Review

Prda

Ruben Baa Department ofb Department of

Accepted 15 August 2009Available online 1 October 2009

currently commercial dynamic tintable smart windows has been carried out.

The technologies of electrochromic, gasochromic, liquid crystal and electrophoretic or suspended-

particle devices were examined and compared for dynamic daylight and solar energy control in

buildings. Presently, state-of-the art commercial electrochromic windows seem most promising to

products for colder climates, and to improve furthermore the commercial products in order to control

the indoor climate in a more energy efcient way by reducing both heating and cooling loads.

. . . . . .

. . . 90

. . 90

. . 91

. . . 92

ARTICLE IN PRESS

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/solmat

Solar Energy Materials & Solar Cells

Solar Energy Materials & Solar Cells 94 (2010) 87105E-mail address: bjorn.petter.jelle@sintef.no (B.P. Jelle).0927-0248/$ - see front matter & 2009 Elsevier B.V. All rights reserved.

doi:10.1016/j.solmat.2009.08.021

Corresponding author at: Department of Building Materials and Structures, SINTEF Building and Infrastructure, NO-7465 Trondheim, Norway. Tel.: +4773593377;fax: +4773593380.3.4.1. Tungsten-based electrochromic windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

3.4.2. Non-tungsten-based electrochromic windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

3.4.3. Photovoltaic integrated electrochromic devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

3.4.4. All-solid-state switchable mirrors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 933.3. Polymer electrochromics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.3.1. Polyaniline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.3.2. Poly(3,4-ethylene-dioxythiophene) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.4. All-solid-state electrochromic windows and devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. Transparent conductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

3. Electrochromic windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

3.1. Tungsten oxide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

3.2. Other electrochromic metal oxides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

3.2.1. Nickel oxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

3.2.2. Iridium oxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

3.2.3. Niobium oxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

3.2.4. Other inorganic electrochromics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90Electrochromic window

Gasochromic window

Liquid crystal window

Suspended-particle window

Electrophoretic window

Daylight control

Solar energy control

Contents

1. Introduction . . . . . . . . . . . . . . . .& 2009 Elsevier B.V. All rights reserved.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88Keywords:

Transparent conductor

Smart window

reduce cooling loads, heating loads and lighting energy in buildings, where they have been found most

reliable and able to modulate the transmittance up to 68% of the total solar spectrum. Their efciency

has already been proven in hot Californian climates, but more research is necessary to validate theArticle history:

Received 11 June 2009A survey on prototype andc Department of Civil and Transport Engineering, Norwegian University of Science and Technology (NTNU), NO-7491 Trondheim, Norwayd Department of Architectural Design, History and Technology, Norwegian University of Science and Technology (NTNU), NO-7491 Trondheim, Norway

a r t i c l e i n f o a b s t r a c tetens a,b, Bjrn Petter Jelle a,c,, Arild Gustavsen d

Building Materials and Structures, SINTEF Building and Infrastructure, NO-7465 Trondheim, Norway

Civil Engineering, Catholic University of Leuven (KUL), B-3001 Heverlee, Belgiumylight and solar energy control in buildings: A state-of-the-art review

ies, requirements and possibilities of smart windows for dynamicopert

ARTICLE IN PRESS

3.5. Gasochromic devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

3.5.1. WO3-based gasochromic windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

3.5.2. Gasochromic switchable mirrors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

4. Liquid crystal devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

5. Electrophoretic or suspended-particle devices (SPD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

6. Commercially available smart windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

6.1. Requirements and expectations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

6.2. Transparent conductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

6.3. All-solid-state electrochromic windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

6.3.1. Properties of available all-solid-state electrochromic widows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

6.3.2. Aesthetic and energetic consequences of available all-solid-state electrochromic windows . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

6.4. Gasochromic windows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

. . .

. . .

. . .

. . .

. . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

. . .

. . .

market: Chromic materials, liquid crystals and electrophoretic or

neglected in

mo

A survey is performed onwhat types of smart windows are currently

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 8710588Smart windows are to be judged on several specic factors. Ofst importance is their transmittance modulation range in the

Before we look closer on the properties of the different types ofsmart windows, transparent conductors (TC) have to be treated.- and thermochromic windows, they will be mostlythis review.

2. Transparent conductorssuspended-particle devices. Here, the chromic devices may bedivided in four categories, i.e. electrochromic, gasochromic,photochromic and thermochromic devices, where the last twopossibilities will respond automatically to, respectively, changesin light and temperature. However, because one can not controlthe outdoor weather conditions and hence neither the propertiesof the photo

available on the market and their properties and potential fordaylight and solar energy control in buildings. The aim is to nd outwhich technologies are of most interest for the mentioned buildingapplications and which are most developed today.6.5. Liquid crystal smart windows . . . . . . . . . . . . . . . . . . . . . . . .

6.6. Suspended-particle smart windows . . . . . . . . . . . . . . . . . . .

6.6.1. Properties of available SP windows. . . . . . . . . . . . .

6.6.2. Applicability of available SP windows. . . . . . . . . . .

7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Appendix A. List of transparent conductor manufacturers . . . . . . .

Appendix B. List of smart window manufacturers . . . . . . . . . . . . . .

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction

Windows are often regarded as a less energy efcient buildingcomponent with a larger maintenance requirement. Nevertheless,their technology has grown by leaps over the last several years. Anew class of windows promises to set the technology bar evenhigher. Dynamic tintable or the so-called smart windows, wherean example is shown in Fig. 1, can change properties such as thesolar factor and the transmission of radiation in the solarspectrum in response to an electric current or to the changingenvironmental conditions themselves. The application of suchwindows may lead towards a drastic reduction of the energyconsumption of highly glazed buildings by reducing cooling loads,heating loads and the demand for electric lighting.

Various techniques are known to derive switchable windows.However, taking into account the specic properties of windowglazing in buildings, one strict rule has to be reckoned with: Atransparent mode of the glazing has to be possible. Presently,three different technologies with external triggering signal arecommonly known for this purpose and start to be available on theFig. 1. Switching sequence of an elec. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

visible and whole solar spectrum. The modulation range is oftenexpressed for a single wavelength in literature, but this gives littleor no information on the overall performance of the smart device.Secondly, the expected lifetime and number of achieved cycleswithout or only minor degradation are of uttermost importance.Thirdly, the switching time for colouration and bleaching isimportant, mostly expressed as the time necessary to reach 90%of its maximummodulation range. The magnitude of the switchingtime is strongly connected to the size of the device, as large devicestend to have long switching times. Furthermore, the achievedwindow size, the total energy consumption, the operating voltageand the operating temperature range are of importance.

Within this work, a review is given on the current technology ofthese smart window technologies. Main questions within thisreview are how do different devices obtain their dynamic tintableproperties, which possibilities are available today to adapt theperformance of these devices and how do the different technologiesperform under laboratory conditions? However, there is no aimwithinthis review to give a historical overview on the subject. The realfocus in this study lies within commercial available smart windows.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99trochromic laminated glass [46].

ARTICLE IN PRESS

(ITO), which may be replaced by heavily doped conductors such as

ii.

iii.

The electroactive layers, often denoted as electrochromics,change their optical properties by switching between theiroxidized and reduced form. Electrochromism may be seen as adevice characteristic instead of material property. Most favourableare electrochromics that are reecting in their coloured stateinstead of absorbing, but this has been found very difcult andmost electrochromics are absorbing. By combining different typeof electrochromics, ion-storage lms and ion conductors, differentproperties can be obtained for the device, where the modulationrange, durability and switching speeds can be optimized.

Many of these electrochromics are well-known today [49].Most important are the metal oxides, of which tungsten oxide isthe most well-known, but also electrochromic polymers areapplied in electrochromic windows and devices.

3.1. Tungsten oxide

The electrochromic phenomenon of materials was originallydiscovered in tungsten oxide WO3 thin lms, and remains untilnow the most promising, most studied and most appliedelectrochromic material in EC windows and devices. Electrochro-

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 87105 89A complete overview on the market of ITO and alternativetransparent conductors is out of the scope of this work, but can befound in NanoMarkets [98]. One may conclude that ITO will farfrom disappear because too many products rely on it, but thefuture for high-performing transparent conductors in electronicdevices as well as smart windows may currently lies in CNTs.

3. Electrochromic windows

Electrochromism is the property of a device to change itsoptical properties reversibly if an external potential is applied,associated with ion insertion and extraction processes. Theelectrochromic device mostly consists of several layers. The basisis a glass or plastic covered by a transparent conducting lm, i.e.mostly ITO (see Ch.2), on which one (or multiple) cathodicelectroactive layer(s) are afxed. These are followed by a layer ofion conductor, on its turn followed by an ion-storage lm or one(or multiple) complimentary anodic electroactive layers and

anohas to be sputtered onto a surface in a vacuum.A third newmaterial is 12CaO 7Al2O3, often denoted as C12A7[57], which itself is an electrochromic material and turnsgreen by a process involving H0-H+ +e after incorporationof hydrogen inside the cages. However, even if the C12A7 typeof TCs is very interesting, the current properties are far frompossible applications in electrochromic windows.PEDOT degrades over time if exposed to light or heat.Carbon nanotubes (CNT) layers seem more promising com-pared with PEDOT, but are far from large-area applications[30,135]. Carbon nanotube layers have a high transparency inthe visible and IR spectrum and sheet resistances around104O/& have been measured. The sheet resistance increasesby increasing transmittance. CNT networks are p-type con-ductors, whereas traditional transparent conductors areexclusively n-type. The availability of a p-type transparentconductor could lead to new cell designs that simplifymanufacturing and improve efciency: They are easier andcheaper than ITO to deposit on glass and plastic surfaces, sincethey can be formed into a solution, compared with ITO whichSnO2:F (FTO), ZnO:Al or ZnO:Ga, while many more are available asshown by Granqvist [52]. However, ITO is in short supply. In fact,one predicts that we could run out of indium, a silvery metalproduced as a by-product of zinc mining, in the next 10 years. Theprice of the metal has raised from around US $100/kg to nearly$1000 in the past 6 years. Several new promising TCs related toenergy applications are offered in the last decade:

i. Based on current technology, PEDOT or poly(3,4-ethylene-dioxythiophene) seems the best alternative as TC [54,55,124].The polymer combines a transmittance above 0.90 with anelectrical resistivity currently below 400O/& for commercialproducts. However, long-term stability remains a problem:TCs are a major issue for all electrically activated devices becauseof the need for a high-quality TC and their costs. TCs need a hightransparency in order to let the active part of the smart windowsregulate as much as possible. Furthermore, the TCs need a highelectronic conductivity to provide a low voltage drop along theconductor surface. For the eld effect devices, i.e. liquid crystalsand electrophoretics or dispersed particles (see Chs. 56), theneed for the lowest resistivity for large areas is less than forelectrochromics [77].

The most widely used transparent conductor in all kind ofdevices, e.g. touch screens, is tin-doped indium oxide In2O3(Sn)ther transparent conducting lm.

Fig. 2. Spectral transmittance vs. wavelength for a smart window (i.e. ITO|WO3|-mism of tungsten oxide is a complex phenomenon and is still notyet completely understood, but it may and can be represented bythe simple reaction

(transparent) WO3+xM+ +xe 2 MxWO3 (deep blue) (1)

where M+ can be H+, Li+ , Na+ or K+, 0oxo1 and where e aredenoting electrons. WO3 turns blue, while doping the oxide withmolybdenum Mo provides colour neutrality. Depending on thecrystallinity of the tungsten layer, tungsten oxide obtains itsmodulation due to reectance or absorbance. One example oftransmittance regulation in an electrochromic window incorpor-ating WO3 is shown in Fig. 2.

Extensive reviews on tungsten oxide lms have been writtenpreviously, where Case study on tungsten oxide in Granqvist [48]and its completion with Granqvist [50], and Colouration oftungsten oxide lms: A model for optically active coatings byBange [11] are among the most comprehensive. Also recentpossibilities have been expressed [26,45] and we would like torefer to these works for more detailed information on thecolouration mechanisms of tungsten oxide.LiNbO3|V2O5|In2O3) (redrawn from [47]).

ARTICLE IN PRESS

3.3. Polymer electrochromics

Besides the oxide lms there are also organic lms availablewith EC properties, but most of them show UV degradation andare hence less likely for possible energy-related applications inexterior smart glazing. Many different polymers have beenincorporated in prototype EC devices, e.g. poly- and monomericpyrrole, viologens, 4,40-diaminodiphenyl sulfone, poly(3-me-tylthiophene) or diclofenac, but most interest lately goes towardspolyaniline (PANI) and poly(3,4-ethylene-dioxythiophene) (PED-OT). As for the inorganic electrochromic materials, the electro-chromic polymers do also need to have a transparent state forapplication in smart windows.

3.3.1. Polyaniline

Polyaniline is a conducting polymer which may undergo colourchanges from a transparent state to violet by both a redox processand proton doping. A simplied formula for PANI consisting ofreduced and oxidized units can be written as [(BN(H)BN(H))x(BNQQQN)1x]y with benzenoid and quinoid units (seeFig. 4). This conversion of benzenoid units into quinoid unitsresults in a typical absorption peak around 600700nm. PANI isone of the most extensively researched electrochromic material

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 87105903.2. Other electrochromic metal oxides

Many other electrochromic metal oxides are known besidesWO3 and applied in prototype electrochromic windows [53], e.g.Bi2O3, CeO2, CoO, CuO, FeOOH, Fe2O3, Fe3O4, FeO, MnO2, MoO3,P2O2y, RhO3, RuO2, SnO2, Ta2O5, TiO2 and V2O5, but most interestlately goes towards nickel oxide NiIIO(1y)Hz, iridium dioxide IrO2and niobium pentoxide Nb2O5. Several of these oxides, besidesWO3, have been studied by Lampert [74,76], e.g. various hydratednickel oxide lms [75,146,147].

3.2.1. Nickel oxide

Films based on NiO have enjoyed much interest lately becausethey combine a reasonable cost with excellent electro-chromic properties, which even can be improved by mixing NiOwith wide band gap oxides such as MgO or Al2O3 [46,36,61,90,92,111,138,149,150]. NiO:X, i.e. where X is Mg, Al, Si, V,Zr, Nb, Ag, Ta, Li, Al or B, has been found complementary withWO3:Xin the visible and near IR: pairing them in a complete device resultsin a very dark colour neutral system [95]. The main effect ofelectrochromic effect takes place in the UV and VIS spectra andreaches a very high colouration efciency between 100cm2/C at340nm and 25cm2/C at 800nm. A 200nm layer has been dened bya Tvis of 0.800.10 at an oxygen concentration somewhat below 1.5%.

(transparent) Ni(OH)22 NiOOH+H+ +e (grey) (2)

(transparent) NiOH+Ni(OH)22 Ni2O3+3H+ +3e (brownish) (3)

3.2.2. Iridium oxide

Also lms based on IrO2 and Ir2O3 have enjoyed an increasedinterest lately [810,69,100]. While IrO2-based lms are exces-sively expensive, good electrochromic properties are obtainedafter dilution with the much cheaper Ta2O5.

(transparent) Ir2O3 xH2O2 Ir2O4(x1)H2O+2H+ +2e (brown)(4)

3.2.3. Niobium oxide

The interest in niobium pentoxide Nb2O5 has increased in thelast decade because of its promising electrochromic properties[58,93,96,97,109,110,116]. Pure Nb2O5 and doped Nb2O5:X, i.e.where X is Sn, Zr, Ti, Li, Mo, WO3 or TiO2, layers change colour byinsertion of H+ or Li+ ions from transparent to brown, blue or greydepending on the crystallinity of the layer.

(transparent) Nb2O5+xM+ +xe 2 MxNb2O5 (blue, brown or

grey) (5)

Undoped Nb2O5 has a high transmittance of 0.800.92 in thevisible region for the bleached state (see Fig. 3) and transmittancesbetween 0.10 and 0.30 are obtained in the coloured state, withrelatively slow colouring and bleaching times. The disadvantage ofthe Nb2O5 layers is their small colouration efciency CE of about 1227cm2/C compared with the CE of 3750cm2/C of tungsten oxide.

Lithiated niobium oxide LixNb2Oy lms exhibits a much higherelectrochromic reversibility, where bleaching is accomplishedafter a few seconds, while colouring times remain the same.

3.2.4. Other inorganic electrochromics

Not all inorganic electrochromics are metal oxides. Oneof them is the widely studied Prussian blue (PB), i.e. K3Fe(CN)6,[1,23,59,6368] which colour reaction may be described as

(transparent) M2FeII[Fe(CN)6]2 MFe

III[Fe(CN)6]+M+ +e (blue) (6)

where M+ is a cation, e.g. K+. PB is the prototype of a polynuclear

transition metal hexacyanometallate with the general formulaM0k[M00(CN)6]l (l,k integers), where M0 and M00 are transition metalswith different formal oxidation numbers [127]. The electrochemicalreduction and oxidation of PB can lead to PrussianWhite and PrussianGreen, respectively. A cycle life of 105 cycles has been found insolutions of pH 23, while also depositing a PB lm on a polyanilinecoating yields a superior cycling lifetime compared with a PB lmdeposited directly onto a platinum or ITO substrate. Besides, asymbiotic relationship between polyaniline and PB is found resultingin enhanced colouration [6368].

Transmittance spectra of pure Prussian Blue have been studiedby Jelle et al. [66] with the hole method in a solid state devicecontaining tungsten oxide, polyaniline and PB. Here, the trans-mittance spectra of a single layer EC at various colouration stageshave been measured and calculated by comparing the transmit-tance spectra of different devices with holes in an electrochromiclayer with the reference device without electrochromic layers.

Fig. 3. Transmittance spectra of undoped and Mo- and Li-doped Nb2O5:X solgeldouble layers on K-glass in the coloured and bleached state at 2.2 and +1V(redrawn from [58]).till date, owing to its good electrochemical cycling stability in

ARTICLE IN PRESS

non-aqueous electrolytes above 106 cycles, its low cost and ease ofprocessing by electrodeposition or liquid casting techniques. Highcolour contrast is usually observed for thick lms (41mm) of PANIbut it is achieved at the expense of switching speed [27].

Recently, a new class of hybrid polymers has been developed atthe Nanyang Technical University of Singapore [141,142,151] basedon aniline, with a 40% enhancement in electrochromic contrast attheir lmax compared with PANI (see Fig. 5) because of moreaccessible doping sites and with increased electrochemicalstability.

The polymer polyaniline-tethered polyhedral oligomeric sil-sesquioxane (POSS-PANI) was synthesized through copolymeriza-tion of aniline with octa(aminophanyl) silsesquioxane (OAPS) inthe presence of dodecylbenzene sulphonic acid (DBSA) or poly(4-styrene sulfonic acid) (PSS) as dopant, with a feed molar ratio ofOAPS to aniline of 0.5/99.5. A loosely-packed structure is formedbecause of covalent binding of the PANI chains to the POSSnanocages, creating a nano-porous structure and allowing easyion movement during redox switching, proven by an increasedionic conductivity by one order of magnitude and a decrease of upto 2 orders of magnitude compared with PANI. When switchedfrom 2.0 to +2.0V, 0.5% POSS-PANI|PSS has a total change inabsorbance DA of 0.66 at lmax compared with 0.44 for PANI/PSS.The coloration time of POSS-PANI|PSS has been found approxi-mately equal to that of PANI/PSS, but the bleaching times havebeen found slightly shorter.

Fig. 4. A simplied formula for PANI consisting of reduced and oxidized units withbenzenoid (B) and quinoid (Q) units that may be written as [(BN(H)BN(H))x(BN=Q=N)1x]y (redrawn from [136]).

Fig. 5. UVvis absorbance spectra of the complementary EC device PET|ITO|PA-NI|Electrolyte|WO3|ITO|PET and PET|ITO|POSS-PANI|Electrolyte|WO3|ITO|PET

switched at 2.0 and 2.0V. (redrawn from [151]).

Table 1Data for WO3-based electrochromic devices found in literature showing materials, samp

the switching time for colouration and bleaching tc/b.

Size (cm2) T ()

WO3-based construction

G|ITO|WO3|PVB:LiClO4|CeTiOx|ITO|G 100 Tsol 0.7

Tvis 0.8

WO3|NiO-based construction

G|ITO|WO3|PMMA-PC-Li+|NiO|ITO|G T600 0.7

G|ITO|WO3|ZrP xH2O|ZrO2|NiO|ITO|G 25 Tvis 0.7Tsol 0.5

G|SnO2|WO3|PVDF-Li+|NiO:Li|SnO2|G 7 Tvis 0.7

G|ITO|WO3|Ta2O5|NiO|ITO|G 2400 Tvis 0.7

Tsol 0.5

G|SnO2|WO3|PEO/PEGMA:Li+|NiO:Li|SnO2|G 144 Tvis 0.7

G/P|ITO|WO3|ZrP|ZrO2|NiO|ITO|G/P 20 Tvis 0.7

P|ITO|WO3|PMMA-PPG-Li+|NiO|ITO|P 220 T550 0.7

G|FTO|WO3|P-3|NiO|FTO|G 12.5 Tvis 0.5

WO3|IrO2-based construction

G/P|ITO|WO3|Ta2O5|IrO2|ITO 30 T550 0.7

WO3|Polymer-based construction

G|ITO|WO3|PAMPS|PANI|ITO|G 2 Tsol 0.7

G|ITO|WO3|PAMPS|PB|PANI|ITO|G 2 Tsol 0.7

G|ITO|WO3|PAMPS|PB|PANI|ITO|G (another version) 2.8 Tsol 0.6

G|SnO2|WO3|PVSA:PVP:H+|PB|SnO2|G 155 T550 0.7

G|ITO|WO3|PAMPS:L|PANI-CSA|AR|ZnSe|AR R1200 0

P|ITO|WO3:H2O|PVDF-HFP-Li+|PANI|ITO|P T800 0.1

G denotes glass, P denotes polymer and the subscript for T signies the wavelength in

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 87105 913.3.2. Poly(3,4-ethylene-dioxythiophene)

Electrochromic applications based on p-type polymers havealso drawn a lot of attention due to their ease of colouring, highelectrochromic contrast and fast response times, of whichpoly(3,4-ethylene-dioxythiophene) (PEDOT) and its derivates aremost researched. PEDOT switches from blue in the neutral stateto transparent in the oxidized state, but has a rather weakelectrochromic contrast. As far as known by the authors, amaximum transmittance change DT610 of 0.65 has beenmeasured [72].

le size, modulation range, the performed number of colouring/bleaching cycles and

cycles tc/b (s) Reference

30.045 16103 60 Schlotter et al. [120]10.07

80.31 104 Lechner and Thomas [80]

40.38 60 Azens et al. [7]

30.25 Karlsson and Roos [70]

50.02 Michalak et al. [95]

30.18 105 Nagai et al. [152]

50.11

00.27 120 Penissi et al. [112]

50.14 103 180/60 Kullman et al. [73]

00.35 5103 200 Granqvist et al. [51]80.06 103 2 Zelazowska [150]

00.18 3.5104 OBrien et al. [101]

40.35 11/11 Jelle and Hagen [64]

30.23 4103 34/23 Jelle and Hagen [64]40.08 300/100 Jelle and Hagen [65]

20.06 2104 30 Ho [59].650.22 900 9 Topart and Hourquebie [134]

20.02 Marcel and Tarascon [94]nm for which the values have been obtained.

ARTICLE IN PRESS

3.4. All-solid-state electrochromic windows and devices

Many of the properties obtained for single electrochromicmaterials are obtained by analysing the potentiastatic response ofa thin lm while immersed in an electrolyte. By combiningdifferent types of these electrochromics, ion-storage lms and ionconductors, different properties can be obtained for the device,where the modulation range, durability and switching speeds canbe optimized. However, a viable electrochromic window (ECW) ismore than the sum of its components. It represents these partsassembled and working together.

For future applications, one must focus on large-area windowsand cyclability, two factors that are mostly neglected in scienticliterature.

A division in electrochromics has been made as organic andinorganic materials, but this division is not possible for ECWsbecause many use both organic and inorganic electrochromicmaterials within the same device. More appropriate here is tomake a division between devices containing tungsten oxide anddevices which do not. Tungsten oxide WO3 is widely known for itsgood electrochromic properties and stability resulting in manytungsten-based prototype and commercial EC windows anddevices, while tungsten-free electrochromic devices are ratherrare.

3.4.1. Tungsten-based electrochromic windows

Tungsten oxide WO3 is used in a wide range of electrochromicdevices. A selection is given on WO3-based electrochromicwindows and their properties in Table 1 based on the previousmentioned electrochromics, but many more devices can be foundin literature.

So far, the largest modulation ranges found in literatureare from WO3-based devices, but modulations ranges higher than0.50 in Tsol or Tvis are rare. A transmittance modulation of0.56 is achieved by Jelle and Hagen [65] for a WO3|PB|PANIdevice for the complete solar spectrum, while a range of 0.55 forthe visible spectrum has been achieved by Nagai et al. [152]for a relatively large WO3|NiO device of 40 60 cm2. However,these values are widely surpassed by the values retrievedby Schlotter et al. [120] with a single active but thicker WO3layer resulting in a modulation range of 0.74 and 0.68 forTvis and Tsol, respectively (e.g. Fig. 6). Degradation occurs rather

A selection is given on non-tungsten electrochromic windowsand their properties in Table 2 based on the previous mentioned

built by the National Renewable Energy Laboratory of Golden, USA.The development of a side-by-side PV-powered ECW [17,18]

sam

(

550

T0550

570

630

630

TvisT56R80

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 8710592Fig. 6. Spectral transmittance of a WO3-based electrochromic window (redrawnfrom [120]).

Table 2Data for non-tungsten electrochromic devices found in literature showing materials,

for colouration and bleaching tc/b.

Size (cm2) T

Nb2O5-based construction

G|SnO2|Nb2O5:Li|Ormolyte-Li+|SnO2:Sb:Mo|SnO2|G 9 T

G|FTO|Nb2O5:Li|PC:LiClO4|(CeO2)x(TiO2)1x|TFO|G 50 DG|FTO|Nb2O5:Mo|PC:LiClO4|(CeO2)x(TiO2)1x|TFO|G 1 200 T

Polymer-based construction

G|ITO|PEDOT|PC|PANI|ITO|G 4 T

P|ITO|PEDOT:PSS|PEO-PC-LiClO4|PANI:CSA|ITO|P 4 T

P|ITO|PEDOT:PSS|PEO-PC-LiClO4|PANI:PSS|ITO|P T

P|ITO|PEDOT:DMF|IL-BF4 |PEDOT:DMF|ITO|P 4 DP|ITO|PEDOT:CVD|salt electrolyte|ITO|PET 1 DP|PANI|Au|PES|PVOH|PET|Au|P Dresulted in the fabrication of a monolithic a-SiC:H PV-poweredECWof 16 cm2 [25,26,34,35] and monolithic dye-TiO2 PV-poweredECWs up to 25 cm2 [16,56,113]. Here, the main concerns for future

ple size, modulation range, the performed number of cycles and the switching time

) cycles tc/b (s) Reference

0.700.20 120 Orel et al. [106]

1000 0.14 2103 Heusing et al. [58]0.600.25 55103 180 Heusing et al. [58]

0.580.14 2104 30 Lin and Ho [153]0.640.22 15103 30 Huang et al. [60]0.650.25

0.51 3/4 Pozo-Gonzalo et al. [115]

6 0.48 150 13/8.5 Lock et al. [91]

01200 0.24 Li et al. [89]electrochromics, but many more devices can be found inliterature.

3.4.3. Photovoltaic integrated electrochromic devices

EC windows with no external wiring are most desirable inthe building industry. Here, an integrated photovoltaic-powered window is an obvious choice, particularly because PVand EC technology have compatible operational characteristics(see Table 3).

Self-powered photovoltaic electrochromic devices have beenfast in this last model, but the resulting modulation ranges after16,000 cycles remain very high compared with otherelectrochromic devices.

3.4.2. Non-tungsten-based electrochromic windows

Compared with the amount of published tungsten-based smartwindows, tungstenless devices are rare. Most of them concern ofniobiumoxide-based electrochromic device and all-polymer de-vices. However, only small devices have been found and thecyclability is rather poor for the polymer devices compared withearlier stated tungsten devices.

ARTICLE IN PRESS

Table 3Data for photovoltaic integrated electrochromic devices found in literature showing m

switching time for colouration and bleaching tc/b.

Photovoltaic EC construction Size (cm2) T (

G|ITO|WO3|TiO2|IC|TCO| T78G|TCO|WO3:Li|LiAlF4|V2O3|TCO|a-SiC:Npin|TCO|G

G|WO3:Li|LiAlF4|V2O5|a-SiC:Hpin|TCO|G T67TsoT50

odu

T (

TvisTsolRsolTvisTvisTvis

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 87105 93large-area applications are the possible loss of the energygenerated by the PV device for larger dimensions, a small rangeof optical modulation and rather low transmittances in the clearstate.

A wide band gap a-Si1xCx:H nip photovoltaic cell isemployed [25,26,34,35] as a semitransparent power supply. ThePV cell has a transmittance of 0.80 over most of the visible lightand maintains a 1-sun1 open-circuit voltage Voc of 0.92V andshort-circuit current Jsc of 2mA/cm

2 for a thickness of 60nm.Whereas a transparent conductor, i.e. SnO2, separated the PV andthe electrochromic device (ECD) in the side-by-side solution, the

G|TCO|WO3|TiO2|LiI:I2:PC|Pt|TCO|G 25

G|ITO|TiO2xNx|NiO|NaHCO3/NaOH|Pt 6

Table 4Data for gasochromic devices found in literature showing materials, sample size, m

and the switching time for colouration and bleaching.

Gasochromic device construction Size (cm2)

G|WO3|Pt|H2/Ar|G 6600

G|H2/Ar|Pd|ITO|Pd|ZrO2|Pd|Zr/ZrO2|MnNiMg|ITO|G

G/P|Pd:W-PTA|Pt|H2

G/P|Pd:W-PTA/ICS-PPG|Pt|H2

G/P|Pt:WO3|Pt|H2 low-voltage ECD is deposited directly on top of the PC device inthe monolithic solution, and consists of lithium based tungstenoxide and vanadium oxide as counter electrode, i.e. LiyWO3|-LiAlF4|V2O5. By properly controlling the thickness of each layer thecolouring and bleaching voltages have been adjusted for compat-ibility with the PV device. Colouration occurred within therange of 0.6 to 1.3V and bleaching within 0.10.6V, withcolouring and bleaching times of approximately 2min. Theabsence of a middle conductor in the monolithic PVECD requiresthat both the colouring and bleaching current ows through thePV and EC device all the time. Therefore, the 2V bleachingvoltage includes the part required to overcome the built-inpotential from the PV of approximately 0.8V. The ECD itself hasa transmittance of 0.630.20 at 670nm at 1.0 and +1.0V, butmuch lower values Tsol of 0.250.08 are obtained for the completePVEC device.

The use of a dye-TiO2 photovoltaic cell [16,56,113] resulted inthe same low value Tsol of 0.260.02 [51] for a simple LiyWO3|dye-TiO2 device. The PV device has a Voc of 0.88V, making it possible

1 Light intensity Ilight is generally expressed in suns, where 1 sun is 1 kW/m2.

As a result, the 1-sun open-circuit voltage VOC represents the maximum voltage (at

zero current) from a solar cell for a light irradiance of 1 sun or 1kW/m2 [71]. Here,

the mentioned VOC of 0.9V is rather high. Silicon solar cells on high quality single

crystalline material have a 1-sun VOC up to 730mV, while commercial devices

typically have a 1-sun VOC around 600mV.to bleach and colour the EC device in about 2min. Compared withthis low transmittance, much better transmittances T500 of 0.870.15 have been obtained by combining a TiO2xNx with nickeloxide as electrochromic layer [61].

3.4.4. All-solid-state switchable mirrors

Recently, all-solid-state switchable mirrors have been achievedat the National Institute of Advanced Industrial Science andTechnology AIST (Japan) based on both the switchable propertiesof Mg4Ni and the electrochromic properties of HxWO3[14,133]. Similar gasochromic switchable mirrors have been

aterials, sample size, modulation range, the performed number of cycles and the

) Cycles tc/b (s) Reference

8 0.700.54 120 Bechinger and Gregg [16]

Deb et al. [25]

0 0.600.10 Deb et al. [25]

l 0.260.02 120 Hauch et al. [56]

0 0.870.15 1h Huang et al. [61]

lation range for the transmittance or reectance, the performed number of cycles

) Cycles tc/b (s) Reference

0.770.06 20,000 20 Wittwer et al. [140]

0.760.05

0.800.15 1 000 Bao et al. [12,13]

0.740.09 50/120 Orel et al. [105]

0.860.05 2000.640.09 500/1000developed, as stated later-on in Ch.3.5, based on both MgNix andMnNiMg.

The device is a multilayer Mg4Ni|Pd|Al|Ta2O5|HxWO3|ITO on aPET sheet. For the switching property, the protons in HxWO3 aretransported to the layer of Mg4Ni by applying voltage whereafterboth layers will turn to their transparent state, while Mg4Ni ishighly reecting in the metallic state. The resulting Tvis are 0.470.01 and switching occurred within 1030 s when a voltage of 5Vwas applied, depending on the ITO sheet resistance.

3.5. Gasochromic devices

The principle behind gasochromic windows is similar to that ofsolid-state gasochromic windows. An electrochromic is switchedbetween a bleached and coloured state by hydrogen gas H2instead of applying a voltage. However, not all electrochromicmaterials can be coloured with hydrogen gas. The gasochromicdevices are claimed to be simple and inexpensive, because only asingle electrochromic layer is sufcient, and transparent electri-cally conducting layers are no longer necessary. The transmittancemodulation of the gasochromic devices exceed these of mostsolid-state EC windows (see Table 4).

Gasochromic switching is an option for window applications,but it requires well controlled gas exchange processes. The effecthas been studied in devices based on NiO, MoO3 [102,144] andV2O5 [123], but the best results for smart windows have beenobtained for tungsten-based gasochromic devices. Usually, a thin

ARTICLE IN PRESS

mechanism of gasochromic colouration is not completely yetknown, but could be summarized as follows [103]:

(transparent) WO3+2H2-WO3+4H- H2WO3+2H-WO2+H2O(coloured)

(7)

(coloured) 2WO2+O2- 2WO3 (transparent) (8)

Both the optical density and the rate of coloration can be variedby the choice of lm thickness and/or gas concentration, but thedependence is weak. For testing, the gases were supplied frompressurized bottles, but generation of the gases by electrolysis isan obvious option and can be integrated into a fac-ade withoutaffecting the rest of the building.

Units with an area of 0.6 1.1m2 have been produced andtested over 20,000 cycles during three years without suffering anyobvious damage. Transmittances Tsol of 0.760.05 and Tvis of 0.77

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 8710594catalytic layer of Pt or Pd is incorporated to facilitate thegasochromic effect, but the oxides remain electrically non-conducting.

3.5.1. WO3-based gasochromic windows

Gasochromic windows have been introduced based on thechromic properties of WO3 [3945,103,104,121,139,140,143] andpalladium doped WO3 [105,107,148] as stated previously in Ch.3.1.Due to its high redox potential, it is possible to colour WO3 withhydrogen gas H2 which is not possible for most other electro-chromic materials.

The optically active material is a sputtered, highly porous,columnar lm of (doped) WO3 with a typical thickness around400nm, coated by a 15nm thin catalytic layer of platinum Pt.Exposed to a low concentration of H2 in a carrier gas, it coloursblue. A more neutral colour, i.e. greyblue, can be obtained byusing a mixture of WO3 with molybdenum oxide WxMoyO3, butthis will result in inferior transmittances. On exposure to O2, thelayer bleaches back to its original state (see Fig. 7). The exact

0.06 have been achieved. The transmittance is reduced to 10% ofits original value within 20 s during the colouring process and isreturned to 95% of this value within a minute by admitting 5%

2

added as counter electrodes. The nal device is H /Ar|Pd|ITO|Pd|Z-Fig. 7. [upper] Mechanism of colouration by H2 in porous columnar WO3.Dissociation of H2 into 2H occurs on the upper surface catalysed by Pt, whereas

the reaction leading to formation of colour centres occurs on the internal pore

surfaces, after diffusions by the protons along the pores in the presence of water

and mechanism of bleaching by O2 (redrawn from [139]). [bottom] Transmittance

for the coloured and bleached state of a gasochromic double-glazed unit with a

WO3 layer of 500nm and a hydrogen concentration of 1% (redrawn from [139]).2

rO2|Pd|Zr/ZrO2|MnNiMg|ITO for which the second Zr/ZrO2 layerhas been added to minimize diffusion of Pd into the MnNiMg layer

Fig. 8. Typical reectance spectra of a MgNi0.15|Pd and a MnNiMg|Zn|Pdoxygen to the system. Colouring could be accelerated byincreasing the lm thickness and/or gas concentration.

3.5.2. Gasochromic switchable mirrors

A new possibility besides WO3-based gasochromic windowshas been brought up recently, resulting in the so-called gaso-chromic mirrors. These devices are based on the same principle: AMgNix alloy [12,15,139,145], MgTi [13] or MnNiMg [3,125] with apalladium Pd catalyst will turn transparent by taking uphydrogen. Similar all-solid-state switchable mirrors have beendeveloped recently [14,133] as stated earlier in Ch.3.4.4 based onboth Mg4Ni and HxWO3.

The gasochromic device developed by Anders et al. [3] uses thecavity between both glass panes for the hydrogen storage (i.e.0.01/0.99 H2/Ar) and for which, as a result, gas transport is nolonger necessary. In order to facilitate ion transport to and fromthe MnNiMg layer, a ZrO ionic conductor and an ITO-layer aregasochromic device (redrawn from [139] and [3]).

ARTICLE IN PRESS

between 80% and 95% performwell (see Fig. 8), while an increased

beznoylocy} benzene, which is known for its possibility to allowand keep a good homeotropic alignment of liquid crystals on

Frontiers Inc. (NY, USA) and their licensees [19]. SPDs consist of 35 layers of which the active layer has adsorbing dipole needle-

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 87105 95switching speed and a decreased upper end of the transmittancerange is noticed with decreasing Mg content.

The stability of these devices still needs to be improved, i.e.typical degradation includes slowing down of the kinetics, relatedto Pd catalyst mobility. Degradation occurs quickly after 100150cycles [3], but up to 1000 cycles have been achieved [12,13] bycoating the metal layer with polyvinylacetate PVAc and celluloseacetate CA. Simultaneously, the transmittance in the non-metallicstate increased with 0.1 for wavelengths above 750nm bycoating with PVAc.

4. Liquid crystal devicesand to protect the layer from oxidation. Films with a Mg content

Fig. 9. Normalnormal transmittance spectra of polymer-dispersed liquid crystallaminated glass for an applied voltage of 50VAc at 60Hz in the ON state (redrawn

from [108]).Switchable devices based on liquid crystals (LC) offer anotherapproach besides electrochromic and gasochromic devices. Themechanism of optical switching is a change in the orientation ofliquid crystal molecules between two conductive electrodes byapplying an electric eld, resulting in a change of theirtransmittance [28,32]. Liquid crystals come in six different types,i.e. nematic, smectic, twisted nematic, cholesteric (ChLCs), guest-host and ferroelectric, but mainly guest-host liquid crystals areapplied in commercial switchable windows since the 1990s, ofwhich polymer-dispersed liquid crystals (PDLC) and encapsulatedliquid crystals (NCAP) are the most common [7779]. Large-areaPDLC windows are already commercially available, i.e. in sizes upto 1.0 2.8m2 by Saint-Gobain glass, and operate between 24 and120V. Recent research on commercial products show goodresults: A 20mm commercial PDLC layer by DM Display Co. Ltd.(Korea) has been used between ITO coated polyester lms toachieve a 20 20 cm2 device with Tuv 0.0180.001, Tvis 0.620.061 and Tnir 0.3840.129 (see Fig. 9), while the haze coefcientchanges from 0.09 to 0.90 [108]. The device exhibits goodtemperature stability between 0 and 60 1C, while a device of 125 350cm2 has been developed and found to be stable for 3million times switching at 100VAc.

However, the devices require continuous power resulting in apower consumption of 5 up to 20W/m2, while also long-term UVstability and high cost remain issues.

the last decades, many glass and coating manufacturers broughtsmart glass on the market for building and automotive purposes,based on EC, LC and SPD technologies. In order to control daylight

and solar energy in buildings, it is important to know how currentcommercially available smart windows perform, which propertiesthey exhibit and how reliable the products are. Answering thesequestions is the main aim in this work.

Before discussing the available products on the market today, itis important to look for which properties are actually desired fordaylight and solar energy control in buildings.

6.1. Requirements and expectations

The required properties for smart windows for solar andenergy applications have been expressed by Lampert [76] (seeshaped or spherical particles, i.e. mostly polyhalide, suspended inan organic uid or gel between two transparent conductors. Theparticles are random and light absorbing in the off state, but willalign by application of an electric eld causing an increase intransmittance. SPD devices have typical transmission ranges of0.790.49 and 0.500.04, a switching time of 100200ms andrequire 65220V AC to operate.

Due to the patents on the technology, only little informationhas been found on recent developments of suspended particledevices [137].

6. Commercially available smart windows

The technology of electrochromism, liquid crystal switchingand electrophoretic switching was discovered and made publiclyin the 1970s and the 1980s, but the progress has been slow. Theglass industry has been trying for some decades, but up torecently no smart windows made it to the market. However, themarket of smart windows is changing due to extensive energyregulations for buildings and the need for alternative solutions. InITO-covered substrates after polymerization. The values areachieved after colouring and bleaching times of, respectively, 2and 0.5min.

Similar to the EC devices, one has developed a liquid crystaldevice including an a-Si1xCx:H nip photovoltaic cell, but thedevice turns out to be relatively dark with transmittances below0.31 [24].

5. Electrophoretic or suspended-particle devices (SPD)

SP devices are similar to LC devices in different ways: Bothtechnologies were initially developed for displays meaning thatthey are relatively fast compared with other technologies andboth rely on the electronic eld to align the active elements to letlight though undisturbed.

Electrophoretic of suspended particle (SP) windows is a lm-based patented technology developed and licensed by ResearchComparable values for the transmittance have been achieved[37] for organosiloxane liquid crystals, but more remarkable is T0.800.01 at lmax (unknown) achieved [22] by using theliquid crystalline monomer 1,4-bis{4-[6-(acryloyloxy)hexyloxy]-Table 5) and surveys have recently been performed [128132] of

ARTICLE IN PRESS

view preservation and furnishing protection, become more and

Unidym (CA, USA) expresses the possibility as large-area trans-

possible exterior building glazing, which are truly based onelectrochromics. An overview is given on the properties of EC

Table 5Requirements for electrochromic windows in the bleached (bl.) and coloured (col.) sta

Solar transmission () Solar reection () S

S

1

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 8710596more important. In general, energy efcient operation and highdurability are desired, while operation using alternating currentvoltage is preferred above direct current voltage. Regarding theperformance of the smart windows, the most desired propertiesare (i) integration with other coatings, e.g. low-e, (ii) glarereduction, (iii) consistent-looking tint changes regardless ofwindow size, (iv) light control to any point between the darkand clear transparent state, (v) a high blockage of UV light, and(vi) fast switching speeds. It is the professionals opinion that amaximum size of 3 2m2 is desired and a median of 500 $/m2 isexpressed as a maximal cost for both commercial and residentialprojects.

Although the awareness of smart windows is moderatelystrong on the commercial market, the knowledge of speciccharacteristics of switchable glass has been found limited. Lack ofknowledge regarding the product category, uncertainties aboutthe actual service life of smart windows and the perception thatthe material costs associated with switchable windows areexcessively high have been found to be the primary inhibitingfactors why architects have not yet applied smart windows.However, the outlook for smart windows is strong. It wasprojected [33] that the dollar value of the smart window demandin the United States will reach $ 1.34109 by 2015, i.e. a 250%increase compared with 2005.

6.2. Transparent conductors

The most-used transparent conductor ITO is widely commer-cially available, but also the poly(3,4-ethylene-dioxythiophene)PEDOT and carbon nanotube CNT alternatives (see Ch.2) are yetlimited available on the market.

Both Asahi glass (Yokohama, Japan) and prazisions glas & optikUnited States architects, LEED2 accredited professionals andwindow manufacturers on the subject of switchable windowsby Research Frontiers Inc.

The attitudes towards smart glass are strongly positive.Leading drivers are the potential for greater energy savings, ademand for sustainable building solutions and the need for lowerlifetime operating costs of buildings, while also aesthetics, e.g.

Tsol Tvis Tnir Rnir

bl. col. bl. col. bl. col. bl. col.

0.500.70 0.100.20 0.500.70 r0.100.20 0.100.20 Z0.70(Germany) have shown the capability to produce a large-area ITOglass of 1m2 with a sheet resistance of 1O/& or lower and lowhaze. On the other hand, Diamond Coatings (United Kingdom) hasshown the ability to produce commercial highly transparent ITO-coatings with transmittances up to 0.95 in the visible spectrumfor a sheet resistance of 1020O/&.

Transparent conductive PEDOT or PEDOT-based polymers havebeen recently introduced on the market by H.C. Starck (Germany)

2 LEED denotes Leadership in Energy and Environmental Design, where LEED

Accredited Professionals indicates that they have passed the accreditation exam

given by the Green Building Certication Institute of the United States Green

Building Council (USGBC).windows on the market today and conclusions on the applic-ability of the commercial windows for daylight and solar energycontrol are stated. Besides these three companies which will betreated extensively, other smaller ones have to be mentioned: AlsoSaint Gobain Sekurit (France), ChromoGenics (Sweden) and Gentex(MI, USA) [20,38,118] produce electrochromic glass on small scale,mainly for automotive applications, but nothing or little is knownon their properties. Further information about various manufac-turers of commercially available electrochromic windows anddevices is found in Appendix B.

6.3.1. Properties of available all-solid-state electrochromic widows

The rst company that provides EC windows on the markettoday is SAGE Electrochromics (NY, USA) [117] which usesparent conductive layer for large-area energy and solar applica-tions. However, no properties on transmittances andconductivities are expressed.

Because PEDOT and CNTs are already limited available on themarket and their high potential for future applications, they seemthe best basis for new sustainable smart windows. However, aslong as their long-term stability have not yet been proven,commercially available smart windows keep on using thestandard transparent conducting oxides as ITO and FTO. Addi-tional information about various manufacturers of transparentconductors is found in Appendix A.

6.3. All-solid-state electrochromic windows

Although many manufacturers claim to have electrochromicsmart windows, only three companies, i.e. SAGE Electrochromics,EControl-Glas and Gesimat [31,46,117] have been found by theauthors of this state-of-the-art review producing windows forand Agfa-Gevaert (Belgium). The coatings have a transmittancearound 0.80, but still a relatively high sheet resistance of 150500O/&. Also CNTs are yet commercially available, but only

te [76].

witching voltage (V) Memory (h) Cycling lifetime Operating temperature

mall Large 112 4104106 cycles520 years

Unprotected

3 1024 3070 1Cabsorptive tungsten oxide as the electrochromic material. SAGEhas been recognized worldwide for its electrochromic technologyand has been referred to by many in the American constructioncommunity as the only one suitable for the building windowindustry. So far, the company has the rst and only commerciallyavailable electrochromic windows for exterior applications whichpassed ASTM E-2141-06, the standard American test methods forassessing the durability of absorptive electrochromic coatings onsealed insulating glass units.

The second company is EControl-Glas (Germany) [31]. Thiscompany has taken over the electrochromic activities of PilkingtonAG (Suisse) and provides electrochromic glass for exterior buildingapplications according to EN ISO 12543-4, the European standard

ARTICLE IN PRESS

for working in symbiosis with photovoltaic devices, but

iii.

iv.

Dat ound

blea 2 K [

T

SAGE Electrochromics, Inc. (WO ) 108150 1.65 00

0

0

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 87105 973

EControl-Glas GmbH and Co. KG (WO3) 120220 1.10.5

Gesimat GmbH (WO3 and CE) 80120 Matesappsinwin

(Gesligionsheandbemeelelayinfobui

in T

avareq

i.

ii.

Fig.Elecle 6a for commercially available electrochromic windows for building applications f

ching cycles, the solar factor SF and the U-value of the entire window in W/m

nufacturer Size (cm2) UTabt methods for the durability of laminated glass for buildinglications. According to literature, both manufacturers use agle WO3-layer for the electrochromic properties of theirdows.A third manufacturer of EC glass has been found in Gesimatrmany) [46], which assembles EC windows conforming ahtly different process. A rst difference may be found in the-conducting layer: An ion-conductive polyvinylbutyral (PVB)et is used, which is commonly used as a layer in safety glasseasily processed within the industry. A second difference mayfound in its electrochromic layers. Where the previouslyntioned manufacturers use a single tungsten oxide layer for itsctrochromic properties, Gesimat applies two complementaryers, i.e. WO3 and an unknown anodic EC. However, normation is found on the applicability of their products forlding envelopes and their maximum sizes are smaller.Properties of these commercially available windows are givenable 6.We may now compare the properties of the commerciallyilable EC windows (Table 6) with the previously stateduirements or desired properties (Table 5):

Both SAGE and EControl-Glas show the capability of producinglarge-area windows of, respectively, 1.6 and 2.6m2, but can notreach the previously expressed desired dimension of 2 3m2.All EC windows operate at 5V DC and have a low powerconsumption of 0.5Wh/m2 or lower, making them suitable

v.

TchrtheactelecapmacomprociaconprostaEC

10. Transmittance spectra of commercial products of Gesimat and of SAGEtrochromics (redrawn from [46,119]).1.65W/m K, whereas a value of 1.1 W/m K has becomestandard in new European building projects.literature with a DTsol of 0.68 and a DTvis of 0.74, while theelectrochromic properties of the Gesimat product has highermodulation ranges due to its two active EC layers.EControl-Glas achieves low U-values of 1.1 and 0.5W/m2K byusing an argon-lled cavity or triple glazing. On the otherhand, SAGE only provides windows with a thermal U-value of

2 2both SAGE and EControl-Glas serve a 10-year guarantee on theirwindows.Eight different types of electrochromic windows from SAGE,EControl-glas and Gesimat are so far found to be available onthe market, of which a summary of the best windows isexpressed in Table 6. Compared with the desired properties asexpressed in Table 5, a few conclusions can be made. Firstly,the transmittance in the visible spectrum for the bestperforming windows of SAGE and Gesimat are satisfying therequirements (see Fig. 10), while the other products (of SAGEand Gesimat) are rather dark with a transmittance between0.48 and 0.35 in the bleached state. No Tvis values are found forEControl-Glas. Secondly, the same can be said about thetransmittance in the complete solar spectrum: Thetransmittance for the coloured state are low enough withvalues between 0.15 and 0.01, but the transmittance for thebleached state does not exceed 0.52 for all available windows.A possible reason here is the incorporation of a low-e coating.In overall, the electrochromic properties of the standardavailable windows of both SAGE and EControl-glas seemcomparable: All have a range DTsol between 0.35 and 0.40which is far from the best-performing EC windows found inmore difcult to integrate in the electric network of abuilding.A lifetime of 105 cycles and 30 years has been expressedwithin the range of 30 to 60 1C by SAGE, which conformsthe desired properties by Lampert [76]. No durabilityproperties are found for the two other manufacturers, but, showing maximum size, modulation range, the guaranteed number of colouring/

31,46,117].

sol () Tvis () SF Cycles

.400.015 0.620.035 0.480.09 105

.500.15 0.360.12 10-year guarantee

.450.14 0.300.10

.520.06 0.750.08 10-year guaranteeo determine which manufacturer provides the best electro-omic windows is not possible because this strongly depends ongoals stated for implementing electrochromic windows in theual building. SAGE Electrochromics is widely known for itsctrochromic windows, while EControl-Glas has shown theability of combining the EC properties with low U-values andnufacturing larger windows and Gesimat has shown thepetence of manufacturing highly contrasting windows. Im-vements on scale and higher transmittances for the commer-l windows in the bleached state are desired, but the overallclusion may be clear: One has shown the possibility toduce large-area electrochromic windows according to thendards for exterior building glazing with a good durability andproperties.

ARTICLE IN PRESS

sta

eleBer

[15

oun2K [

TabDat oun

blea 2 K [

M

Ty ss c

A

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 8710598performances for realistic ofce conditions. The conclusions ofthe eld study can be recapitulated as follows:

i. Switching from fully coloured to fully bleached takes 67minat temperatures above 10 1C for windows of 0.46 0.89m2,while switching times between 40 and 85min have beennoted at colder temperatures and at low solar irradiance.These values are high, especially for low temperatures, andfaster switching speeds are desired. Also, the switching rangedecreased over the 2.5-year installed period.

ii. Controlling the commercial EC window wall to maximizepubsimiii.

Ared.2. Aesthetic and energetic consequences of available all-solid-

te electrochromic windows

A eld study has been performed on large-area commercialctrochromic windows of SAGE (Table 6) by the Lawrencekeley National Laboratory (CA, USA) [8188]; Carmody et al.4] Clear et al. [155]) of which the nal results have beenlished by Lee et al. [86]. These eld tests, together withulations, allowed the quantication of the EC window6.3le 8a for commercially available SP-based smart windows for building applications f

ching cycles, the solar factor SF and the U-value of the entire window in W/m

anufacturer

pical values, e.g. SmartGlass, American glass products, pleotint and innovative gla

merican glass products companyTable 7Data for commercially available LC-based smart windows for building applications f

bleaching cycles, the solar factor SF and the U-value of the entire window in W/m

Manufacturer Size (cm2)

Typical values, e.g. SmartGlass, DreamGlass and SGG. 100280Nippon Sheet Glass Co., Ltd. 180275Innovative Glass Corporation 260488daylight and energy efciency resulted in a high average dailylighting energy saving of 44% compared with a reference casewith fully lowered blinds and without daylight-controlledelectric lighting. However, the values are much lower, i.e. 26%and 10% if well-tuned daylighting control is applied asreference case and if the EC window is controlled to meetvisual comfort, respectively. Controlling the commercial ECwindow wall for glare hardly reduced the average daily coolingloads due to solar heat gains, i.e. 8% and 3%, respectively, for areference window without and with fully lowered blinds. Evenmore, using a split-fac-ade EC window raised the averagecooling loads due to solar heat gains by 11% compared with ashaded reference case.The EC system provides potentially large savings in electricdemand for perimeter zones onwarm days in hot climates: thewindows strongly reduced peak cooling loads, i.e. by 26% and19%, respectively, for a reference window without and withfully lowered blinds. However, much lower impacts will benoticed in cooler climates, in built-up urban areas (castingshadows) or for small windows.

ccording to the above studies, achieving maximum energyuction via electrochromic windows requires controlling the6.4. Gasochromic windows

The transmittance modulation of prototype gasochromicwindows exceeds these of most all-solid-state electrochromicwindows. A pilot production plant for windows up to 1.5 2.0m2has been built by The Fraunhofer Institute for Solar EnergySystems and Interpane (Germany) where transmittances Tsol of0.760.05 and Tvis of 0.770.06 have been achieved,window and lighting systems in order to minimize lighting energyrather than to reduce solar heat gains, primarily because thecoefcient of performance of conventional heating, ventilatingand air conditioning (HVAC) systems shows little daily variation.However, the estimated energy impacts are highly dependent onthe assumed reference case and the actual EC window perfor-mance properties, and more long-term studies are needed on thehuman behaviour factor.d, showing maximum size, modulation range, the guaranteed number of colouring/

29,62,99,122,126].

U Tsol () Tvis () SF Cycles

DT 0.02 DT 0.02 106

0.690.12

1.6 0.770.56

d, showing maximum size, modulation range, the guaranteed number of colouring/

2,62,114,126].

Size (cm2) U Tsol () Tvis () SF Cycles

orporation. 100280 0.500.04 0.350.18 0.500.30 106110 () 0.790.49 105but no commercially available gasochromic windows have beenfound.

6.5. Liquid crystal smart windows

Liquid crystal devices are commercial widely available forlarge-area applications. An overview is given in Appendix Bon the most-known manufacturers of LC windows (seealso Table 7), but many more manufacturers are on the markettoday.

Commercial LC windows have been found stable for tempera-tures between 0 and 60 1C and large-area windows of up to 12m2

are on the market, but long-term UV stability remains an issue.However, the resulting modulation ranges seem less favourablefor energy-related applications in building envelopes. Averagemodulation ranges are 0.02 for both DTvis and DTsol. The mainmodulation by switching LC windows is obtained in hazing thelight transmission. The haze coefcient switches from about 0.90to 0.08, but the haze coefcient stays relatively high in thetransparent mode (i.e. less clear). Two manufacturers, i.e. NipponSheet Glass Corporation (Japan) and Innovative Glass Corporation(NY, USA), have higher modulation ranges DTvis of 0.57 and 0.21,respectively.

ARTICLE IN PRESS

app

about various manufacturers of commercially availablesuspended particle-based smart windows and devices is foundin Appendix B.

with a Tvis of 0.790.49, but the company mainly providesautomotive glazing and no exterior building glazing productshave been found.

Electrochromic and electrophoretic or suspended-particlewindows seem highly promising for dynamic daylight and solarenergy applications in buildings based on the achieved transmit-

modulation range in the visible spectrum is much higher forelectrochromic windows, though. On the other hand, the

switching.Currently, based on this literature survey, electrochromic

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 87105 99i. The device could only be cycled for less then 1000 switchesbefore breakdown, due to stresses caused by abrupt changes inthe applied voltage, which is signicantly poorer than expectedby the information given by the company.

ii. The optical direct transmittance in the clear state is poor, i.e.below 0.25.

Conclusion of the study stated that future research shouldfocus on new materials and manufacturing processes to improve6.6.2. Applicability of available SP windows

A commercial 0.28 0.22m2 suspended-particle window ofCricursa Cristales Curvados (Spain) [21] has been tested on theapplicability as smart window for building purposes [137]. Here,the response times have been found much shorter, i.e. between 2and 3 s, compared with the longer switching times in EC windows.However, two major concerns are expressed regarding possiblewide-spread building applications:6.6.1. Properties of available SP windows

Many commercial SP windows are available today fromdifferent manufacturers with sizes up to 2.6 4.9m2, but allhave a similar transmittance modulation with a Tvis of 0.500.04and stability for more than 106 cycles between 20 and 65 1C (seee.g. Table 8). Darker windows are available, but it is questionedwhether this is favourable with the already low maximumtransmittance of 0.50. Similar to LC windows, all available SPwindows operate between 65 and 220V AC, where a constantpower is necessary to maintain the clear state, resulting in apower consumption of 1.9 up to 16W/m2 in contrast toelectrochromic windows where power only is necessary duringswitching.

One exception to the relatively dark windows may be noticed:American Glass Products Company (TN, USA) [2] provides windows6.6. Suspended-particle smart windows

The market on suspended-particle (SP) based windows ispatented technology and completely licensed by Research FrontiersInc. (NY, USA) and their licensees. An overview is given on theproperties of SP windows on the market today and conclusions onthe applicability of the commercial windows for daylightand solar energy control are stated. Additional informationsolcycitching.As a result of this, liquid crystal devices are commonly onlylied in buildings for aesthetic or privacy applications instead ofar energy control.eleswAll available LC windows operate between 65 and 230V AC,but constant power is necessary in the clear state resulting in apower consumption of 3.5 up to 15.5W/m2 in contrast to

ctrochromic windows where power only is necessary duringlability and obtain better transmittance ranges.windows seem to be the most promising state-of-the-arttechnology for daylight and solar energy purposes. The reliabilityof the current commercially available windows has been proven,their properties are within expectations and room for improve-ments has been demonstrated in literature. The windows havebeen found to be able to reduce up to 26% of lighting energycompared with well-tuned daylighting control by blinds, andaround 20% of the peak cooling loads in hot climates as California(USA). However, little is known about their efciency in colder,e.g. Nordic, climates.

Gasochromic windows are recently being developed andshow promising results. Due to its simple device structureand the absence of transparent conductors, very high transmit-tance modulation ranges compared with the short researchperiod have been achieved. This may also mean that futurecommercial gasochromic windows may become an economicallyattractive high performance alternative for current smartwindow technologies. However, negative aspects such as the useof gas and a limited available number of cycles must bementioned.

Acknowledgements

This work was supported by the Research Council of Norway,AF Gruppen, Glava, Hunton Fiber as, Icopal, Isola, Jackon, maxit,Moelven ByggModul, Rambll, Skanska, Statsbygg and Takprodu-sentenes forskningsgruppe through the SINTEF/NTNU researchproject Robust Envelope Construction Details for Buildings of the21st Century (ROBUST).

Appendix A. List of transparent conductor manufacturers

See Table A1.

Appendix B. List of smart window manufacturerstransmittance modulation has been found poor for commercialliquid crystal windows. In addition, the liquid crystal windowshave been found instable for UV radiation and as a resultinappropriate for long-term exterior building applications. Liquidcrystal and suspended-particle windows share the same dis-advantages: Both need an electric eld to be maintained as long asthe transparent mode of the glass is required, resulting in ahigher energy consumption compared with electrochromic win-dows which normally only require an electric eld duringtance modulation ranges. The transmittances in the solarspectrum, the guaranteed number of cycles and the maximumwindow sizes are similar for the commercial products of bothtechnologies. The maximum transmittance as well as the7. ConclusionsSee Table A2.

ARTICLE IN PRESS

Table A1

Manufacturer Illustration Product Tvis Tsol Tuv Sheet resistance Size Further Information

Transparent conducting ITO

prazisions glas & optik, Gmbh,Im Langen Busch 14, D-58640

Iserlohn, GERMANY

CEC-series 0.70 o5 up to10,000O/&

- ITO-coated glass

Tel.: +492371776790;

fax: +4923717767970

info@pgo-online.com, www.

pgo-online.com

Diamonds Coatings Limited,Unit 2a, Harvey Works Ind. State,

Shelah Road, Halesowen, UK

DIAMOX ITO Tmax 0.790.89 1050O/& 100100 cm2 - ITO-coatings

Tel.: +448451630603; fax:

+44845 1360604

DIAMOX+ITO Tmax 0.930.85

www.diamondcoatings.co.uk

AGC Asahi Glass co.Ltd,Strawinskylaan 1525, 1077

Amsterdam, NETHERLANDS

Down to 1O/& - ITO-coated glass

Tel.: +31205736040;

fax: +31205753191

www.agc.co.jp

Transparent conducting PEDOT

H.C. Starck GmbH, Im Schleeke7891, 38642 Goslar, GERMANNY

CleviosTM P

grades

40.80 - - Down to 150O/&

- - PEDOT coating

Tel.: +4953217510; fax:

+4953217516192

www.clevios.com

Agfa-Gevaert, Septestraat 27, B-2640Mortsel, BELGIUM

OrgaconTM - - - o200O/& - - PEDOT coating

Tel.: +3234442111; fax:

+3234444485

Info.orgacon@agfa.com, www.agfa.

com

Transparent conducting carbon nanotubes

NANOCS, 244 Fifth Avenue, #2949,10001 NY, USA

ITO-series 40.85 8 up to 100O/& - ITO

Tel.: +18889088803; fax:

+19175912212

CNT-series - Carbon nanotubes

CNT

sales@nanocs.com, www.nanocs.

com

NCS-series - Nano Gold Coating

Chengdu Organic Chemicals Co.Ltd.

SWCNT EC 4100S/cm - Single-wall Carbonnanotubes NTC

Tel.: +86288523676585228839;

fax: +86288521506985223978

carbon@cioc.ac.cn, www.

timesnano.com

Unidym, Inc., 1430 OBrien Drive,Suite G, Menlo Park, CA 94025, USA

HIPCO s Single-

Wall Carbon

Nanotube

large-area - Carbon nanotubes

CNT

Tel.: +16504621935; fax:

+16504621939

www.unidym.com

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 87105100

ARTIC

LEIN

PRESS

Table A2

Manufacturer Illustration Product Uwindow(W/m2K)

Tvis Tsol Tuv SF Durability Max size Electricaldemand

Further Information

Commercially available electrochromic windows and devices

SAGE Electrochromics, Inc.,One Sage Way, Faribault, MN 55021,USA

Classic TM 1.65 0.620.035 0.400.015 0.0560.008 0.480.09 100 000cycles, 30yrs, 30 to60 1C

Up to108150cm2

5 VDc Electrochromicwindows for buildingapplicationsSwitch-ing time of 35min10years warrantyOnlycommerciallyavailable smartwindows for exteriorapplications whichpassed ASTM E-2141-06WO3-based

1.65 0.620.035 0.400.015 0.0040.001 0.480.09Tel.: +1 877 7243321;fax: +15073330145 See Green

TM 1.65 0.480.028 0.190.01 0.040.006 0.440.09

sales@sage-ec.com,http://www.sage-ec.com/ Cool View Blue

TM1.65 0.400.023 0.300.01 0.0000.000 0.460.09

Clear-as-Day TM 1.65 0.350.019 0.310.01 0.0000.000 0.460.09

EControl-Glas GmbH & Co. KG,Glaserstr.1, 93437 Furth im Wald,GERMANY

EControls

Double Glass1.1 0.500.15 0.050.005 0.360.12 120220

cm25 VDco0.5Wh/m2

Electrochromicwindows for buildingapplicationsAccord-ing to DIN EN ISO12543-4 for exteriorinsulating glass-WO3-based

Tel.: +499973858 330;fax: +499973858331

EControls TripleGlass

0.5 0.450.14 0.020.003 0.300.10

info@econtrol-glas.de,www.econtrol-glas.de

GESIMAT GmbH,Koepenicker Str. 325, 12555 Berlin,GERMANY

0.750.08 0.520.06 12080 cm2 0.5/2.2 V DC,0.04Wh/m2

Electrochromicwindow based on ECand active counter-ECWO3+active CETel.: +493047389 251;

fax: +4947389252www.gesimat.de

ChromoGenics AB,Marstagatan4, SE-75323, Uppsala,SWEDEN

Electrochromicwindow based on WOand NiO

Tel.: +4618 43004 30info@chromogenics.be,www.chromgenics.se

Saint Gobain Sekurit,Viktoriaallee 3-5, 52066 Aachen,GERMANY

SGSElectrochromicGlass

0.400.04 0.200.02 2590 1C 1.5V DCo0.1Wh/m2

Electrochromicwindow forautomotiveapplicationsTel.: +49241947 2604

contactSekurit@Saint-Gobain.com,www.sekurit.com

IP Glass Technology B.V.,159 Groenendaal, 3011 SRRotterdam, NETHERLANDS

SPD Glass Electrochromic Glassin cooperation withSAGE Electrochromics

Tel.: +311021367 52;fax: +31102131709

ECD Glass

info@intraprojects.com,www.intraprojects.com

LCD Glass

GENTEX Corporation,600 North Centennial St., Zeeland,49464 MI, USA

Gentex Auto-Dimming

Electrochromicmirrors forautomotiveapplications.Tel.: +1 616 7721800;

fax: +16167727348www.gentex.com

AJJER LLC.,4541E Fort Lowell Road, Tucson, AZ85715, USA

ElectrochromixInc

Electrochromicautomotive mirrors &building glazing

Tel.: +1 5203217680x28;fax: +15203210030

aagrawal@gwestofce.net,www.ajjer.com

R.Baeten

set

al./SolarEnerg

yMateria

ls&

SolarCells

94(2010

)87105

101

ARTIC

LEIN

PRESS

Table A2 (continued )

Manufacturer Illustration Product Uwindow(W/m2K)

Tvis Tsol Tuv SF Durability ze Electricaldemand

Further Information

Commercially available liquid crystal-based smart windows and devices

SmartGlass International, Ltd.,(See SPD-based smart windows)

LC SmartGlass TM 1.3 0.650.58 0.510.49 0.060.06 3106cycles, 12yrs, 20 to50 1C

29065110 Vac,o200mA/m2,5W/m2

Polymer-dispersedliquid crystal device(PDLC), Diffuse lightwhen switched on

www.SmartGlassinternational.com

DreamGlass,C/ Canada, 15. 28860 Paracuellos delJarama, Madrid, SPAIN

DreamGlass V1s 0.750.71(hazed)

1050 1C 00 6080 VAc,10 W/m2

Liquid crystal device(LC)

Tel.: +349165842 45;fax: +34 916581451

DreamGlass V2 s 0.790.75(hazed)

065 1C 00 100120VAc, 8 W/m2

www.dreamglass.es

Saint-Gobain Glass,Les Miroirs 18, avenue dAlsace,92400 Courbevoie, FRANCE

Priva-Lite 55.2(11mm)

5.65.8 0.770.76 0.630.64 2060 1C 80 230 VAc, o 5W/m2

NCAP LC; Haze0.0750.90;Reectance 0.190.18

Tel.: +3314762 3000 Priva-Lite 55.2(28mm)

1.32.6 0.690.68 0.590.59 NCAP LC; Haze0.0750.90;Reectance 0.230.23

glassinfo@saint-gobain-glass.com,www.sggprivalite.com

LTI Smart Glass Inc.,175 Crystal Straat PO Box, 2305Lenox, MA 01240, USA

110 VAc Liquid crystal device(LC)-Bulletproofprivacy glazing

Tel.: +1 4136375001;fax: +1 4136375004

http://ltisg.com/3.0/switchable.php

Nippon Sheet Glass Co., Ltd.,2-1-7 Kaigan, Minato-ku, Tokyo,105-8552, JAPAN

UMU TM 0.690.12Ref. [78]

UV instable 2060 1C 75 100 VAc,3.5W/m2

NCAP LC forautomotive andbuilding applications

Tel.: +81354439522http://www.nsg.co.jp/en/

Scienstry, Inc.,1110E Collins Blvd.Ste120,Richardson, TX 75081, USA

NPD-LCD Liquid crystal glazing(LC)0.145 up to3.3mmo5up to 10,000O/UTel.: +1 9726905880;

fax: +1 9726905888info@scienstry.us, www.scienstry.us

Innovative Glass Corporation(see SPD-based smart windows)

LC Glass TM 1.6 0.770.56 0.01 2070 1C 88 120 VAc,130mA/m2,15.5W/m2

Liquid crystal device(LC)

www.innovativeglasscorp.com

Commercially available suspended-particle-based smart windows and devicesResearch Frontiers Inc.,240 Crossways Park Drive,Woodbury, 11797 NY, USA

SPD Smart TM 40100 1C idth SPD for automotiveand buildingapplications

Tel.: +1 5163641902;fax: +1 5163643798

http://www.refr-spd.com

SmartGlass International, Ltd.,Unit 21, Cookstown IndustrialEstate, Tallaght, Dublin 24, IRELAND

SPD SmartGlassTM

1.3 0.490.0024 0.0050.005 3106cycles, 12yrs, 20 to50 1C

29065110 VAc,o200mA/m2,5W/m2

Suspended particledevice (SPD)

Tel.: +35314629945;fax: +353 14629951

info@smartglassinternational.com,www.SmartGlassinternational.com

R.Baeten

set

al./SolarEnerg

yMateria

ls&

SolarCells

94(2010

)87105

102Max si

Up to98.6cm2

1203cm2

1003cm2

1002cm2

1802cm2

2604cm2

Max. w150 cm

Up to98.6cm2

ARTICLE IN PRESS

CRICURSACristalesCurvadosS.A.,

CamideCanFerrans/n,Pol.

IndustrialColldela

Maya,08403

Granollers

(Barcelona),SPAIN

CRI-Regulite

A1.29

0.170.004

0.230.11

0.350.21

41,000,000

cycles,30

901C

Upto

100300

cm2

Supply:80220VAc,

Power:o5W/m

2Suspen-

ded

particle

device

(SPD)

Tel.:+34938404470;

fax:+34938401460

B1.29

0.2760.012

0.2690.132

0.390.23

cricursa@cricursa.com,

www.cricu

rsa.com

C1.29

0.3330.029

0.2980.142

0.430.22

D1.29

0.4450.091

0.3440.180

0.480.30

AmericanGlass

ProductsCompany,

AGPEuropeGmbH,Ruhralle9D,

44139Dortmund,GERMANY

VARIO

PlusPolar

0.790.49

0

100000nds

operations,

20651C

Maxim

um

width:1.1m

110220Vac,

16W/m

2Suspendedparticle

device(SPD)H

azing

from

8to

95%

Tel.:+432319525286

VARIO

PlusSky

0.500.04,

0.400.01,

0.100.01

0

Maxim

um

width:1.0m

110220Vac,

5W/m

2Suspendedparticle

device(SPD)

agpeurope@agpglass.com,

http://w

ww.agpglass.com/

PLEOTINTL.C.C.,

7705West

OliveRoad,West

Olive,

MI49460,USA

Sunlight

Responsive

Therm

och

romic

SRT

0.540.04

Therm

och

romic

device

Tel.:+16167388540;

fax:+16167387661

info@pleotint.com,

www.pleotint.com

InnovativeGlass

Corporation,

130NewtonRoadPlainview,NY,

USA

SPDSmartglass

TM

1.6

0.500.01

0.01

30901C

260488

cm2

120VAc,

16mA/m

2,

1.9W/m

2

Suspendedparticle

device(SPD)

Tel.:+15167771100;

fax:+15167771106

info@innovativeglasscorp.com,

www.innovativeglasscorp.com

R. Baetens et al. / Solar Energy Materials & Solar Cells 94 (2010) 87105 103References

[1] S.A. Agnihotry, P. Singh, A.G. Joshi, D.P. Singh, K.N. Sood, K.N. Shivaprasad,Electrodeposited Prussian Blue lms: annealing effect, Electrochimica Acta51 (2006) 42914301.

[2] American Glass Products, 2009, retrieved April 16, 2009, /http://www.agpglass.comwww.agpglass.comS.

[3] A. Anders, J.L. Slack, T.J. Richardson, Electrochromically switched, gas-reservoir metal hydride devices with application to energy-efcientwindows, Thin Solid Films 517 (2008) 10211026.

[4] E. Avendano, A. Azens, G.A. Niklasson, C.G. Graqvist, Electrochromism innickel oxide thin lms containing Mg, Al, Si, V, Zr, Nb, Ag or Ta, Solar EnergyMaterials and Solar Cells 84 (2004) 337350.

[5] E. Avendano, L. Berggren, G.A. Niklasson, C.G. Granqvist, A. Azens,Electrochromic materials and devices: brief survey and new data on opticalabsorption in tungsten oxide and nickel oxide lms, Thin Solid Films 496(2006) 3036.

[6] E. Avendano, A. Azens, G.A. Niklasson, C.G. Granqvist, Sputter depositedelectrochromic lms and devices based on these: progress on nickel-oxide-based lms, Materials Science and Engineering B 138 (2007)112117.

[7] A. Azens, L. Kullman, G. Vaivars, H. Nordberg, C.G. Granqvist, Sputter-deposited nickel oxide for electrochromic applications, Solid State Ionics113115 (1998) 449456.

[8] J. Backholm, A. Azens, G.A. Niklasson, Electrochemical and optical propertiesof sputter deposited IrTa and Ir oxide thin lms, Solar Energy Materials andSolar Cells 90 (2006) 414421.

[9] J. Backholm, E. Avendano, A. Azens, G. deM Azevedo, E. Coronel, G.A. Niklasson,C.G. Granqvist, Solar Energy Materials and Solar Cells 92 (2008) 9196.

[10] J. Backholm, G.A. Niklasson, Optical properties of electrochromic iridiumoxide and iridiumtantalum oxide thin lms in different colouration states,Solar Energy Materials and Solar Cells 92 (2008) 13881392.

[11] K. Bange, Colouration of tungsten oxide lms: a model for optically activecoatings, Solar Energy Materials and Solar Cells 58 (1999) 1131.

[12] S. Bao, Y. Yamada, M. Okada, K. Yoshimura, The effect of polymer coating onswitching behaviour and cycling durability of Pd/MgNi thin lms, AppliedSurface Science 253 (2007) 62686272.

[13] S. Bao, K. Tajima, Y. Yamada, M. Okada, K. Yoshimura, Color-neutralswitchable mirror based on magnesiumtitanium thin lms, AppliedPhysics A 87 (2007) 621624.

[14] S. Bao, K. Tajima, Y. Yamada, M. Okada, K. Yoshimura, Metal buffer layerinserted switchable mirrors, Solar Energy Materials and Solar Cells 92(2008) 216223.

[15] S. Bao, K. Tajima, Y. Yamada, M. Okada, K. Yoshimura, Magnesiumtitaniumalloy switchable mirrors, Solar Energy Materials and Solar Cells 92 (2008)224227.

[16] C. Bechinger, B.A. Gregg, Development of a new self-powered electrochromicdevice for light modulation without external power supply, Solar EnergyMaterials and Solar Cells 54 (1998) 405410.

[17] D. Benson, R. Crandall, S.K. Deb, J.L. Stone, Stand-alone photovoltaic poweredelectrochromic windows, US Patent 5.384.653, January 24, 1995.

[18] J.N. Bullock, C. Bechinger, D.K. Benson, H.M. Branz, Semi-transparant a-SiC:Hsolar cells for self-powered photovoltaic-electrochromic devices, Journal ofNon-Crystalline Solids 198200 (1996) 11631167.

[19] S. Chakrapani, S.M. Slovak, R.L. Saxe, B. Fanning, Research Frontiers USPatent 65416827SPD Films and Light Valves Comprising Same, 2002.

[20] Chromogenics, 2009, retrieved April 16, 2009, /www.chromgenics.seS.[21] Cricursa Cristales Curvados, 2009, retrieved April 16, 2009, /www.cricursa.

comS.[22] D. Cupelli, F.P. Nicoletta, S. Manfredi, G. De Filpo, G. Chidichimo, Electrically

switchable chromogenic materials for external glazing, Solar EnergyMaterials and Solar Cells 93 (2009) 329333.

[23] K.C. Chen, F.R. Chen, J.-J. Kai, Electrochromic property of nano-compositePrussian Blue based thin lm, Electrochimica Acta 52 (2007) 33303335.

[24] C.-Y. Chen, Y.-L. Lo, Integration of a-Si:H solar cell with novel twist nematicliquid crystal cell for adjustable brightness and enhanced power character-istics, Solar Energy Materials and Solar Cells doi:10.1016/j.solmat.2009.01.017.

[25] S.K. Deb, S.-H. Lee, C.E. Tracy, J.R. Pitts, B.A. Gregg, H.M. Branz, Stand-alonephotovoltaic-powered electrochromic smart window, Electrochimica Acta46 (2001) 21252130.

[26] S.K. Deb, Opportunities and challenges in science and technology of WO3 forelectrochromic and related applications, Solar Energy Materials and SolarCells 92 (2008) 245258.

[27] M. Deepa, S. Ahmad, K.N. Sood, J. Alam, S. Ahmad, A.K. Srivastava,Electrochromic properties of polyaniline thin lm nanostructures derivedfrom solutions of ionic liquid/polyethylene glycol, Electrohimica Acta 52(2007) 74537463.

[28] J.W. Doane, G. Chidichimo