Electrochromic performance of nanocomposite nickel oxide counterelectrodes containing lithium and zirconium

Feng Lin a,b, Manuel Montano c, Chixia Tian c, Yazhou Ji b, Dennis Nordlund d,Tsu-Chien Weng d, Rob G. Moore e, Dane T. Gillaspie a, Kim M. Jones a, Anne C. Dillon a,Ryan M. Richards b,c, Chaiwat Engtrakul a,n

a National Renewable Energy Laboratory, Golden, CO 80401, USAb Materials Science Program, Colorado School of Mines, Golden, CO 80401, USAc Department of Chemistry and Geochemistry, Colorado School of Mines, Golden, CO 80401, USAd Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USAe Stanford Institute for Materials and Energy Sciences, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA

a r t i c l e i n f o

Available online 2 December 2013

Keywords:Nickel oxideNanocompositeOxidation stateElectrochromicLi stoichiometry

a b s t r a c t

Nickel oxide materials are suitable for counter electrodes in complementary electrochromic devices.The state-of-the-art nickel oxide counter electrode materials are typically prepared with multipleadditives to enhance peformance. Herein, nanocomposite nickel oxide counter electrodes werefabricated via RF magnetron co-sputtering from Ni–Zr alloy and Li2O ceramic targets. The as-depositednanocomposite counter electrodes were characterized with inductively coupled plasma mass spectro-metry (ICP-MS), scanning electron microscopy (SEM), high-resolution transmission electron microscopy(HRTEM), X-ray diffraction (XRD), and X-ray absorption spectroscopy (XAS). It was found that thestoichiometry, crystal structure and electronic structure of the nickel oxide-based materials could bereadily tuned by varying the Li2O sputter deposition power level. Comprehensive electrochromicevaluation demonstrated that the performance of the nickel oxide-based materials was dependent onthe overall Li stoichiometry. Overall, the nanocomposite nickel oxide counter electrode containinglithium and zirconium synthesized with a Li2O deposition power of 45 W exhibited the optimalperformance with an optical modulation of 71% and coloration efficiency of 30 cm2/C at 670 nm inLi-ion electrolyte.

& 2014 Elsevier B.V. All rights reserved.

1. Introduction

Electrochromic effects in transition metal oxide materials (e.g.,nickel oxide, tungsten oxide, titanium oxide) have received extensiveattention for smart windows, rear-view mirrors and non-emissivedisplays. A cathode/anode complementary electrode configuration inelectrochromic devices is the most prevalent configuration due to theadvantages this configuration offers, i.e., color neutrality and colorationefficiency [1,2]. Tungsten oxide and nickel oxide are the mostinvestigated electrochromic cathodic and anodic materials, respec-tively [2,3]. The successful operation of a complementary electrochro-mic device requires compatibility between the cathodic and anodicelectrodes. The critical electrode compatibility requirements include abalanced charge capacity [4–6] and equivalent switching kinetics [7].Since the 1970s, extensive efforts have been devoted to improve theperformance and mechanistic understanding of the electrochromic

process in cathodic tungsten oxide materials [8–11]. Further researchhas been directed towards developing cost effective syntheses andmanufacturing processes for these cathodic materials [2,12]. Conver-sely, many challenges remain unresolved for anodic nickel oxidecounter electrodes, including slow switching kinetics, inferior opticalmodulation, and poor bleached-state transparency [13–16].

In recent years, significant efforts have been dedicated tooptimize the performance of nickel oxide counter electrodematerials, with the emphasis on controlling morphology on thenanoscale (nanocomposites) [1,7,17], crystal structure (latticeorder, crystallinity) [17,18] and electronic structure [19]. The mostpromising nickel oxide counter electrodes are composed of com-plex mixtures of transition metal oxides including secondaryadditives such as lithium oxide and lithium peroxide [20].Recently, multicomponent nickel oxide counter electrodes (i.e.,nickel oxide materials that contain at least two additives) haveshown superior performance relative to the traditional nanocom-posite nickel oxide electrodes that contain only one additive[1,20]. The addition of co-additives (i.e., a transition metal and

Contents lists available at ScienceDirect

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

Solar Energy Materials & Solar Cells

0927-0248/$ - see front matter & 2014 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.solmat.2013.11.023

n Corresponding author. Tel.: +1 303 384 6646; fax: +1 303 384 6655.E-mail address: chaiwat.engtrakul@nrel.gov (C. Engtrakul).

Solar Energy Materials & Solar Cells 126 (2014) 206–212

lithium) to the nickel oxide electrode material enables simulta-neous improvement of bleached-state transparency, switchingkinetics, and optical modulation [1,19,20]. In addition, the nickeloxidation state and hole concentration in nickel oxide-basedelectrochromic materials are critical for favorable electrochromicperformance and relevant specifically to the coloration/bleachingmechanism [19,20]. In our previous study, we demonstrated thatlithium is an effective additive for tuning the nickel oxidation stateand hole concentration in nickel oxide materials containingaluminum [19]. However, the electrochromic performance (e.g.,optical modulation and bleached-state transparency) of these Al-containing nickel oxide nanocomposites was modest compared tothe performance reported for nickel oxide materials containinglithium and tungsten [1] or lithium and zirconium [20] additives.The nickel oxide nanocomposite materials containing lithium andzirconium exhibited extremely high electrochromic performancerelative to previous electrochromic transition metal oxide materi-als [20]. However, it is unclear whether or not the hole concentra-tion in these high performing nickel oxide materials containinglithium and zirconium can be optimized further to improveelectrochromic performance.

Herein, we extend our previous studies by systematicallytuning the Li stoichiometry in radio frequency (RF) magnetronco-sputtering deposition of nanocomposite nickel oxide materials.It is found that the sputtering conditions (i.e., deposition powerlevels) provide an efficient route for tuning the Li stoichiometry,crystal structure, and electronic structure in nanocomposite nickeloxide materials. The chemical, structural, and electronic propertiesof these nickel oxide-based materials are characterized in detailand an optimal lithium concentration is determined.

2. Experimental methods

2.1. Electrode preparation

RF magnetron co-sputtering was preformed on an AngstromEvoVac deposition system housed in a glove box under an argonatmosphere following a previously described method (seeScheme 1) [19]. Three-inch diameter metal alloy target, Ni–Zr(80–20 at%), was purchased from ACI Alloys, while a three-inchdiameter Li2O ceramic target (99.9%) supported on a molybdenumbacking plate was purchased from Plasmaterials, Inc. The deposi-tion power level for the metal alloy target was fixed at 60 W, whilethe deposition power level for the Li2O ceramic target was

adjusted to 0 W, 15 W, 30 W, 45 W and 60 W to control the Listoichiometry. The resulting nickel oxide-based samples aredenoted as Li2O/0 W, Li2O/15 W, Li2O/30 W, Li2O/45 W and Li2O/60 W according to the Li2O deposition power level. The target–substrate distance was 10 cm and remained constant throughoutthe study, and no additional heating was applied to the substrate.The substrate holder was rotated during the sputtering process.The base pressure and total deposition pressure were 10�7 Torrand 2 mTorr, respectively. The Ar/O2 gas mixture was fixed at 1/2throughout the study. The fluorine-doped tin oxide (FTO) glasssubstrates were purchased from Hartford Glass CO, Inc. (TEC 15,1.5 in.�0.82 in.�2.3 mm). The substrates were cleaned succes-sively with soapy water, deionized water, acetone, isopropanol,and dried under flowing N2. ICP-MS and XRD samples weredeposited on aluminum foil and microscope glass, respectively.

2.2. Materials characterization

The crystal structures of the resulting films were characterizedon a Philips X-ray diffractometer Model PW1729 operated at 45 kVand 40 mA using CuKα radiation. The samples were prepared fortransmission electron microscopy (TEM) using a Nano-Lab 200Dual Beam FIB and analyzed on a FEI G2-T30 TEM operating at300 kV. The ICP-MS used is a Perkin Elmer NexION 300q with anS10 autosampler. Calibration standards for Li, Ni, and Zr weremade using SPEX Certiprep ICP-MS standard solutions in 2% HCl(Optima trace metal grade). Dilutions of thin film digest solutionswere made in 2% HCl. All dilutions were made in 15 mL Falconpolypropylene centrifuge tubes and analyzed immediately. SEMwas done on a JEOL JSM-7000F Field Emission Scanning ElectronMicroscope with an EDAX Genesis EDS. Soft X-ray absorptionspectroscopy (XAS) measurements were performed on the 31-polewiggler beam line 10-1 at Stanford Synchrotron Radiation Light-source (SSRL) using a ring current of 350 mA and a 1000 l mm�1

spherical grating monochromator with 20 μm entrance and exitslits, providing �1011 ph s�1 at 0.2 eV resolution in a 1 mm2 beamspot. During the measurements, all samples were attached to analuminum sample holder and the surface was connected to theisolated holder using conductive carbon. Data were acquired in asingle load at room temperature and under ultra-high vacuum(10�9 Torr). Detection was performed in total electron yield (TEY)and fluorescence yield (FY) modes. For TEY, we collected thesample drain current, and a silicon diode (IRD AXUV-100) wasused to collect the FY positioned near the sample surface.Contributions from visible light were carefully minimized beforethe acquisition, and all spectra were normalized by the currentfrom freshly evaporated gold on a fine grid positioned upstream ofthe main chamber.

2.3. Electrochromic performance evaluation

Electrochromic properties were measured in a liquid electro-lyte half-cell where the electrolyte (Novolyte Technologies, part ofthe BASF Group) was 1 M lithium perchlorate (LiClO4) dissolved inpropylene carbonate (PC). Cyclic voltammetry (CV) was carried outusing a BioLogic VMP3 multichannel potentiostat with a scan rateof 20 mV/s and a voltage range of 1.7–4.2 V vs. Li/Liþ , and theinitial voltage was set at the open circuit voltage. In situ transmit-tance was measured using a diode laser at 670 nm. Switchingkinetics (i.e., coloration and bleaching) was measured underpotential step cycling from 1.7 to 4.2 V vs. Li/Liþ , where eachpotential step was maintained for 2 min. The switching speed isdefined as the time required to achieve �90% of total transmit-tance change within a potential step. Both switching kinetics andCV measurements were performed for 200 cycles. All electroche-mical measurements were carried out under an argon atmosphere

Scheme 1. Schematic representation of RF magnetron co-sputtering chamber,where the relative Li2O deposition power level was adjusted to control the Listoichiometry in the nickel oxide-based materials.

F. Lin et al. / Solar Energy Materials & Solar Cells 126 (2014) 206–212 207

in a glove box. The samples were transferred from the sputteringchamber to testing cells without exposure to air or moisture.

3. Results and discussion

The stoichiometric information of the as-deposited counterelectrode materials was quantified using ICP-MS and shown inTable 1 and Fig. 1. The atomic ratio between Li and Ni (Li/Ni)generally increases with the increase of Li2O deposition power,while no obvious difference in the Li/Ni ratio is observed betweenLi2O/45 W and Li2O/60 W. The atomic ratio between Zr and Ni isfairly stable when the Li2O deposition power was varied. We notethat the stoichiometry of Li2O/45 W electrode is slightly differentfrom our previous study under similar co-sputtering condition[20]. The difference is attributed to the extended sputtering timein the present study. Based on the ICP-MS study, we conclude thatvarying the Li2O deposition power successfully modifies the Li

stoichiometry in the as-deposited nanocomposite nickel oxidematerials.

Electrochromic properties of transition metal oxide materialsare influenced by their crystallinity (e.g., WO3 and NiO)[11,17,19,21]. Highly crystalline phases tend to lead to slowswitching kinetics and inferior optical modulation, while amor-phous phases compromise long-term durability. Previous studiesshowed that a mixture of crystalline and amorphous phases isbeneficial for the overall electrochromic performance of the activematerials [1,11]. The crystallinity of the as-deposited nanocompo-site nickel oxide materials was studied by XRD and shown in Fig. 2.Overall, the as-deposited electrodes have extremely poor crystal-linity. The crystallinity of nickel oxide materials is slightlyimproved as the Li2O deposition power level is increased from0W to 45 W and then decreases as the Li2O power level is furtherramped up to 60 W. The Li2O/45 W electrode is composed ofmixture of crystalline and amorphous phases, which is a prere-quisite for excellent electrochromic performance according toprevious reports [1,11]. Fig. 3 shows the cross-sectional SEM andHRTEM images for the Li2O/45 W electrode. The thickness of thenanocomposite nickel oxide layer is �200 nm. Strong adhesion isobserved between the FTO and nanocomposite nickel oxide layers,which is beneficial for electron transport between the FTO andexternal circuit. Furthermore, the nanocomposite nickel oxidelayer is less dense compared to the nanocomposite nickel oxideelectrodes containing lithium and aluminum [19], which mayaccount for the enhanced performance of the nanocompositenickel oxide electrodes in the present study. The HRTEM imagein Fig. 3b shows that the nanocomposite nickel oxide material iscomposed of a mixture of crystalline and amorphous phases.Similar crystalline/amorphous morphology was previouslyobserved in nickel oxide-based electrochromic materials contain-ing complex additives [1,17,20]. The nickel oxide crystallite size isca. 5 nm and consistent with the XRD results in Fig. 2.

The nickel oxidation state and hole concentration in nickeloxide-based electrochromic materials is critical for favorableelectrochromic performance and relevant specifically to the col-oration/bleaching mechanism [19,20]. Therefore, a XAS study was

Table 1Stoichiometric information for the as-deposited nanocomposite nickel oxide electrochromic electrodes determined by ICP-MS.

Li2O/0W Li2O/15W Li2O/30W Li2O/45W Li2O/60 W

Li/Ni 0.23070.017 0.66170.052 1.96670.013 2.07170.097Zr/Ni 0.20670.024 0.21070.014 0.24070.011 0.23170.009 0.21570.014Formula NiZr0.206Ox Li0.23NiZr0.21Ox Li0.661NiZr0.24Ox Li1.966NiZr0.231Ox Li2.071NiZr0.215Ox

Fig. 1. The atomic ratios for (a) Li/Ni and (b) Zr/Ni in the as-deposited nanocom-posite nickel oxide electrochromic electrodes.

Fig. 2. XRD patterns for the as-deposited nanocomposite nickel oxide electrodes,where the crystal indices are indicated.

F. Lin et al. / Solar Energy Materials & Solar Cells 126 (2014) 206–212208

performed to investigate the influence of the Li stoichiometry onthe nickel oxidation state and hole concentration in the nickeloxide nanocomposites containing Li and Zr. In general, Ni L-edgeXAS is sensitive to the nickel oxidation state and hole concentra-tion in nickel oxide-based materials [22,23]. Depending on thedetection mode of XAS, the technique can be sensitive to thesurface (TEY mode) or bulk (FY mode) properties of the material[24]. Recording TEY and FY signals allows for a thorough under-standing of the electronic properties of our nanocomposite nickeloxide thin films. Fig. 4 displays the Ni L-edge XAS spectra for theas-deposited nanocomposite nickel oxide thin films in the TEYmode. We note that the signal-to-noise ratio is lower in thenanocomposite nickel oxide electrodes with high Li stoichiometry(i.e., Li2O/45 W and Li2O/60 W) than those with low Li stoichio-metry (i.e., Li2O/0 W, Li2O/15 W, and Li2O/30 W) due to theformation of a lithium peroxide rich surface layer [20]. The featurecentered at 855 eV is assigned to the Ni LIII (2p3/2) edge absorptionand the feature centered at 872.5 eV is due to Ni LII (2p1/2) edgeabsorption. Spin–orbit effects cause the separation observed forthese two features (LII and LIII) and further splitting of thesefeatures has been assigned to the Ni2p–Ni3d electrostatic interac-tion and crystal field effects [25]. An increase in the intensity of theshoulder at higher energy of the Ni LIII edge absorption

(highlighted in Fig. 4) is indicative of an increase in the formalnickel oxidation state (i.e., increase in hole concentration) in thenickel oxide electrode materials [23]. The intensity of this shoulderintensifies as the lithium concentration increases (i.e., Li2O sputterpower level increases) inferring that the addition of lithium affectsthe nickel oxidation state and hole concentration in the electrode.Furthermore, the XAS spectra in FY mode (bulk sensitive, Fig. 5)show similar results as the surface sensitive TEY mode (Fig. 4).These Ni L-edge spectra (Figs. 4 and 5) are consistent withprevious reports [19,22,23]. Therefore, we conclude that theaddition of lithium increases the nickel oxidation state (Ni2þ toNi3þ) and hole concentration throughout the entire nanocompo-site nickel oxide film. This synthetic approach can be easilyimplemented for tuning the structural and electronic propertiesof the nanocomposite nickel oxide materials.

The electrochromic performance of the resulting nanocompositenickel oxide electrodes was analyzed in a Li-ion electrolyte dis-solved in propylene carbonate. Fig. 6 summarizes the in situ opticalmodulation results using a CV cycling technique at the wavelengthof 670 nm. A noticeable activation period is observed for thebleached-state transmittance (upper curve) for the studied electro-des. Optical modulation is directly related to the Li stoichiometry in

Fig. 3. (a) Cross-sectional SEM image of the Li2O/45 W electrode, where the glass/FTO and FTO/nickel oxide boundaries were indicated by dashed lines; and(b) HRTEM image for the same electrode, where the crystalline nickel oxide phasesare circled.

Fig. 4. Ni LII,III-edge X-ray absorption spectra for the as-deposited nanocompositenickel oxide electrodes in the total electron yield (TEY) mode, where the spectrahave been normalized to the most intense peaks and offset for clarity.

Fig. 5. Ni LII,III-edge X-ray absorption spectra for the as-deposited nanocompositenickel oxide electrodes in the fluorescence yield (FY) mode, where the spectra havebeen normalized to the most intense peaks and offset for clarity.

F. Lin et al. / Solar Energy Materials & Solar Cells 126 (2014) 206–212 209

the nanocomposite nickel oxide electrodes. Here, the optical mod-ulation of the nickel oxide-based materials is compromised if thelithium concentration is low or high (i.e., Li2O/0 W, Li2O/15W, andLi2O/60 W). The electrochromic performance of the Li2O/30W andLi2O/45 W samples is enhanced relative to the performance pre-viously reported for nickel oxide-based nanocomposite materials[1,7]. Colored and bleached states are improved for the Li2O/45Welectrode relative to the corresponding states for the Li2O/30Welectrode. We attribute the improvement in the colored state to theincrease in hole doping in the nickel oxide material [19]. However,this reasoning does not explain the superior bleached stateobserved for the Li2O/45W electrode. The bleached-state transpar-ency of nickel oxide is known to be extremely dependent on theaddition of additives (e.g., Al, Zr, and W) [26]. However, in our casethere is only a minor difference in the Zr additive concentration forthe various electrodes (Table 1) indicating that the enhancedbleached-state transparency is independent of the Zr/Ni ratio. Sincethe nickel oxide bleached-state transparency is related to the extentof hole removal during the Liþ/e� intercalation [20], it is feasiblethat a mixture of crystalline and amorphous phases (containinglithium rich species) is beneficial for hole removal [1,14]. Therefore,we suggest that the complex crystalline/amorphous morphology ofthe nanocomposite nickel oxide materials imparts to it an attributethat allows for the efficient removal of holes and enhances thebleached-state transparency.

The evolution of charge/discharge capacities during long-termelectrochemical cycling provides useful information to determinethe activation and deactivation of electrochromic materials. Fig. 7depicts the charge/discharge capacities for the electrodes cycledusing CV. There is a minimal activation period for the charge/discharge capacities for the Li2O/30 W electrode. Conversely, theLi2O/60 W electrode requires more than 200 cycles for an activa-tion period. Both of these electrodes are stable in terms ofmaintaining capacities. Interestingly, the Li2O/45 W electrodeexperiences a shorter activation period (�50 cycles) and thendegrades significantly to 88% of its maximum capacity after 200cycles. However, this degradation does not adversely affect theoptical modulation observed in Fig. 6.

The charge reversibility (%R) for an electrochemical cycle is ofimportance for accessing the long-term electrochromic perfor-mance of a material. Accumulation of irreversible Liþ/e� inter-calation products can result in overall performance degradation,

including decreasing optical modulation and switching kinetics.The %R was calculated and shown in the inset of Fig. 7. All of thestudied electrodes maintain extremely high %R (499%) and aresuperior to high performing porous tungsten oxide electrodes [11].

Coloration efficiency accounts for the amount of charge neces-sary to produce a change in optical density (ΔOD). In practice, highcoloration efficiency is desirable to enable significant energysavings during the electrochromic switching process. Fig. 8 showsthe dependence of coloration efficiency on the CV cycle number.Overall, the coloration efficiency is competitive to the previouslyreported value for a nanocomposite Li1.2Ni0.1WOx electrode [1].The Li2O/30 W electrode exhibits the highest coloration efficiencyamong the studied electrodes, which can be attributed to thesmallest charge capacity observed in the Li2O/30 W electrode(Fig. 7). The Li2O/45 W and Li2O/60 W electrodes exhibit improvedcoloration efficiency with increasing CV cycle number. This con-tinuous improvement in coloration efficiency was similar to thatobserved by Gillaspie et al. for nanocomposite nickel oxideelectrodes containing lithium and tungsten and was attributed tothe decrease of lithium loss in noncoloring processes [1].

The coloration and bleaching kinetics were studied for theselected nanocomposite nickel oxide electrodes and summarizedin Fig. 9. All of the electrodes exhibit similar trends with increasing

Fig. 6. Dependence of in situ transmittance on the CV cycle number, where thesame color curves represent the respective colored and bleached states for theelectrodes synthesized under varying Li2O sputter power levels. The differencebetween the upper and lower curves is interpreted as the optical modulation. Thearrow illustrates the optical modulation at the 110th cycle for a nanocompositenickel oxide electrodes with high Li stoichiometry. The in situ transmittance istaken at 670 nm.

Fig. 7. Dependence of charge and discharge capacities on the CV cycle number forthe selected nanocomposite nickel oxide electrodes, where (þ) and (�) representcharge and discharge capacities, respectively. The inset shows the charge reversi-bility (%R) of the Liþ/e� intercalation/deintercalation process.

Fig. 8. Dependence of coloration efficiency on the CV cycle number for the selectednanocomposite nickel oxide electrodes.

F. Lin et al. / Solar Energy Materials & Solar Cells 126 (2014) 206–212210

cycle number. The bleaching kinetics is significantly improvedwhile the coloration kinetics degrades with the increase in cyclenumber. The Li2O/45 W electrode exhibits slightly enhancedbleaching kinetics compared to the Li2O/30 W and Li2O/60 Welectrodes, while the Li2O/45 W and Li2O/30 W electrodes exhibitenhanced coloration kinetics compared to the Li2O/60 W samples.These observed switching kinetics are improved relative to nickeloxide nanocomposite electrodes synthesized utilizing wet-chemical methods and tested under similar condition [7].

4. Conclusions

Nanocomposite nickel oxide counter electrodes were fabricatedvia RF magnetron co-sputtering using Ni–Zr alloy and Li2O ceramictargets. The Li2O sputter power level was tuned to control the Listoichiometry in the resulting nanocomposite nickel oxide counterelectrodes containing lithium and zirconium. Controlling the Listoichiometry (i.e., nickel oxidation state, hole concentration) wascritical for obtaining superior nanocomposite nickel oxide electro-des. The optimal electrochromic performance was observed withthe electrode composition of Li1.966NiZr0.231Ox (Li2O/45 W). Theelectrochromic performance data for the various nickel oxide-based materials is summarized in Table 2 and clearly demonstratesthat the electrochromic performance is relevant to the Li stoichio-metry in the nanocomposite nickel oxide electrodes. The Li2O/45 W electrode exhibits the most favorable electrochromic perfor-mance in terms of optical modulation, bleached-state transpar-ency, coloration efficiency and switching kinetics. We anticipatethat this understanding will provide further insight into thedevelopment of general nanocomposite electrochromic materials.

Acknowledgments

This paper is dedicated to the loving memory of Anne C. Dillon.This research was supported by the U.S. Department of Energyunder Contract number DE-AC36-08-GO28308 with the NationalRenewable Energy Laboratory as part of the DOE Office of EnergyEfficiency and Renewable Energy Office of Building TechnologiesProgram. Portions of this research were carried out at the StanfordSynchrotron Radiation Laboratory, a national user facility operatedby Stanford University on behalf of the U.S. Department of Energy,Office of Basic Energy Sciences. F. Lin and C. Engtrakul would liketo acknowledge D. Weir, N. Sbar, and J.-C. Giron (Sage Electro-chromics) for precious discussion.

References

[1] D. Gillaspie, A. Norman, C.E. Tracy, J.R. Pitts, S.H. Lee, A. Dillon, Nanocompositecounter electrode materials for electrochromic windows, J. Electrochem. Soc.157 (2010) H328–H331.

[2] D.T. Gillaspie, R.C. Tenent, A.C. Dillon, Metal-oxide films for electrochromicapplications: present technology and future directions, J. Mater. Chem. 20(2010) 9585–9592.

[3] G.A. Niklasson, C.G. Granqvist, Electrochromics for smart windows: thin filmsof tungsten oxide and nickel oxide, and devices based on these, J. Mater. Chem.17 (2007) 127–156.

[4] A. Azens, L. Kullman, C.G. Granqvist, Ozone coloration of Ni and Cr oxide films,Sol. Energy Mater. Sol. Cells 76 (2003) 147–153.

[5] J. Padilla, Increasing performance of electrochromic devices by modifyingsurface and charge ratio between electrodes, Sol. Energy Mater. Sol. Cells 99(2012) 56–61.

[6] C.Y. Kim, S.G. Cho, T.Y. Lim, Cycle test and degradation analysis of WO(3)/PCþLiClO(4)/CeO(2)center dot TiO(2) electrochromic device, Sol. EnergyMater. Sol. Cells 93 (2009) 2056–2061.

[7] R.C. Tenent, D.T. Gillaspie, A. Miedaner, P.A. Parilla, C.J. Curtis, A.C. Dillon, Fast-switching electrochromic Liþ-doped nio films by ultrasonic spray deposition,J. Electrochem. Soc. 157 (2010) H318–H322.

[8] F. Lin, C.-P. Li, G. Chen, R.C. Tenent, C.A. Wolden, D.T. Gillaspie, A.C. Dillon,R.M. Richards, C. Engtrakul, Low-temperature ozone exposure technique tomodulate the stoichiometry of WOx nanorods and optimize the electrochro-mic performance, Nanotechnology 23 (2012) 255601.

[9] J. Zhang, X.L. Wang, X.H. Xia, C.D. Gu, J.P. Tu, Electrochromic behavior of WO(3) nanotree films prepared by hydrothermal oxidation, Sol. Energy Mater. Sol.Cells 95 (2011) 2107–2112.

[10] T. Brezesinski, D. Fattakhova-Rohlfing, S. Sallard, M. Antonietti, B.M. Smarsly,Highly crystalline WO(3) thin films with ordered 3D mesoporosity andimproved electrochromic performance, Small 2 (2006) 1203–1211.

[11] F. Lin, J.F. Cheng, C. Engtrakul, A.C. Dillon, D. Nordlund, R.G. Moore, T.C. Weng,S.K.R. Williams, R.M. Richards, In situ crystallization of high performing WO3-based electrochromic materials and the importance for durability and switch-ing kinetics, J. Mater. Chem. 22 (2012) 16817–16823.

[12] C.P. Li, F. Lin, R.M. Richards, C. Engtrakul, A.C. Dillon, R.C. Tenent, C.A. Wolden,Ultrasonic spray deposition of high performance WO3 films using template-assisted sol–gel chemistry, Electrochem. Commun. 25 (2012) 62–65.

[13] I. Bouessaya, A. Rougier, P. Poizot, J. Moscovici, A. Michalowicz, J.M. Tarascon,Electrochromic degradation in nickel oxide thin film: a self-discharge anddissolution phenomenon, Electrochim. Acta 50 (2005) 3737–3745.

[14] S.H. Lee, Y.S. Park, S.K. Joo, Characterization of Ni–Woxide thin film electrodes,Solid State Ion. 109 (1998) 303–310.

[15] N. Penin, A. Rougier, L. Laffont, P. Poizot, J.M. Tarascon, Improved cyclability bytungsten addition in electrochromic NiO thin films, Sol. Energy Mater. Sol.Cells 90 (2006) 422–433.

[16] J. Smulko, A. Azens, R. Marsal, L.B. Kish, S. Green, C.G. Granqvist, Application of1/f current noise for quality and age monitoring of electrochromic devices, Sol.Energy Mater. Sol. Cells 92 (2008) 914–918.

[17] H. Moulki, D.H. Park, B.-K. Min, H. Kwon, S.-J. Hwang, J.-H. Choy, T. Toupance,G. Campet, A. Rougier, Improved electrochromic performances of NiO basedthin films by lithium addition: from single layers to devices, Electrochim. Acta74 (2012) 46–52.

[18] T. Kubo, Y. Nishikitani, Y. Sawai, H. Iwanaga, Y. Sato, Y. Shigesato, Electro-chromic properties of LixNiyO films deposited by RF magnetron sputtering,J. Electrochem. Soc. 156 (2009) H629–H633.

[19] F. Lin, D. Nordlund, T.-C. Weng, R.G. Moore, D.T. Gillaspie, A.C. Dillon,R.M. Richards, C. Engtrakul, Hole doping in Al-containing nickel oxidematerials to improve electrochromic performance, ACS Appl. Mater. Interfaces5 (2013) 301–309.

[20] F. Lin, D. Nordlund, T.-C. Weng, D. Sokaras, K.M. Jones, R.B. Reed, D.T. Gillaspie,D. Weir, R.G. Moore, A.C. Dillon, R.M. Richards, C. Engtrakul, Origin ofelectrochromism in high-performing nanocomposite nickel oxide, ACS Appl.Mater. Interfaces 5 (2013) 3643–3649.

Fig. 9. Switching kinetics of the selected nanocomposite nickel oxide electrodes asa function of cycle number, where the open and solid dots represent the bleachingand coloration kinetics, respectively.

Table 2Electrochromic characteristics for the nanocomposite nickel oxide electrodes,where the average values were calculated for charge capacity (CC), reversibility(%R), coloration efficiency (CE), bleaching kinetics (tb), and coloration kinetics (tc)by averaging 200 cycles. The optical modulation was calculated after the 50th(ΔT50), 100th (ΔT100) and 200th (ΔT200) CV cycle.

Electrodes CC/mC %R CE/cm2 C�1 tb/s tc/s ΔT50 ΔT100 ΔT200

Li2O/30 W 44.1 99.4 30.8 15.1 53.6 51.2% 55.8% 56.4%Li2O/45 W 64.4 99.6 29.8 13.6 56.4 68.1% 70.7% 68.3%Li2O/60 W 59.0 99.5 26.4 15.1 62.5 29.2% 39.2% 51.6%

F. Lin et al. / Solar Energy Materials & Solar Cells 126 (2014) 206–212 211

[21] S.H. Lee, R. Deshpande, P.A. Parilla, K.M. Jones, B. To, A.H. Mahan, A.C. Dillon,Crystalline WO3 nanoparticles for highly improved electrochromic applica-tions, Adv. Mater. 18 (2006) 763–766.

[22] J. Vanelp, B.G. Searle, G.A. Sawatzky, M. Sacchi, Ligand hole induced symmetrymixing of d8 states in LixNi1�xO, as observed in Ni 2p X-ray absorption-spectroscopy, Solid State Commun. 80 (1991) 67–71.

[23] M.A. Vanveenendaal, G.A. Sawatzky, Doping dependence of Ni 2p X-ray-absorptionspectra of M(x)Ni(1�x)O (M¼Li,Na), Phys. Rev. B 50 (1994) 11326–11331.

[24] H.K. Jeong, H.J. Noh, J.Y. Kim, L. Colakerol, P.A. Glans, M.H. Jin, K.E. Smith,Y.H. Lee, Comment on “near-edge X-ray absorption fine-structure investiga-tion of graphene”, Phys. Rev. Lett. 102 (2009) 099701.

[25] W.S. Yoon, K.Y. Chung, J. McBreen, D.A. Fischer, X.Q. Yang, Changes inelectronic structure of the electrochemically Li-ion deintercalated LiNiO2

system investigated by soft X-ray absorption spectroscopy, J. Power Sources163 (2006) 234–237.

[26] E. Avendano, A. Azens, G.A. Niklasson, C.G. Granqvist, Electrochromism innickel oxide films containing Mg, Al, SiV, Zr, Nb, Ag, or Ta, Sol. Energy Mater.Sol. Cells 84 (2004) 337–350.

F. Lin et al. / Solar Energy Materials & Solar Cells 126 (2014) 206–212212