-in-Publication Data

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 <http://dnb.d-nb.de>.

© 2015 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: 978-3-527-33610-4

ePDF ISBN: 978-3-527-67988-1

ePub ISBN: 978-3-527-67987-4

Mobi ISBN: 978-3-527-67986-7

oBook ISBN: 978-3-527-67985-0

Printing and Binding Markono Print Media Pte Ltd, Singapore

In memoriam

1956 – 2015

We record with deep sadness the untimely death of our co-editor Roger J. Mortimer, a fine academic colleague and friend, and initiator of this book. We dedicate this book to his memory.

Preface

Electrochromic materials have the property of a change, evocation or bleaching, of colour as effected either by an electron-transfer (redox) process or by a sufficient electrochemical potential. Although materials are usually said to be electrochromic when light is modulated by reflectance or absorbance in the visible region of the electromagnetic spectrum – colour changes perceptible to the human eye – interest in electrochromic devices (ECDs) for multispectral energy modulation, to include the infrared and microwave regions, has extended the working definition [1].

While the topic of electrochromism has a history dating back to the nineteenth century, only in the last quarter of the twentieth has its study gained a real impetus. So, applications have hitherto been limited, apart from the astonishing success of the 250 million Gentex anti-dazzle car mirrors that have been sold since 1987 and the adjustable-darkening windows of the Boeing Dreamliner aircraft. The ultimate goal of contemporary studies is the provision of large-scale electrochromic ‘smart’ windows/glazing for buildings at modest expenditure which, applied widely in the United States, would save billions of dollars in air-conditioning costs. In tropical and equatorial climes, savings would be proportionally greater: Singapore, for example spends one-quarter of its Gross Domestic Product (GDP) on air conditioning, a sine qua non for tolerable living conditions there. Importantly, note that only weak anti-thermal protection is provided by colour alone, the electrochrome itself getting heated. Hence, transparent metal oxides that can be electrochemically or otherwise reduced to form shiny metallic reflectors are of fundamental importance, though this process is widely under-emphasised. Numerous other applications have been contemplated, and for some, prototype devices have been developed. Applications include electrochromic strips as battery state-of-charge indicators, electrochromic sunglasses, reusable price labels, protective eyewear, controllable aircraft canopies, glare-reduction systems for offices, devices for frozen-food monitoring, camouflage materials, chameleonic fabrics, spacecraft thermal control, an optical iris for a camera lens and (non-emissive) controllable light-reflective or light-transmissive display devices for optical information and storage.

This edited book follows our earlier research monographs [2, 3] now with invited contributions from the main experts across the globe. Part One concerns electrochromic materials and processing and covers metal oxides, Prussian blue, viologens, conjugated conducting polymers, transition metal coordination complexes and polymers, organic near-infrared materials and metal hydrides. Part Two concerns nanostructured electrochromic materials and device fabrication and covers nanostructures in electrochromic materials, advances in polymer electrolytes for ECD applications, gyroid-structured electrodes for electrochromic (and supercapacitor) applications, layer-by-layer assembly of electrochromic materials and plasmonic electrochromism of metal oxide nanocrystals. Part Three describes the applications of electrochromic materials and covers solution-phase ECDs and systems, electrochromic smart windows and fabric electrochromic displays. Part Four covers device case studies, environmental impact issues and elaborations and includes a case study of an electrochromic foil, life cycle analysis (LCA) of electrochromic windows, a case study of the installation, operation, monitoring and user experience of electrochromic glazing in a UK office and photoelectrochromic devices. The book closes with an Appendix, where electrochromic materials and device performance parameters are defined, to include some cautions about their comparisons between different research laboratories.

References

1. Rauh, R.D. (1999) Electrochromic windows: an overview. Electrochim. Acta, 44, 3165–3176.

2. Monk, P.M.S., Mortimer, R.J., and Rosseinsky, D.R. (1995) Electrochromism – Fundamentals and Applications, Wiley-VCH Verlag GmbH, Weinheim.

3. Monk, P.M.S., Mortimer, R.J., and Rosseinsky, D.R. (2007) Electrochromism and Electrochromic Devices, Cambridge University Press, Cambridge.

Acknowledgements

We are indebted to numerous colleagues and correspondents with whom we have collaborated over many years in electrochromics research, in particular Professor Hassan Kellawi of Damascus University, who introduced DRR to Electrochromism, which then led to RJM and PMSM becoming interested in this research field. Special thanks are due to Professor Paulo R. Bueno of São Paulo State University (UNESP), whose papers reviewed in Chapter 3 helped clarify our general summary of that research area.

From the staff of Wiley-VCH, we wish to thank Dr Bente Flier (Associate Commissioning Editor), who first commissioned the book, Dr Martin Preuss (Senior Commissioning Editor), Lesley Belfit (Project Editor, STMS Books) and the production team at Wiley-VCH, for their professionalism, care and help.

Professor Roger J. Mortimer

Loughborough, UK

Dr. David R. Rosseinsky

Exeter, UK

Rev. Dr. Paul M. S. Monk

Oldham, UK

List of Contributors

Harlan J. Byker

Pleotint LLC

18th Avenue

Jenison, MI 49428

USA

Hsin-Wei Chen

National Taiwan University

Department of Chemical Engineering

No 1, Sec. 4, Roosevelt Road

Taipei

Taiwan

Susana I. Córdoba de Torresi

Universidade de São Paulo

Instituto de Química

C.P 26077

05513-970 São Paulo, SP

Brazil

Aubrey L. Dyer

Clayton State University

Department of Natural Sciences

College of Arts and Sciences

Clayton State Blvd.

Morrow, GA 30260

USA

and

Center for Organic Photonics and Electronics

Georgia Institute of Technology

School of Chemistry and Biochemistry

School of Materials Science and Engineering

Atlantic Dr. NW

Atlanta, GA 30332

USA

Alice Lee-Sie Eh

Nanyang Technological University

School of Materials Science and Engineering

Nanyang Avenue

Singapore 639798

Singapore

Claes-Göran Granqvist

Uppsala University

Department of Engineering Sciences

The Ångström Laboratory

PO Box 534, SE-75121

Uppsala

Sweden

Matthias Harsch

LCS Life Cycle Simulation GmbH

Aspacher Strasse 9

D-71522 Backnang

Germany

Kuo-Chuan Ho

National Taiwan University

Department of Chemical Engineering

No 1, Sec. 4, Roosevelt Road

Taipei

Taiwan

Chih-Yu Hsu

Electronic Functional Materials Group, Polymer Materials Unit

National Institute for Materials Science (NIMS)

1-1 Namiki

305-0044 Tsukuba

Japan

Chih-Wei Hu

National Taiwan University

Department of Chemical Engineering

No 1, Sec. 4, Roosevelt Road

Taipei

Taiwan

and

National Institute of Advanced Industrial Science and Technology (AIST)

Anagahora 2266-98

Shimoshidami, Moriyama-ku

Nagoya 463-8560

Japan

Fritz Huguenin

Universidade de São Paulo

Faculdade De Filosofia

Ciências e Letras de Ribeirão Preto

Av Bandeirantes 3900

14040-901 Ribeirão Preto, SP

Brazil

Michael A. Invernale

University of Connecticut

Department of Chemistry and the Polymer Program

North Eagleville Road U-2136

Storrs, CT 06269-3136

USA

Bjørn Petter Jelle

SINTEF Building and Infrastructure

Department of Materials and Structures

Høgskoleringen 7A

NO-7465 Trondheim

Norway

and

Norwegian University of Science and Technology (NTNU)

Department of Civil and Transport Engineering

Høgskoleringen 7B

NO-7491 Trondheim

Norway

Keith E. Johnson

Center for Organic Photonics and Electronics

Georgia Institute of Technology

School of Chemistry and Biochemistry

School of Materials Science and Engineering

Atlantic Dr. NW

Atlanta, GA 30332

USA

Pooi See Lee

Nanyang Technological University

College of Engineering

School of Materials Science and Engineering

Nanyang Avenue

639798

Singapore

Anna Llordes

The University of Texas at Austin

McKetta Department of Chemical Engineering

Austin, TX 78712

USA

and

The Molecular Foundry

Lawrence Berkeley National Laboratory

Berkeley, CA 94720

USA

Sebastien D. Lounis

The University of Texas at Austin

McKetta Department of Chemical Engineering

Austin, TX 78712

USA

and

The Molecular Foundry

Lawrence Berkeley National Laboratory

Berkeley, CA 94720

USA

and

The University of California

Graduate Group in Applied Science & Technology

Berkeley, CA 94720

USA

Xuehong Lu

Nanyang Technological University

College of Engineering

School of Materials Science and Engineering

Nanyang Avenue

Singapore 639798

Singapore

John Mardaljevic

Loughborough University

School of Civil and Building Engineering

Loughborough

Leicestershire LE11 3TU

UK

Jose R. Martins Neto

Universidade de São Paulo

Instituto de Química

C.P 26077

05513-970 São Paulo, SP

Brazil

Delia J. Milliron

The University of Texas at Austin

McKetta Department of Chemical Engineering

Austin, TX 78712

USA

Paul M. S. Monk

St Barnabas' Vicarage

Arundel Street

Clarksfield

Oldham OL4 1NL

UK

Roger J. Mortimer†

Loughborough University

Department of Chemistry

Loughborough

Leicestershire LE11 3TU

UK

Anna M. Österholm

Center for Organic Photonics and Electronics

Georgia Institute of Technology

School of Chemistry and Biochemistry

School of Materials Science and Engineering

Atlantic Dr. NW

Atlanta, GA 30332

USA

Michael T. Otley

University of Connecticut

Department of Chemistry and the Polymer Program

North Eagleville Road U-2136

Storrs, CT 06269-3136

USA

Birgit Painter

De Montfort University

Institute of Energy and Sustainable Development

The Gateway, Leicester

LE1 9BH

UK

Uwe Posset

Center for Applied Electrochemistry

Fraunhofer-Institut für Silicatforschung ISC

Neunerplatz 2

D-97082 Würzburg

Germany

John R. Reynolds

Center for Organic Photonics and Electronics

Georgia Institute of Technology

School of Chemistry and Biochemistry

School of Materials Science and Engineering

Atlantic Dr. NW

Atlanta, GA 30332

USA

David R. Rosseinsky

University of Exeter

School of Physics

Exeter EX4 4QL

UK

Evan L. Runnerstrom

The University of Texas at Austin

McKetta Department of Chemical Engineering

Austin, TX 78712

USA

and

The Molecular Foundry

Lawrence Berkeley National Laboratory

Berkeley, CA 94720

USA

and

Department of Materials Science and Engineering

Berkeley, CA 94720

USA

Maik R.J. Scherer

Adolphe Merkle Institute

Chemin des Verdiers

Fribourg

Switzerland

D. Eric Shen

Center for Organic Photonics and Electronics

Georgia Institute of Technology

School of Chemistry and Biochemistry

School of Materials Science and Engineering

Atlantic Dr. NW

Atlanta, GA 30332

USA

Ullrich Steiner

Adolphe Merkle Institute

Chemin des Verdiers

Fribourg

Switzerland

Gregory A. Sotzing

University of Connecticut

Department of Chemistry and the Polymer Program

North Eagleville Road U-2136

Storrs, CT 06269-3136

USA

Thomas S. Varley

University College London

Department of Chemistry

Gower Street

WC1H 0AJ London

UK

Marcio Vidotti

Universidade Federal de Paraná

Departamento de Química

CP 19081

81531-990 Curitiba

Brazil

Xinhua Wan

Peking University

Beijing National Laboratory for Molecular Sciences

Key Laboratory of Polymer Chemistry and Physics of Ministry of Education

College of Chemistry and Molecular Engineering

No.5 Summer Palace Road, Haidian District

Beijing 100871

China

Ruth Kelly Waskett

Institute of Energy and Sustainable Development

De Montfort University

Leicester

The Gateway LE1 9BH

UK

Shanxin Xiong

Xi'an University of Science and Technology

College of Chemistry and Chemical Engineering

Yanta Road

Xi'an

R China

Bin Yao

Peking University

Beijing National Laboratory for Molecular Sciences

Key Laboratory of Polymer Chemistry and Physics of Ministry of Education

College of Chemistry and Molecular Engineering

No.5 Summer Palace Road, Haidian District

Beijing 100871

China

Kazuki Yoshimura

Structural Materials Research Institute

National Institute of Advanced Industrial Science and Technology (AIST)

Anagahora, Shimoshidami, Moriyama-ku

Nagoya

Aichi 463-8560

Japan

Jie Zhang

Peking University

Beijing National Laboratory for Molecular Sciences

Key Laboratory of Polymer Chemistry and Physics of Ministry of Education

College of Chemistry and Molecular Engineering

No.5 Summer Palace Road, Haidian District

Beijing

China

Yu-Wu Zhong

Beijing National Laboratory for Molecular Sciences

CAS Key Laboratory of Photochemistry

Chinese Academy of Sciences, Institute of Chemistry

Bei Yi Jie, Zhong Guan Cun

Beijing 100190

China

Part I

Electrochromic Materials and Processing

Chapter 1

Electrochromic Metal Oxides: An Introduction to Materials and Devices

Claes-Göran Granqvist

1.1 Introduction

Electrochromic materials are able to change their properties under the action of an electrical voltage or current. They can be integrated in devices that modulate their transmittance, reflectance, absorptance or emittance. Electrochromism is known to exist in many types of materials. This chapter considers electrochromic metal oxides and devices based on these.

Figure 1.1 shows a generic electrochromic device comprising five superimposed layers on a single transparent substrate or positioned between two transparent substrates [1]. Its variable optical transmittance ensues from the electrochromic films, which change their optical absorption when ions are inserted or extracted via a centrally positioned electrolyte. The ion transport is easiest for small ions, and protons (H+) or lithium ions (Li+) are used in most electrochromic devices. Transparent liquid electrolytes as well as ion-containing thin oxide films were employed in early studies on electrochromics [2], but polymer electrolytes became of interest subsequently [3, 4] and paralleled the developments in electrical battery technology.

Figure 1.1 Generic five-layer electrochromic device design. Arrows indicate movement of ions in an applied electric field. From Ref. [1].

The ions are moved in the electrochromic device when an electrical field is applied between two transparent electrical conductors, as illustrated in Figure 1.1. The required voltage is only of the order of 1 V DC, so powering is, in general, easy and can be achieved by photovoltaics [5]. In small devices, the voltage can be applied directly to the transparent conductors but large devices – such as ‘smart’ windows for buildings – require ‘bus bars’, that is, a metallic frame partly or fully around the circumference of the transparent conducting thin film in order to achieve a uniform current distribution and thereby sufficiently fast and uniform colouring and bleaching. The transparent substrates are often of flat glass, but polymers such as polyethylene terephthalate (PET) or polycarbonate can also be used. The permeation of gas and humidity through foils may or may not be an issue for devices; barrier layers can be applied if needed [6].

An electrochromic device contains three principally different kinds of layered materials: The electrolyte is a pure ion conductor and separates the two electrochromic films (or separates one electrochromic film from an optically passive ion storage film). The electrochromic films conduct both ions and electrons and hence belong to the class of mixed conductors. The transparent conductors, finally, are pure electron conductors. Optical absorption occurs when electrons move into the electrochromic film(s) from the transparent conductors along with charge-balancing ions entering from the electrolyte. This very simplified explanation of the operating principles for an electrochromic device emphasises that it can be described as an ‘electrical thin-film battery’ with a charging state that translates to a degree of optical absorption. This analogy has been pointed out a number of times but has only rarely been taken full advantage of for electrochromics.

Electrochromic devices have a number of characteristic properties that are of much interest for applications. Thus, they exhibit open circuit memory, just as electrical batteries do, and can maintain their optical properties and electrical charge for extended periods of time without drawing energy (depending on the quality of electrical insulation of the electrolyte). The optical absorption can be tuned and set at any level between states with minimum and maximum absorption. The optical changes are slow and have typical time constants from seconds to tens of minutes, depending on physical dimensions, which means that the optical changes can occur on a timescale comparable with the eyes' ability to light-adapt. Furthermore, the optical properties are based on processes on an atomic scale, so electrochromic windows can be without visible haze; this latter property has been documented in detailed spectrally resolved measurements of scattered light [7]. By combining two different electrochromic films in one device, one can adjust the optical transmittance and reach better colour neutrality than with a single electrochromic film. Finally, the electrolyte can be functionalised, provided it is a solid and adhesive bulk-like polymer, so that the smart window combines its optical performance with spall shielding, burglar protection, acoustic damping, near-infrared damping and perhaps even more features.

This chapter is organised as follows: Section 1.2 serves as a background and gives some notes on early work on electrochromic materials and devices. Section 1.3 provides an in-depth discussion of EC materials and covers optical and electronic effects and, specifically, charge transfer in tungsten oxide. It also treats ionic effects with foci on the inherent nanoporosity in electrochromic oxides and on possibilities to augment the porosity by choosing appropriate thin-film deposition parameters. A number of concrete examples on the importance of the deposition conditions are reported, and Section 1.3 ends with a discussion of the electrochromic properties of tungsten–nickel oxide films across the full compositional range. Section 1.4 surveys properties of transparent conducting electrode materials as well as transparent electrolytes. Section 1.5 gives a background to electrochromic devices, specifically delineating a number of hurdles for practical device manufacturing as well as principles for some large-area devices. Conclusions are given in Section 1.6.

1.2 Some Notes on History and Early Applications

Electrochromism in thin films of metal oxides seems to have been discovered several times through independent work. A vivid description of electrically induced colour changes in thin films of tungsten oxide immersed in sulfuric acid was given in an internal document at the Balzers AG in Liechtenstein in 1953 (cited in a book on inorganic electrochromic materials [1]). Later work on W oxide films by Deb at the American Cyanamid Corporation during the 1960s yielded analogous results, which were reported in two seminal papers in 1969 and 1973 [8, 9]; these publications are widely seen as the starting point for research and development of electrochromic devices. Deb's early work was discussed in some detail much later [2, 10]. Another very important electrochromic oxide is Ni oxide, whose usefulness became clear in the mid-1980s [11, 12]. Parallel developments took place in the Soviet Union, and a paper from 1974 [13] quotes ‘USSR Author’s Certificates' and patents by Malyuk et al. dating back to 1963; this work dealt with Nb oxide films.

Early research on electrochromic materials and devices in the United States, Soviet Union, Japan and Europe was motivated by potential applications in information displays, and there were strong research efforts during the first half of the 1970s at several large companies. Generally speaking, these efforts became of less relevance towards the end of the 1970s, and liquid-crystal-based constructions then started to dominate the market for small displays. Electrochromic-based variable-transmission glass was of interest in the context of cathode ray tubes for some time [14]. At present, there is a strong resurgence in the interest in electrochromic-based display-oriented devices (such as ‘electronic paper’), and much research and development is devoted to organics-based full-colour electrochromic displays with excellent viewing properties, cheap printable electrochromic ‘labels’ and, very recently, to ‘active’ authentication devices [15].

Electrochromic-based rear-view mirrors for cars and trucks provide another applications area, and it appears that research and development goes back to the late 1970s [16]. However, the market for ‘active’ rear-view mirrors was largely taken by a device design that is not identical to the one discussed here [17].

Oxide-based electrochromics came into the limelight during the first half of the 1980s, when it became widely accepted that this technology can be of great significance for energy-efficient fenestration [18, 19]. The term ‘smart’ window (alternatively ‘intelligent’ or ‘switching’ window) was coined in 1984/1985 [20, 21] and got immediate attention from researchers as well as from media and the general public.

Electrochromic eyewear has attracted interest on and off for decades. Photochromic glass or plastic has been widely used in sunglasses and goggles, but photochromic devices have clear limitations: their coloration relies on ultraviolet solar irradiation, so they are not of much use inside buildings and vehicles, and the coloration and bleaching dynamics are undesirably slow especially for bleaching at low temperatures. Electrochromic-based eyewear does not have these drawbacks, which motivates a lingering interest. Studies have been reported for sun goggles [22] and helmet visors [23].

A different type of application regards electrochromic-controlled thermal emittance for thermal control of spacecraft. Their exposure to cold space or solar irradiation can yield temperature differences between −50 and +100 °C, which may lead to unreliability in electronic components, detectors and so on. Mechanical shutters can be used for emittance control but tend to be impractical. Investigations on electrochromic-based thermal management have been conducted at least since the early 1990s and remain active today [24–26]. Related devices can be employed for military camouflage in the infrared [27].

The aforementioned applications are not the only ones of interest for electrochromic devices, but the possibility of achieving colour-changing systems based on inorganic oxides or organic materials seems to serve as a catalyst for human inventiveness, and recently mentioned potential uses for electrochromics include fingerprint enhancement in authentication devices and forensics [28] as well as fashion spandex (Lycra — a polyurethane–polyurea copolymer) [29].

1.3 Overview of Electrochromic Oxides

Electrochromic materials and associated devices have been researched continuously ever since the discovery of electrochromism. Looking specifically at electrochromic metal oxides, the literature up to 2007 has been covered in some detail in several prior publications [1, 30–33], and it was noted that some 50–100 scientific and technical papers were published per year. Section 1.3.1 gives a rather superficial presentation of work on electrochromic oxide films presented from 2007. From the sheer number of publications, one may conclude that the research field is thriving and that a large amount of work is devoted to materials based on oxides of W and Ni, as has been the case for the last two decades. Section 1.3.2 discusses optical and electronic properties in electrochromic oxides from a general perspective, and Section 1.3.3 treats these properties in detail for the most widely used electrochromic material, that is, W oxide. Section 1.3.4 discusses ion conduction in electrochromic oxides, again with focus on W oxide. The composition of the electrochromic oxides and their deposition parameters are very important, and some examples from recent studies are given in Section 1.3.5. Finally, Section 1.3.6 reports data on electrochromic properties for films in the full compositional range of W–Ni oxide.

It should be pointed out that much can be inferred about the electrochromic oxides also from studies on alternative applications such as gas sensors and gasochromics, electrical batteries, photocatalysts and so on [34], so the actual knowledge basis for the electrochromic oxides is much larger than the one outlined in this chapter.

1.3.1 Recent Work on Electrochromic Oxide Thin Films

There are two principally different kinds of electrochromic oxides: ‘cathodic’ ones colouring under ion insertion and ‘anodic’ ones colouring under ion extraction. Figure 1.2 indicates metals capable of forming oxides of these two categories and also points out that oxides based on vanadium are intermediate [1]. This scheme serves as a natural starting point for a survey of the recent scientific literature on electrochromic oxides. In the following, we first treat the cathodic oxides, then the anodic oxides and finally the intermediate one.

Figure 1.2 Periodic system of the elements (except the lanthanides and actinides). Differently shaded boxes indicate transition metals with oxides capable of giving cathodic and anodic electrochromism. From Ref. [1].

Tungsten oxide remains the most widely studied electrochromic oxide, and films of this cathodically colouring material have been prepared by a huge number of techniques, including traditional thin-film preparation with physical and chemical vapour deposition, a plethora of chemical methods, electrochemical methods and others. Surveys of these techniques can be found in a number of books and review papers [35–40]. Nanoparticles have sometimes been used as intermediate steps for the films, and substrate templating has been employed to improve the electrochromic performance.

Regarding physical vapour deposition, data have been reported on W oxide films prepared by thermal evaporation, sputtering and pulsed laser deposition. Other work has used chemical vapour deposition and related spray pyrolysis, and a large number of investigations have been based on chemical routes for making films. Furthermore, electrodeposition, anodisation and electrophoretic deposition have been employed. A useful review has been published on properties, synthesis and applications of nanostructured W oxide [41].

Mixed oxides based on tungsten can exhibit properties that are superior, in one way or another, to those of the pure oxide. One well-studied option is W–Ti oxide, where the addition of titanium leads to significantly enhanced durability under electrochemical cycling, as has been known for many years [42, 43]. Recent work on electrochromism of W–Ti oxide has been reported for films made by sputtering, spray pyrolysis, chemical techniques, electrodeposition and anodisation. A key result is that addition of Ti stabilises a highly disordered structure [44, 45], as also reported in other work on W–Ti oxide that was not specifically on electrochromism [46]. Additional research on mixed oxides has been carried out on films of W–MW oxide with MW being Li, C, N, V, Ni, Nb, Mo, Ru, Sn and Ta. Furthermore, investigations have been reported on W oxide containing coinage metal nanoparticles capable of giving plasmon-induced optical absorption of visible light, specifically for nanoparticles of Ag, Pt and Au; using, instead, nanoparticles of indium-tin oxide (ITO) makes it possible to have plasmon absorption at infrared wavelengths. Finally, a number of hybrid nanomaterials have been studied, such as WO3–PEDOT:PSS, where the second component is poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate).

Electrochromic Mo oxide has many similarities to W oxide, and studies have been performed on films prepared by evaporation, chemical vapour deposition, wet chemical techniques and electrodeposition. Mixed oxides have been researched for Mo–MMo oxide with MMo being C, Ti, V, Nb and Ce.

Concerning Ti oxide, electrochromic properties have been reported for films made by sputtering, chemical vapour deposition and spray pyrolysis, various wet chemical techniques, doctor blading and anodisation. This oxide has also been used as anchoring agent for organic chromophores (such as viologens, see Chapter 3) and for inorganic compounds (such as ‘Prussian Blue’, see Chapter 2). Work has been published for films of Ti–MTi with MTi being V and Zr. Finally, studies have been reported on electrochromism of Nb oxide and Nb–Mo oxide.

Regarding anodically colouring electrochromic oxides, we first note work on Ir oxide prepared by sputter deposition and by sol–gel technology, and Ir–Sn oxide prepared by evaporation and Ir–Ta oxide by sputtering. However, iridium oxide is too expensive for most applications, even if diluted with less costly Sn or Ta. Ni oxide is a good alternative whose electrochromic activity was discovered many years ago, and Ni-oxide-based films are included in several of today's electrochromic devices. Recent work has been reported on Ni oxide films prepared by thermal evaporation, sputter deposition, chemical vapour deposition and spray pyrolysis, various chemical techniques and electrodeposition. Furthermore, Ni oxide pigments have been deposited from water dispersion. Mixed binary oxides of Ni–MNi have been studied for MNi being Li, C, N, F, Al, Ti, V, Mn, Co, Cu and W. One reason why additions to Ni oxide are of interest is that the luminous transmittance can be enhanced [47]. Further improvements are possible with ternary or more complex oxides with MNi being (Li,W) [48], (Li,Al) [49], (Li,Zr) [50] and LiPON [51]. Films of the latter materials were produced recently by sputter deposition and are of particular interest for practical applications in future electrochromic-based fenestration. Hybrid films have been made of Ni oxide and one of several polymers or graphene oxide.

V-pentoxide-based films with intermediate electrochromic properties have been prepared by vacuum evaporation, sputter deposition, spray pyrolysis, chemical techniques, electrodeposition and inkjet printing. Mixed oxides based on V–MV have been investigated with MV being C, Na, Ti, Mo, Ag, Ta and W. Mixed V–Ti oxides can serve as good counter electrodes in electrochromic devices [52, 53].

1.3.2 Optical and Electronic Effects

Most oxide-based electrochromic devices employ two electrochromic films, and it is clearly advantageous to combine one cathodic oxide (e.g. based on W, Mo, Ti or Nb) and another anodic oxide (e.g. based on Ni or Ir). Applying a voltage in order to transport ions and electrons between the two electrochromic films in one direction makes both of these films colour, and transporting ions and electrons in the other direction makes both of them bleach. By combining cathodic and anodic oxides, one can accomplish a rather neutral visual appearance. The most commonly used oxides are based on tungsten and nickel, which exhibit cathodic and anodic electrochromism, respectively, according to the highly schematic reactions [1]

1.1 equation

and

1.2

for the case of proton insertion/extraction. Electrons are denoted e−.

We now consider the origin of the electrochromic properties in these oxides and first look at their crystalline nature. The pertinent structures fall into three categories: (defect) perovskites, rutiles and layer/block configurations. These structures can be treated within a unified framework with ‘ubiquitous’ MO6 octahedra (where M denotes metal) connected via joint corners and/or joint edges [1, 54]. Edge-sharing is associated with some degree of octahedral deformation. Only two oxides are problematical within this description: the first one is vanadium pentoxide (V2O5), whose crystal structure can be constructed from heavily distorted VO6 octahedra or, alternatively, from square pyramidal VO5 units [55, 56]; and the second example is hydrous nickel oxide – which is the actual electrochromic material rather than pure NiO [57, 58] – which is thought to contain layers of edge-sharing NiO6 octahedra.

Octahedral coordination is essential for the electronic properties of electrochromic oxides [1, 54]. It was shown in detail many years ago [59] that oxygen 2p bands are separated from metal d levels, and octahedral symmetry leads to splitting of these latter levels into bands with the conventional designations eg and t2g. Figure 1.3 illustrates three cases of importance for the electrochromic oxides. Left-hand panel, for HxWO3, indicates that the O2p band is separated from the split d band by an energy gap. Pure WO3 has a full O2p band and an empty d band, and the band gap is wide enough to render films of this material transparent. Inserting small ions and charge-balancing electrons – according to the aforementioned ‘battery’ model – leads to a partial filling of the d band along with optical absorption as discussed later. The middle panel in Figure 1.3 is adequate for anodically colouring electrochromic oxides. The pure oxides have unoccupied t2g states, and insertion of ions and electrons may fill these states to the top of the band, so the material exhibits a gap between the eg and t2g levels. The material then becomes transparent, assuming that the band gap is large enough. Finally, the right-hand part of Figure 1.3 indicates that V2O5 – with both cathodic and anodic features in its electrochromism – has a principally different electronic structure. The deviation from octahedral coordination is significant enough that the d band displays a narrow split-off part in the band gap. Insertion of ions and electrons into V2O5 may fill this narrow band so that the optical band gap is widened. These features of the band structure can account for electrochromism of V2O5 [60], at least in principle, as well as band gap widening during photo-injection of hydrogen into this material [61].

Figure 1.3 Schematic band structures for different types of electrochromic oxides, as discussed in the main text. Shaded regions signify filled states and E denotes energy. From Ref. [1].

The detailed mechanism for optical absorption in the electrochromic oxides is often poorly understood. The exception may be W oxide, which is considered in the following section, whose properties were reviewed some years ago [33]. However, optical absorption in electrochromic oxides is generally believed to be connected with charge transfer, and polaron absorption accounts for at least most of the significant features [33, 62–65]. A simplified, but closely related, model for the absorption considers intervalence charge transfer transitions [66].

1.3.3 Charge Transfer Absorption in Tungsten Oxide

Electrons inserted together with ions are localised on metal ions and, for the case of tungsten oxide, change some of the W⁶+ sites to W⁵+. Charge transfer between sites i and j can be expressed, schematically, as

1.3 equation

More specifically, the electrons are thought to enter localised states positioned 0.1–0.2 eV below the conduction band and displace the atoms surrounding them so that they form a potential well; strong electron–phonon interaction then leads to the creation of small polarons with a size of 0.5–0.6 nm [33].

Figure 1.4, taken from the work of Berggren et al. [65], reports data on the optical absorption coefficient of sputter-deposited W oxide films electrochemically intercalated with Li+ ions to a number of different levels x (defined as the number of Li+ ions per W atom). It is found that Li+ intercalation yields a broad and asymmetric peak at an energy of ∼1.3 eV. For large intercalation levels, this peak is shifted slightly towards higher energies.

Figure 1.4 Spectral absorption coefficient for W oxide films with different intercalation levels, where x denotes the Li+/W ratio. From Ref. [65].

Figure 1.5 shows data from a comparison of the spectral absorption coefficient at two intercalation levels with Bryksin's theory of polaron absorption [63], which is based on intraband transitions between localised energy levels in a Gaussian density of states [65]. Theory and experiments are found to agree well both for x = 0.04 and x = 0.36, as seen from Figure 1.5(a) and (b), respectively. A comparison with the more recent theory by He [64] does show equally good correspondence, especially not at low intercalation. Successful comparison between Bryksin's theory and experimental optical data was presented several years ago also in other work on W oxide films [67], and further data on polaron absorption in the same material have been given recently [68].

Figure 1.5 Spectral absorption as measured (cf. Figure 1.4) and calculated from two theories of polaron absorption at two intercalation levels, where x denotes Li+/W ratio. From Ref. [65].

The simple model for charge exchange, outlined earlier, is credible only as long as transitions can take place from a state occupied by an electron to another state capable of receiving the electron. If the ion and electron insertion levels are large, this assumption is no longer the case and then ‘site saturation’ [69] becomes important, as considered next.

A detailed investigation of the optical properties at different amounts of lithium insertion into sputter-deposited W oxide films was reported recently by Berggren and Niklasson [70] and Berggren et al. [71], and data are available also for hydrogen-containing material [72]. Figure 1.6a reports spectral optical absorption coefficient α for different levels of Li+ intercalation into a W oxide film. A broad absorption band evolves at an energy of ∼1.5 eV for low values of x, as also observed earlier, while the absorption peak shifts towards higher energies for large amounts of Li. The strong increase of α above 3.5 eV is associated with the fundamental band gap of W oxide. It is interesting to subtract the absorption due to the unintercalated material, and the corresponding absorption coefficients, denoted α+, are given in Figure 1.6b which indicates that the data have unambiguous peak structures. The data could be modelled with three Gaussian peaks, whereas modelling with only two peaks was unsuccessful. Two of the peaks were at the positions shown in Figure 1.6b, and the third peak could be located at intermediate energies. Figure 1.7 indicates the integrated strengths of these peaks.

Figure 1.6 Panel (a) shows spectral absorption coefficient α for a slightly sub-stoichiometric W oxide film intercalated to the shown Li+/W ratios x, and panel (b) shows corresponding absorption α+ when the absorption of the unintercalated film in panel (a) has been subtracted. From Ref. [71].

Figure 1.7 Integrated absorption coefficient as a function of Li+/W ratio x for three Gaussian peaks representing the data in Figure 1.6b. From Ref. [71].

The origin of the Gaussian peaks can be reconciled with ‘site saturation’ by considering three types of sites, namely W⁴+, W⁵+ and W⁶+. Beginning with empty states (x = 0), most of the states will be singly occupied at the start of the intercalation, and electron transitions between empty and singly occupied states will be prevalent. As more single states are filled, the probability that doubly occupied states also will be formed will be higher. Analytical expressions for the number of possible electronic transitions can be given for W⁶+ ↔ W⁵+, W⁵+ ↔ W⁴+ and W⁶+ ↔ W⁴+ by 2x(2 − x)³, 2x³(2 − x) and x²(2 − x)², respectively [71]. Corresponding curves are given in Figure 1.8, and it is clear that there is good similarity with the integrated peak structure in Figure 1.7, which hence indicates that ‘site saturation’ takes place. The intensities of the curves do not coincide, which is not surprising since the absorption strength per transition is likely to be different for the three cases. Considering a practical electrochromic device, long-term cycling durability demands that x is kept low, perhaps not exceeding 0.3–0.35, and then the W⁵+ ↔ W⁶+ transitions are clearly dominating.

Figure 1.8 Relative number of transitions of the types W⁶+ ↔ W⁵+, W⁵+ ↔ W⁴+ and W⁶+ ↔ W⁴+ as a function of Li+/W ratio x. From Ref. [71].

1.3.4 Ionic Effects

Most electrochromic oxides consist of octahedral units in various arrangements, as discussed in detail in Section 1.3.2. These oxides are appropriate both because of their electronic features and because the spaces between the octahedral units are sufficiently large to allow facile transport of small ions. Clusters of octahedra are able to form disordered and more or less loosely packed aggregates with large porosity, and it follows that nanostructures enter at two or more length scales. The following discussion is again focused on W oxide, which has been investigated in great detail. It is not obvious that the anodic electrochromic oxides can be understood on the same premises however, and grain boundaries may then be of large significance, but the situation remains unclear.

Figure 1.9 illustrates nanostructural features of W oxide and shows WO6 octahedra, each with six oxygen atoms surrounding a tungsten atom. Stoichiometric WO3 has a structure wherein each octahedron shares corners with neighbouring octahedra. WO3 and similar transition-metal-based oxides easily form sub-stoichiometric oxides, which include a certain amount of edge-sharing octahedra. The three-dimensional structure comprised by the octahedra yields a three-dimensional ‘tunnels’ structure conducive for ion transport.

Figure 1.9 Schematic image of corner-sharing and edge-sharing octahedra in slightly sub-stoichiometric crystalline W oxide. From Ref. [1].

The crystalline structure in Figure 1.9 is simplified and refers to a cubic structure (cf. Figure 1.10a), and a tetragonal structure usually prevails in WO3 at normal temperature and pressure. The tetragonal structure is more favourable for ion transport than the cubic one since the separations among the octahedral units are larger, as shown in Figure 1.10b. Hexagonal structures, indicated in Figure 1.10c, are easily formed in thin films [1, 73, 74] of W oxide, and the structure is then even better suited for the transport of ions.

Figure 1.10 W oxide with cubic (a), tetragonal (b) and hexagonal (c) structure. Dots signify sites for ion insertion in spaces between the WO6 octahedra. Dashed lines show extents of the unit cells. From Ref. [1].

Figure 1.11 indicates structural data based on modelling of X-ray scattering from films made by evaporation onto substrates at different temperatures, from room temperature to 300 °C [75]. Cluster-type structures are apparent and include hexagonal-type units, which grow in size and interconnect at high substrate temperatures.

Figure 1.11 Structural models for WO6 octahedra in W oxide films prepared by evaporation onto substrates at the shown temperatures (RT denotes room temperature). Arrows in the x and y directions are 2 nm in length. After Ref. [75].

Larger nanostructures than those created by the aggregation of octahedral units can occur as a consequence of the limited mobility of the deposition species, and this mechanism is important for thin films prepared by most techniques. Regarding sputter deposition, the main features are captured in a ‘zone diagram’, commonly referred to as a ‘Thornton diagram’ [76]. It is shown in Figure 1.12. More elaborated or specialised versions of this diagram have been presented in subsequent work. It is evident that low substrate temperature and high pressure in the sputter plasma lead to nanoporous features appropriate for ion transport across the film thickness and hence for electrochromic devices. A direct correlation between sputter gas pressure and electrochromic performance was reported recently [77]. Oblique angle deposition can promote porosity still further, as specifically shown for W-oxide-based films [78–80]. Annealing (cf. Figure 1.11) and ion irradiation [81] are other ways to influence film density.

Figure 1.12 Schematic nanostructures in sputter-deposited thin films formed at different pressures in the sputter plasma and at different substrate temperatures. The melting temperature of the material is denoted Tm. From Ref. [76].

W oxide films can involve molecular deposition species, and it is of interest to consider the lowest energy structures of (WO3)q clusters, which may form the deformable building blocks for octahedra-based nanostructures. Such clusters were evaluated from first-principle calculations in recent work by Sai et al. [82]. Data for 2 ≤ q ≤ 12 are depicted in Figure 1.13, from which it can be inferred that small clusters (with q equal to 3 and 4) have ring-like configurations with alternating W–O arrangements, while larger clusters (with q ≥ 8) can be described as symmetric spherical-like cages. Trimeric W3O9 molecules form during evaporation [83] and correspond to q = 3 in the cluster model; these structures are characterised by a hexagonal configuration. Aqueous solutions, useful for liquid-phase film deposition, can have a preponderance of (W6O19)²− ions, known as ‘Lindqvist anions’ [84].

Figure 1.13 Lowest-energy structures of (WO3)q clusters with 2 ≤ q ≤ 12 and metastable isomers (denoted 5b, 6b and 10b). Crystallographic symmetry designations are indicated at the various clusters. From Ref. [82].

1.3.5 On the Importance of Thin-Film Deposition Parameters

Irrespective of the technique for making EC thin films, the detailed deposition conditions usually play a decisive role. However, each technique is unique to a considerable extent and little can be said in general apart from the guidelines inferred from the previous section.

We first look at films made by physical vapour deposition, which is notable for its possibilities to accomplish process control and reproducibility. A clear illustration of the applicability of the Thornton diagram in Figure 1.12 for electrochromic films was found in recent work on titanium-oxide-based films made by sputtering at various pressures p onto substrates at different temperatures τs [85, 86]. Figure 1.14, reproduced from work by Sorar et al. [85], shows cyclic voltammograms indicating current density under charge insertion and extraction as a function of applied voltage. Clearly, p > 10 mTorr and τs < 100 °C are required to accomplish a structure porous enough for facile ion transport.

Figure 1.14 Cyclic voltammograms for sputter-deposited TiO2-based film in a Li+-conducting electrolyte. Film deposition took place at the pressure p and oxygen/argon ratio γ onto substrates at temperature τs. Data in panels (a) and (b) refer to the effects of varying p and τs, respectively. From Ref. [85].

The composition of the sputter gas is very important, in addition to its pressure, as elucidated in Figure 1.15 for water vapour added to Ar + O2 during sputter deposition of Ni oxide films; the results were presented recently by Green et al. [87]. Addition of water vapour to a pressure of ∼7 × 10−2 Pa increased the current density during ion insertion and extraction and enlarged the optical modulation by allowing a darker state. Other work on sputtering of Ni oxide in the presence of water vapour has been reported elsewhere [88, 89].

Figure 1.15 Current density during ion insertion/extraction (a) and spectral transmittance at maximum and minimum intercalation (b) for sputter-deposited Ni oxide films prepared with the shown partial pressures of H2O in the sputter gas. From Ref. [87].

The properties of the substrate for the electrochromic film may be of large significance. One example of this dependence is reported in Figure 1.16a, taken from work by Yuan et al. [90], which shows optical transmittance in bleached and coloured states for a Ni oxide film prepared by electrodeposition onto a substrate with and without a layer of self-assembled polystyrene nanospheres. Clearly, the sphere templating yields a lower coloured-state transmittance. The visual appearance of films on untreated and templated substrates is shown in Figure 1.16b.

Figure 1.16 Spectral transmittance (a) and visual appearance (b) of an electroplated Ni oxide film on a substrate with and without a template layer consisting of polystyrene nanospheres. The film is in its fully coloured and bleached states. From Ref. [90].

The deposition rate is a very important parameter, which often more or less determines the cost for industrial coating production. However, this parameter is not commonly reported in scientific papers. Among high-rate techniques, it is particularly interesting to consider reactive-gas-flow sputtering, which was studied recently by Oka et al. [91] to make W oxide films. Figure 1.17 shows deposition rate, oxygen/tungsten stoichiometry and density for films made by sputtering at two different oxygen flows but under otherwise identical conditions. The deposition rate was found to be as high as ∼4 nm s−1, which is considerably higher than for typical reactive DC magnetron sputtering. The density depended on the oxygen gas flow and could be as low as half of the bulk value for WO3 (7.16 g cm−3). Hence, the film structure is highly porous and suited for electrochromic devices. Not surprisingly, investigations of the electrochromic performance showed good results. Another contemporary high-rate deposition technique, appropriate for W oxide, is high-power impulse magnetron sputtering (known as HiPIMS) [92].

Figure 1.17 Deposition rate, oxygen/tungsten stoichiometry and density of thin films of electrochromic W oxide prepared by reactive-gas-flow sputtering at different oxygen gas flows. The total gas pressure was 70 Pa, and for two of the data sets also 90 Pa. From Ref. [91].

1.3.6 Electrochromism in Films of Mixed Oxide: The W–Ni-Oxide System

Electrochromism has been investigated in many binary, ternary and so on, oxides, as surveyed in Section 1.3.1. Nevertheless, there is a general lack of comprehensive investigations across the full compositional range between the component oxides. An exception is the W–Ni oxide system, denoted NixW1−x oxide, which was studied by Green et al. [93–98] and clearly combines cathodically and anodically colouring oxides. Tungsten-rich sputter-deposited films were found to consist of a mixture of amorphous WO3 and nanocrystalline NiWO4, with equal amounts of W and Ni the structure was dominated by NiWO4, and nickel-rich films were made up of nanocrystalline NiO and NiWO4.

Figure 1.18 shows optical absorption coefficients at a mid-luminous wavelength of 550 nm for fully coloured and bleached films. The electrochromism is seen to be much stronger in W oxide than in Ni oxide. For W-rich films, the absorption coefficient drops as Ni fraction x is increased, except at the composition x ≈ 0.12 where a pronounced peak can be seen in the absorption coefficient. For Ni-rich films, the absorption coefficient rises as x approaches unity. Films in the compositional range 0.3 < x < 0.7 do not display electrochromism. Another important parameter for electrochromic films is coloration efficiency η, defined by η = OD/ΔQ where optical density (OD) is absorption coefficient multiplied by film thickness and ΔQ is inserted/extracted charge density. Figure 1.19 shows data on coloration efficiency and allows easy comparison with Figure 1.18. Obviously, η increases slightly for increasing values of x, except for compositions around x ≈ 0.5 where η is approximately zero.

Figure 1.18 Mid-luminous absorption coefficient as a function of composition for electrochromic thin films of W–Ni oxide. Filled and open symbols denote fully coloured and bleached states, respectively. From Ref. [95].

Figure 1.19 Mid-luminous coloration efficiency as a function of composition for electrochromic thin films of W–Ni oxide. From Ref. [95].

The strong optical absorption at x ≈ 0.12 is interesting for electrochromic device applications, and Figure 1.20 reports spectral coloration efficiency for films of pure W oxide and for NixW1−x oxide films with two values of x. It is evident that good performance of films with x ≈ 0.12 is found in the whole luminous wavelength range, that is, for 400–700 nm.

Figure 1.20 Spectral coloration efficiency for W–Ni oxide films with the shown compositions. From Ref. [94].

1.4 Transparent Electrical Conductors and Electrolytes

An electrochromic device does not only include electrochromic thin films but also incorporates transparent electrical conductors and an electrolyte, as seen in Figure 1.1. The transparent electrical conductors may be the most costly part in the device – especially if they are based on an indium-containing oxide – and clearly deserve attention. They are of critical importance not only in electrochromics but also for thin-film solar cells, light emitting devices and so on. There are several recent reviews covering the field of transparent conductors [32, 99–101], and therefore only a bird's-eye view is given here, though with some attention to very recent work and to transparent conductors suitable for flexible substrates. There are several types of transparent conductors with specific pros and cons; semiconductor-based films are treated in Section 1.4.1, metal-based films in Section 1.4.2 and nanowire-based and other alternatives in Section 1.4.3. Electrolytes for electrochromic devices are surveyed in Section 1.4.4 with foci on thin films and polymer layers.

1.4.1 Transparent Electrical Conductors: Oxide Films

Thin films of heavily doped wide-bandgap conducting oxides are commonly used in electrochromic devices. These materials include In2O3:Sn (ITO), In2O3:Zn, ZnO:Al, ZnO:Ga, ZnO:In, ZnO:Si, ZnO:B, SnO2:F (FTO), SnO2:Sb and TiO2:Nb; their doping levels are typically a few atomic percent. Several of the oxides can combine a resistivity as low as ∼1 × 10−4 Ω cm with excellent luminous transmittance and durability. Films of ITO, ZnO:Al and ZnO:Ga deposited by reactive DC magnetron sputtering onto glass and PET typically have a resistivity of ∼2 × 10−4 and ∼4 × 10−4 Ω cm, respectively. High-quality FTO films are normally made by spray pyrolysis in conjunction with float glass production, and high temperatures are necessary also for TiO2:Nb films. All of these oxides are transparent across most of the solar spectrum, as seen in Figure 1.21 for FTO-coated glass [102]. The oxide-based transparent conductors are very well understood theoretically [103–105], which means that detailed and accurate simulations of optical properties can be made for multilayer configurations such as electrochromic